The Bullpen Has Blown 7 Games for Trachsel

[Frobby]

Here are the sweepstakes winners!

Trachsel 7
Bedard 4
Guthrie 4
Loewen, Cabrera, Burres, Shuey, Bell & Bradford 1 each

I guess Trachsel has been paid back in spades for his lucky 15-win total a year ago.

[os1971]

Here are the sweepstakes winners!

Trachsel 7
Bedard 4
Guthrie 4
Loewen, Cabrera, Burres, Shuey, Bell & Bradford 1 each

I guess Trachsel has been paid back in spades for his lucky 15-win total a year ago.

The way he's pitched, its hard to believe we can't move him.

[OregonBird]

I suppose most teams don't want to part w/young, CHEAP talent the way player salaries are skyrocketing every year.

[DrungoHazewood]

I suppose most teams don't want to part w/young, CHEAP talent the way player salaries are skyrocketing every year.

And the Orioles apparently won't take a random AA arm for him. Oh well, maybe he'll net us a compensation pick if he signs elsewhere.

[Frobby]

And the Orioles apparently won't take a random AA arm for him. Oh well, maybe he'll net us a compensation pick if he signs elsewhere.

I think the Orioles are justified if they decide to exercise the $4.75 mm option they have on Trachsel. He has been good a lot more often than he has been bad this year. We don't know what to expect from most of our young guys, and we need a viable alternative of one or more of them isn't ready to step up, or gets hurt. Trachsel gets no respect, and you can make all the K/BB, xFIP, DIPS and age arguments you want, but he has done a good job this year and has been a consistent performer for a very long time.

[tim27]

I think the Orioles are justified if they decide to exercise the $4.75 mm option they have on Trachsel. He has been good a lot more often than he has been bad this year. We don't know what to expect from most of our young guys, and we need a viable alternative of one or more of them isn't ready to step up, or gets hurt. Trachsel gets no respect, and you can make all the K/BB, xFIP, DIPS and age arguments you want, but he has done a good job this year and has been a consistent performer for a very long time.

I'm starting to agree with this as well. The only problem is he should be behind glass next year which reads "Break in case of emergency." He shouldn't get a single start next year over a guy like Olson, Liz, or Penn. I have no problem keeping him around he's performed admirably this year, but I just worry that if the O's did they'd run him out there every fifth day for his veteranosity.

P.S. He's worth more than some random AA arm.

[DrungoHazewood]

I think the Orioles are justified if they decide to exercise the $4.75 mm option they have on Trachsel. He has been good a lot more often than he has been bad this year. We don't know what to expect from most of our young guys, and we need a viable alternative of one or more of them isn't ready to step up, or gets hurt. Trachsel gets no respect, and you can make all the K/BB, xFIP, DIPS and age arguments you want, but he has done a good job this year and has been a consistent performer for a very long time.

As ugly as Trachsel's peripherals have been this year, the records of players like him in year X+1 are downright hideous.

Here's everyone in the expansion era over 35, with 140+ IP, < 3.5 K/9, and > 3.5 BB/9:


1961-2006

AGE >= 35
INNINGS PITCHED >= 140
STRIKEOUTS/9 IP <= 3.5
WALKS/9 IP >= 3.5

RSAA YEAR RSAA IP SO/9 IP BB/9 IP
1 Phil Niekro 1986 -4 210.1 3.47 4.06
2 Claude Osteen 1975 -10 204 2.78 4.06
3 Bob Knepper 1989 -30 165 3.49 4.09

Their ERAs the next year were 6.30, out of majors, and 5.68 (in 44 innings). They were all retired before year X+2.

If you lower the innings threshold to 100 you get Mike Cuellar's last year in Baltimore. The next season he pitched three innings, allowed seven runs, and retired. You also get Bob Forsch in '89, and he retired immediately after the year. And you get Wilbur Wood in 1977. In '78 he actually pitched 168 innings, but to a 5.20 ERA in a league with a 3.68.

What Trachsel is doing right now is at the very edge of what's possible. Continuing that next year would be historically unique. If they pick up his option I'll be bitterly disappointed.

[Frobby]

As ugly as Trachsel's peripherals have been this year, the records of players like him in year X+1 are downright hideous.

Here's everyone in the expansion era over 35, with 140+ IP, < 3.5 K/9, and > 3.5 BB/9:


1961-2006

AGE >= 35
INNINGS PITCHED >= 140
STRIKEOUTS/9 IP <= 3.5
WALKS/9 IP >= 3.5

RSAA YEAR RSAA IP SO/9 IP BB/9 IP
1 Phil Niekro 1986 -4 210.1 3.47 4.06
2 Claude Osteen 1975 -10 204 2.78 4.06
3 Bob Knepper 1989 -30 165 3.49 4.09

Their ERAs the next year were 6.30, out of majors, and 5.68 (in 44 innings). They were all retired before year X+2.

If you lower the innings threshold to 100 you get Mike Cuellar's last year in Baltimore. The next season he pitched three innings, allowed seven runs, and retired. You also get Bob Forsch in '89, and he retired immediately after the year. And you get Wilbur Wood in 1977. In '78 he actually pitched 168 innings, but to a 5.20 ERA in a league with a 3.68.

What Trachsel is doing right now is at the very edge of what's possible. Continuing that next year would be historically unique. If they pick up his option I'll be bitterly disappointed.

Isn't this just a variation of what you said (1) before this season started based on last year's numbers, and (2) back in May when Trachsel had a good start? Yet here he is, with the ERA of a no. 3 starter.

I wouldn't go in to 2008 with a plan that gives me 5 viable starting pitcher opitions, one of whom is Trachsel. But if he's one of about 7-8 options, that's fine. Make Penn and Olson prove they are worthy. Make Loewen prove he's healthy and not too rusty. Make Cabrera prove he can take a step forward instead of running in place.

[tim27]

As ugly as Trachsel's peripherals have been this year, the records of players like him in year X+1 are downright hideous.

Here's everyone in the expansion era over 35, with 140+ IP, < 3.5 K/9, and > 3.5 BB/9:


1961-2006

AGE >= 35
INNINGS PITCHED >= 140
STRIKEOUTS/9 IP <= 3.5
WALKS/9 IP >= 3.5

RSAA YEAR RSAA IP SO/9 IP BB/9 IP
1 Phil Niekro 1986 -4 210.1 3.47 4.06
2 Claude Osteen 1975 -10 204 2.78 4.06
3 Bob Knepper 1989 -30 165 3.49 4.09

Their ERAs the next year were 6.30, out of majors, and 5.68 (in 44 innings). They were all retired before year X+2.

If you lower the innings threshold to 100 you get Mike Cuellar's last year in Baltimore. The next season he pitched three innings, allowed seven runs, and retired. You also get Bob Forsch in '89, and he retired immediately after the year. And you get Wilbur Wood in 1977. In '78 he actually pitched 168 innings, but to a 5.20 ERA in a league with a 3.68.

What Trachsel is doing right now is at the very edge of what's possible. Continuing that next year would be historically unique. If they pick up his option I'll be bitterly disappointed.

Didn't we hear the same things about Traschel last year? Now, maybe he does fall off the cliff, but I was hearing the same things about Traschel last April and it hasn't happened yet. What makes his failure inevitable next year, when supposedly it was inevitable this year and it didn't happen?

[tim27]

Isn't this just a variation of what you said (1) before this season started based on last year's numbers, and (2) back in May when Trachsel had a good start? Yet here he is, with the ERA of a no. 3 starter.

I wouldn't go in to 2008 with a plan that gives me 5 viable starting pitcher opitions, one of whom is Trachsel. But if he's one of about 7-8 options, that's fine. Make Penn and Olson prove they are worthy. Make Loewen prove he's healthy and not too rusty. Make Cabrera prove he can take a step forward instead of running in place.

Agree 100%.

[wildcard]

I think the Orioles are justified if they decide to exercise the $4.75 mm option they have on Trachsel. He has been good a lot more often than he has been bad this year. We don't know what to expect from most of our young guys, and we need a viable alternative of one or more of them isn't ready to step up, or gets hurt. Trachsel gets no respect, and you can make all the K/BB, xFIP, DIPS and age arguments you want, but he has done a good job this year and has been a consistent performer for a very long time.

You could be right, but I think you about a month early. Let's see what he does in September.

[ATO UMD]

If the team is going to base its future around the young pitching we have throughout the organization, there's no point in keeping Trachsel around at all.

Any combination of: Bedard, Cabrera, Olson, Loewen, Guthrie, Liz and/or Penn works for me.

[osfan83]

Just because you blindfolded yourself, ran across 6 lanes of traffic in rush hour and lived to tell about it, doesn't mean it's safe, and you will be able to do it again.

[DrungoHazewood]

Didn't we hear the same things about Traschel last year? Now, maybe he does fall off the cliff, but I was hearing the same things about Traschel last April and it hasn't happened yet. What makes his failure inevitable next year, when supposedly it was inevitable this year and it didn't happen?

So you're fine with betting on a historically unique performance? I think I said at the time of the Trachsel signing that occasionally you'll see someone with ugly K rates have a decent year. It does happen. But those players, almost without exception, drastically decline the next season. Trachsel doesn't even have the benefit of an exceptionally low walk rate, or any other exceptional performance measures. It's all smoke and mirrors, unless you believe he's figured out an entirely new, unique way of pitching successfully.

Just because someone has run 100 yards on the edge of the Grand Canyon without falling over doesn't mean you should bet on him making it through the next 100 yards. I'm not sure there's a single player in all of major league baseball with a higher likelihood of collapse than Steve Trachsel.

[BigDawg819]

Trax, the wily vet, has done exactly what I expected him to do so far this year. He goes out there and battles and gives you your money's worth every time he toes the rubber. Some playoff team or contender or a playoff spot is going to regret grabbing him for the stretch run.

[BigDawg819]

So you're fine with betting on a historically unique performance? I think I said at the time of the Trachsel signing that occasionally you'll see someone with ugly K rates have a decent year. It does happen. But those players, almost without exception, drastically decline the next season. Trachsel doesn't even have the benefit of an exceptionally low walk rate, or any other exceptional performance measures. It's all smoke and mirrors, unless you believe he's figured out an entirely new, unique way of pitching successfully.

Just because someone has run 100 yards on the edge of the Grand Canyon without falling over doesn't mean you should bet on him making it through the next 100 yards. I'm not sure there's a single player in all of major league baseball with a higher likelihood of collapse than Steve Trachsel.

There is, his name is Danny Baez......

[tim27]

So you're fine with betting on a historically unique performance? I think I said at the time of the Trachsel signing that occasionally you'll see someone with ugly K rates have a decent year. It does happen. But those players, almost without exception, drastically decline the next season. Trachsel doesn't even have the benefit of an exceptionally low walk rate, or any other exceptional performance measures. It's all smoke and mirrors, unless you believe he's figured out an entirely new, unique way of pitching successfully.

Just because someone has run 100 yards on the edge of the Grand Canyon without falling over doesn't mean you should bet on him making it through the next 100 yards. I'm not sure there's a single player in all of major league baseball with a higher likelihood of collapse than Steve Trachsel.

Didn't you say that last season too? I mean, if you keep predicting his decline and you wait long enough I'm sure it'll come to pass. I mean, he is like 38 years old.
FWIW, I'm not saying it won't happen; I expect it to as well. Hell, he could get bombed all September and this makes this pretty much a moot point. But if he continues to pitch well for the rest of the season then I have no problem bringing him back next year as insurance for the young pitchers we have in the rotation. Like Frobby said, I wouldn't go into a year I was going to compete with Traschel as one of my five starting options (although the Mets did last year and they did all right) but in a rebuilding year, solely as insurance, there is nothing wrong with him being here.

[RShack]

Didn't we hear the same things about Traschel last year? Now, maybe he does fall off the cliff, but I was hearing the same things about Traschel last April and it hasn't happened yet. What makes his failure inevitable next year, when supposedly it was inevitable this year and it didn't happen?
This is the *classic* OH misuse of stats... using normative data to predict individual performance... it's nice to be aware of normative data, you'd be a dope not to consider what is normal... but reaching predictive conclusions about individuals based only on normative data is simply a very wrong thing to do... yet it happens around here every time... what I am saying is not some radical idea, it's one of the basics ... it's from "Stats 101", people.... (well, maybe Stats 201)...

[DrungoHazewood]

Didn't you say that last season too? I mean, if you keep predicting his decline and you wait long enough I'm sure it'll come to pass. I mean, he is like 38 years old.
FWIW, I'm not saying it won't happen; I expect it to as well. Hell, he could get bombed all September and this makes this pretty much a moot point. But if he continues to pitch well for the rest of the season then I have no problem bringing him back next year as insurance for the young pitchers we have in the rotation. Like Frobby said, I wouldn't go into a year I was going to compete with Traschel as one of my five starting options (although the Mets did last year and they did all right) but in a rebuilding year, solely as insurance, there is nothing wrong with him being here.

Do you really think that the Orioles would bring back Steve Trachsel as insurance? If they pick up his $4.75M option for 2008 he's going to be in the rotation come hell or high water. That money virtually guarantees him 15 starts.

I wouldn't guarantee him anything. The only way I'd even consider bringing him back is if they decine his option and resign him at the league minimum, or on a minor league deal. And we know that's not happening.

[tim27]

This is the *classic* OH misuse of stats... using normative data to predict individual performance... it's nice to be aware of normative data, you'd be a dope not to consider what is normal... but reaching predictive conclusions about individuals based only on normative data is simply a very wrong thing to do... yet it happens around here every time... what I am saying is not some radical idea, it's one of the basics ... it's from "Stats 101", people....

So I guess you think Wang is a fluke because he doesn't k enough batters, right?;)

[tim27]

Do you really think that the Orioles would bring back Steve Trachsel as insurance? If they pick up his $4.75M option for 2008 he's going to be in the rotation come hell or high water. That money virtually guarantees him 15 starts.

I wouldn't guarantee him anything. The only way I'd even consider bringing him back is if they decine his option and resign him at the league minimum, or on a minor league deal. And we know that's not happening.

Hey, I already admitted that is a concern of mine. If- IF- they use Traschel as an emergency option for the young starters, and keep him in the pen for the rest of the time to be long-relief for the inevitable 2.1 inning start that Olson/Liz is going to give you every other week, then I have no problem with them picking up his option. But if I had to choose between 20 starts of Traschel and no Traschel for next year and I had to choose right now, I'd choose no Traschel.

[Sports Guy]

Didn't you say that last season too? I mean, if you keep predicting his decline and you wait long enough I'm sure it'll come to pass. I mean, he is like 38 years old.
FWIW, I'm not saying it won't happen; I expect it to as well. Hell, he could get bombed all September and this makes this pretty much a moot point. But if he continues to pitch well for the rest of the season then I have no problem bringing him back next year as insurance for the young pitchers we have in the rotation. Like Frobby said, I wouldn't go into a year I was going to compete with Traschel as one of my five starting options (although the Mets did last year and they did all right) but in a rebuilding year, solely as insurance, there is nothing wrong with him being here.
So, in a year where you want to compete, you don't keep the vet pitcher....In a year where you want to rebuild, you keep the vet pitcher and use him over the young pitcher?

That makes no sense.

[tim27]

So, in a year where you want to compete, you don't keep the vet pitcher....In a year where you want to rebuild, you keep the vet pitcher and use him over the young pitcher?

That makes no sense.


What part of SOLELY AS INSURANCE don't you understand?

[Sports Guy]

This is the *classic* OH misuse of stats... using normative data to predict individual performance... it's nice to be aware of normative data, you'd be a dope not to consider what is normal... but reaching predictive conclusions about individuals based only on normative data is simply a very wrong thing to do... yet it happens around here every time... what I am saying is not some radical idea, it's one of the basics ... it's from "Stats 101", people.... (well, maybe Stats 201)...
Yea, you are right...Let's ignore stats, history and trends.

Let's just say that every player can buck any trend, so wtf, let's see what happens. :rolleyes:

We are playing the odds.

[DrungoHazewood]

This is the *classic* OH misuse of stats... using normative data to predict individual performance... it's nice to be aware of normative data, you'd be a dope not to consider what is normal... but reaching predictive conclusions about individuals based only on normative data is simply a very wrong thing to do... yet it happens around here every time... what I am saying is not some radical idea, it's one of the basics ... it's from "Stats 101", people....

So you're of the opinion that Trachsel is likely to continue his currently adequate ERA?

Don't you ever consider that a player who isn't just slightly bucking trends, but is orders of magnitude and many standard deviations from the mean, is all that more unlikley to work out? This is the same as the Tejada/Hernandez argument. When you've deviated a bit from normal I can accept that you may have unique attributes that mean you'll buck overall trends. But when you're off the scale, in a bad way, there's almost no chance of that.

Luis Hernandez is so bad that it's almost inconceivable that he has some kind of unique set of skills that'll let him eventually succeed. Steve Trachsel's secondary numbers are so terrible that it almost guarantees failure in the future. Both of these players have profiles that don't have future success even a trivial percentage of the time.

[Old#5fan]

This is the *classic* OH misuse of stats... using normative data to predict individual performance... it's nice to be aware of normative data, you'd be a dope not to consider what is normal... but reaching predictive conclusions about individuals based only on normative data is simply a very wrong thing to do... yet it happens around here every time... what I am saying is not some radical idea, it's one of the basics ... it's from "Stats 101", people.... (well, maybe Stats 201)...

Yeah if baseball was like a computer game then analysis of stats should be able to predict the winners each season. Of course stats don't account for the human factor, or unexplained anamolies.

I would love some statistician for example to explain why Corey Patterson for example reversed his trend from last season where he could hit LH pitching but hit well against RH pitching and this year it is just the reverse.

My theory on this is he worked hard during the offseason to improve and maybe took a lot of extra batting practice and worked out against a LH pitcher. However, by doing so he diminished his previous ability to hit a RH pitcher. I don't know if this is a valid theory but it is the only one I can think of. Otherwise it just doesn't make a lick of sense!

[DrungoHazewood]

What part of SOLELY AS INSURANCE don't you understand?

So what part of spending $4.75M on a pitcher precludes him from being insurance don't you understand? What part of $4.75M is about $4M too much for an insurance policy don't you understand?

[Sports Guy]

What part of SOLELY AS INSURANCE don't you understand?

I understand that you are completely wrong if you think he is insurance..If they pick up his option, he is in the rotation regardless of his or others performance until at least May, if not longer.

To think otherwise is ridiculous.

And in a rebuilding year, you know who is insurance for a young pitcher? ANOTHER YOUNG PITCHER!

As of now, the Orioles have these guys who should be in the majors next year:

Bedard
Loewen
DCab
Guthrie
Penn
Olson
Burres
Liz

We also have guys like JJ Johnson who could be in the mix.

This is before you make any trades, sign any MiL FAs or sign the next RLo or Guthrie(if you do).

We don't need a 4.75 million dollar insurance policy, especially one who is likely to be the worse starter in the AL next year.

BTW, the premise of this thread is Trax has been unlucky because of the pen...Well he has been very lucky to even have a chance to win most of these games so it evens out.

[Old#5fan]

So you're of the opinion that Trachsel is likely to continue his currently adequate ERA?

Don't you ever consider that a player who isn't just slightly bucking trends, but is orders of magnitude and many standard deviations from the mean, is all that more unlikley to work out? This is the same as the Tejada/Hernandez argument. When you've deviated a bit from normal I can accept that you may have unique attributes that mean you'll buck overall trends. But when you're off the scale, in a bad way, there's almost no chance of that.

Luis Hernandez is so bad that it's almost inconceivable that he has some kind of unique set of skills that'll let him eventually succeed. Steve Trachsel's secondary numbers are so terrible that it almost guarantees failure in the future. Both of these players have profiles that don't have future success even a trivial percentage of the time.


Yeah, and Corey Patterson was so bad last year at hitting LH pitchers that many Orioles fans wanted him gone or at least out of the lineup against lefties. Now this year the opposite is true! Splain that one Ricky!:)

IMHO that is just another example of how you simply cannot predict the future based on past stats even in the major leagues from year to year. To predict or project stats from the minors to majors is nothing more than a crapshoot period.

[RShack]

This is the *classic* OH misuse of stats... using normative data to predict individual performance... it's nice to be aware of normative data, you'd be a dope not to consider what is normal... but reaching predictive conclusions about individuals based only on normative data is simply a very wrong thing to do... yet it happens around here every time... what I am saying is not some radical idea, it's one of the basics ... it's from "Stats 101", people.... (well, maybe Stats 201)...
Yea, you are right...Let's ignore stats, history and trends.

SG, for you, I recommend "Reading Comprehension 101" before "Stats 101"....

[DrungoHazewood]

Yeah if baseball was like a computer game then analyzation of stats should be able to predict the winners each season. Of course stats don't account for the human factor, or unexplained anamolies.

I would love some statistician for example to explain why Corey Patterson for example reversed his trend from last season where he could hit LH pitching but hit well against RH pitching and this year it is just the reverse.

My theory on this is he worked hard during the offseason to improve and maybe took a lot of extra batting practice and worked out against a LH pitcher. However, by doing so he diminished his previous ability to hit a RH pitcher. I don't know if this is a valid theory but it is the only one I can think of. Otherwise it just doesn't make a lick of sense!

Patterson has reverse splits this year because most players have a relatively similar platoon split. Given 150 or 200 PAs an individual's performance can vary much more than that platoon split. Patterson wasn't as bad against lefties as he appeared last year, nor as good as he looks this year.

Going forward you'd do much better to assume he's going to hit his career averages of .734 vs. righties and .657 vs. lefties than you would to assume he's suddenly decided to work on his hitting and figured out how to hit lefties at the age of 27.

[DrungoHazewood]

Yeah, and Corey Patterson was so bad last year at hitting LH pitchers that many Orioles fans wanted him gone or at least out of the lineup against lefties. Now this year the opposite is true! Splain that one Ricky!:)

IMHO that is just another example of how you simply cannot predict the future based on past stats even in the major leagues from year to year. To predict or project stats from the minors to majors is nothing more than a crapshoot period.

See post #31 in this thread.

[Sports Guy]

Yeah if baseball was like a computer game then analyzation of stats should be able to predict the winners each season. Of course stats don't account for the human factor, or unexplained anamolies.

I would love some statistician for example to explain why Corey Patterson for example reversed his trend from last season where he could hit LH pitching but hit well against RH pitching and this year it is just the reverse.

My theory on this is he worked hard during the offseason to improve and maybe took a lot of extra batting practice and worked out against a LH pitcher. However, by doing so he diminished his previous ability to hit a RH pitcher. I don't know if this is a valid theory but it is the only one I can think of. Otherwise it just doesn't make a lick of sense!

Things like this can vary year to year....In 2003 and 2004, CPat was very solid versus lefties...In 2002, 2005 and 2006, he was very poor versus them.

[Old#5fan]

Patterson has reverse splits this year because most players have a relatively similar platoon split. Given 150 or 200 PAs an individual's performance can vary much more than that platoon split. Patterson wasn't as bad against lefties as he appeared last year, nor as good as he looks this year.

Going forward you'd do much better to assume he's going to hit his career averages of .734 vs. righties and .657 vs. lefties than you would to assume he's suddenly decided to work on his hitting and figured out how to hit lefties at the age of 27.

Okay, if you say so. Under your stats based projections what will he do next season? In other words will he revert back to hitting poorly against LH pitching and better against RH pitching? If what you say is true than he is due for a year where he hits about the same against both or so it would seem?:confused:

[Sports Guy]

Yeah, and Corey Patterson was so bad last year at hitting LH pitchers that many Orioles fans wanted him gone or at least out of the lineup against lefties. Now this year the opposite is true! Splain that one Ricky!:)

IMHO that is just another example of how you simply cannot predict the future based on past stats even in the major leagues from year to year. To predict or project stats from the minors to majors is nothing more than a crapshoot period.There is that word again Eli.

[Sports Guy]

SG, for you, I recommend "Reading Comprehension 101" before "Stats 101"....
And for you, i recommend "i don't understand what i am talking about 101".

End of the day, you should always use the best future stats to figure things out...Yes they are humans and yes things happen but to ever build a team like that or sign a player like that is beyond stupid.

[Old#5fan]

There is that word again Eli.

I am not trying to cause anyone a problem but I don't think the use of that word would be considered illegal would it?:confused:

[DrungoHazewood]

Okay, if you say so. Under your stats based projections what will he do next season? In other words will he revert back to hitting poorly against LH pitching and better against RH pitching? If what you say is true than he is due for a year where he hits about the same against both or so it would seem?:confused:

You're misunderstanding the predictive value of individual numbers. No one says they're a crystal ball. They just give us a likely range of what's going to happen.

If someone put a gun to my head and ask what Patterson's platoon splits would be next year I'd give them his career averages. But I'd know full well that they could vary significantly from that because that's the nature of a 200 at bat trial.

[Sports Guy]

I am not trying to cause anyone a problem but I don't think the use of that word would be considered illegal would it?:confused:
No, its not illegal.

[Old#5fan]

And for you, i recommend "i don't understand what i am talking about 101".

End of the day, you should always use the best future stats to figure things out...Yes they are humans and yes things happen but to ever build a team liek that or sign a player like that is beyond stupid.

Actually, I think it would make much more sense to use scouting reports along with stats as a backup. But what do I know?:(

[DrungoHazewood]

I am not trying to cause anyone a problem but I don't think the use of that word would be considered illegal would it?:confused:

Not illegal, no. But in the context you're using it it's incorrect.

[RShack]

So you're of the opinion that Trachsel is likely to continue his currently adequate ERA?
No, I am of the opinion that past performance is not only the best predictor of future performance, it is also a pretty lousy predictor. Just because it's the best predictor we have, that does not make it trustworthy.

Furthermore, the best predictor of an individual's performance is the past performance of that very individual, not normative data about everybody. And even that is a poor predictor.

Don't turn this into something it's not. Don't turn it into some goofy "let's play Luis everyday" claim. Just be reasonable. You are looking for stats-based certainty when the actual truth is a great deal more uncertain than the conclusions you reach based on normative data about past performance.

According to the discussions around here, Jeff Kent could never happen, Steve Finley could never happen, Roger Clemens could never happen, Melvin Mora could never happen, Mike Cuellar could never happen. LOTS of things could never happen.

(Now, how long until somebody says, "Oh, you think Trax is like Clemens huh?")

[Sports Guy]

Actually, I think it would make much more sense to use scouting reports along with stats as a backup. But what do I know?:(Scouting reports should be used for certain things but even scouts get things wrong.

For example...We have seen ONE scout say they think Luis Hernandez can be a ML SS. Well, we have yet to see more than that and while he may be ready defensively, anyone who thinks he is ready with the bat has no idea what they are talking about.

So, sometimes they have to be taken with a grain of salt.

Scouting reports should be used on players you haven't seen but even those aren't accurate a lot.

I remember wanting Kaz Matsui...I had never seen him play but all the scouting reports i saw raved about him....Well, i ignored his stats that would have told me he wasn't going to be that good of a MLer.

[hoosiers]

I love threads like these - especially when they contain ridiculous phrases like "historically unique" to explain being confounded by the great Master Traschel.

Among obvious mistakes, this has been a quality pickup by our FO. I hope we deal him soon, but I see no reason to accept any average AA arm for him.

[DrungoHazewood]

Actually, I think it would make much more sense to use scouting reports along with stats as a backup. But what do I know?:(

Why not use all of the information at your disposal, weighting it by what makes sense for each individual case? There are players for whom a statistical analysis makes sense, players with less of a track record for whom scouting reports should hold more weight, and players for whom other information could be invaluable.

[Sports Guy]

No, I am of the opinion that past performance is not only the best predictor of future performance, it is also a pretty lousy predictor. Just because it's the best predictor we have, that does not make it trustworthy.

Furthermore, the best predictor of an individual's performance is the past performance of that very individual, not normative data about everybody. And even that is a poor predictor.

Don't turn this into something it's not. Don't turn it into some goofy "let's play Luis everyday" claim. Just be reasonable. You are looking for stats-based certainty when the actual truth is a great deal more uncertain than the conclusions you reach based on normative data about past performance.

According to the discussions around here, Jeff Kent could never happen, Steve Finley could never happen, Roger Clemens could never happen, Melvin Mora could never happen, Mike Cuellar could never happen. LOTS of things could never happen.

(Now, how long until somebody says, "Oh, you think Trax is like Clemens huh?")
This isn't really true.

Let's say a player is 35 years old....They were very good up until age 32(10 year period)...In the last 3 seasons, they have been on the decline....What do you think happens...They revert back to the form of their 20s and their prime or they keep declining as they get into their mid to upper 30s?

[Mackus]

What part of SOLELY AS INSURANCE don't you understand?The part that costs us $5M.

[DrungoHazewood]

I love threads like these - especially when they contain ridiculous phrases like "historically unique" to explain being confounded by the great Master Traschel.

Among obvious mistakes, this has been a quality pickup by our FO. I hope we deal him soon, but I see no reason to accept any average AA arm for him.

I love comments like this - especially when they contain ridiculous phrases like Master Trachsel. You want to keep relying on players with (whatever you want to call a 45:69 K:BB ratio in 140 innings), have at it.

[BigDawg819]

The part that costs us $5M.

Well insurance rates are on the rise..... :D

[tim27]

[QUOTE=Sports Guy;1034339]I understand that you are completely wrong if you think he is insurance..If they pick up his option, he is in the rotation regardless of his or others performance until at least May, if not longer.

To think otherwise is ridiculous.

I'd be glad to pay him 4.75 next year solely as insurance because I wouldn't ever base my playing time arrangements on financial concern. To argue that the O's would is a valid point I've adressed. I'm not arguing what the O's would do but what I would do.


And in a rebuilding year, you know who is insurance for a young pitcher? ANOTHER YOUNG PITCHER!

As of now, the Orioles have these guys who should be in the majors next year:

Bedard Obviously
Loewen Injury concern, not to mention his ability to pitch 175 inn. due to ineffeciency.
DCabObviously
GuthrieI thought you weren't ready to hand him a rotation spot next year because he's a fluke.
PennInjury concern and poor major league performance.
OlsonShould probably be here and start but I wouldn't hand him a rotation spot.
BurresIMO, he shouldn't really be looked at as a serious option to start. Talk about smoke and mirrors.
LizMaybe he should be here and maybe not. He could very well benefit from another year in the high minors.

We also have guys like JJ Johnson who could be in the mix.JJ Johnson is not ML starter material. We might as well hand LF to Fio.

By my count that's two guys we can legitemately count on to give us 200 ip next year. We need to get those innings somewhere. It could be a lot worse than Steve Traschel. Now you'll make the argument that it can't be any worse but that 4.40 era begs to differ.

[BigDawg819]

I love comments like this - especially when they contain ridiculous phrases like Master Trachsel. You want to keep relying on players with (whatever you want to call a 45:69 K:BB ratio in 140 innings), have at it.

Are you trying to say that Trax is not a Jedi? Blasphemy!

<a href="http://photobucket.com" target="_blank"><img src="http://i44.photobucket.com/albums/f12/wrobel8/storm-trooper.jpg" border="0" alt="Photo Sharing and Video Hosting at Photobucket"></a>

[tim27]

You're misunderstanding the predictive value of individual numbers. No one says they're a crystal ball. They just give us a likely range of what's going to happen.

If someone put a gun to my head and ask what Patterson's platoon splits would be next year I'd give them his career averages. But I'd know full well that they could vary significantly from that because that's the nature of a 200 at bat trial.

So basically you couldn't tell them anything then could you?

[DrungoHazewood]

According to the discussions around here, Jeff Kent could never happen, Steve Finley could never happen, Roger Clemens could never happen, Melvin Mora could never happen, Mike Cuellar could never happen. LOTS of things could never happen.

This is the fatal flaw in your argument. None of those players had a season where his peripherals were so fundamentally, completely in conflict with his results.

This is all about playing the odds, and I'm not betting the 2008 Orioles season on Trachsel bucking massive odds against him.

[Sports Guy]

[QUOTE]

I'd be glad to pay him 4.75 next year solely as insurance because I wouldn't ever base my playing time arrangements on financial concern. To argue that the O's would is a valid point I've adressed. I'm not arguing what the O's would do but what I would do.



By my count that's two guys we can legitemately count on to give us 200 ip next year. We need to get those innings somewhere. It could be a lot worse than Steve Traschel. Now you'll make the argument that it can't be any worse but that 4.40 era begs to differ.How many staffs have more than 2 guys who throw 200 innings a year.

I pretty much disagree with your entire post but it is a waste of my time to get into it....Just answer that question.

[vatech1994]

Trachsel has pitched well this year, but I wouldn't give him a dime for next year. My eyes tell me the same thing the stats tell me. He is on the verge of being out of baseball. He survives right now on throwing enough curveballs and splitters just off the edge of the plate. A really patient team will tear him up 3 out of 4 starts. Even last night he was hit very hard early but many of them went directly at outfielders. The Devil Rays got upset about that and became impatient which played directly into Trachsel's hands. We did the same thing with the D'Rays starter on Tuesday night.

I'm not stats only. I'm not eyes only. However, when the stats and my eyes both agree with each other, I go with them regardless of current results. Trachsel is on the hairy edge right now. He may find his way through another season on that edge, but I wouldn't want to be a nickel on it.

BTW, this is also the reason why I think SG's stance on Guthrie is nuts and why I think Baez can pitch slightly above replacement level. My eyes tell me that Guthrie has very good stuff and awesome peripheral characteristics. The stats may not make him out to be a future #2/3 starter, but I'm going with my eyes and his results this year. The stats say Baez isn't very good, but my eyes see that he can be absolutely nasty at times. He just needs to be more consistent in his execution.

[DrungoHazewood]

So basically you couldn't tell them anything then could you?

About exactly what Corey Patterson was going to do in 2008? Nope. If I could I'd be a very rich man. But I could give them a range of outcomes with varying liklihoods that would almost certainly encompass what he'll actually do.

[Old#5fan]

Why not use all of the information at your disposal, weighting it by what makes sense for each individual case? There are players for whom a statistical analysis makes sense, players with less of a track record for whom scouting reports should hold more weight, and players for whom other information could be invaluable.

Actually I think a combination of all of the above should be used on every player. Even if there are stats available you still need to get a reliable scouting report.

[Old#5fan]

About exactly what Corey Patterson was going to do in 2008? Nope. If I could I'd be a very rich man. But I could give them a range of outcomes with varying liklihoods that would almost certainly encompass what he'll actually do.

So what I take from this is stats really just aren't very reliable to predict what will occur in the future as there are just too many variables. I distinctly recall a few years ago on this forum many posters were all hyped up about Kurt Ainsworth because he had a great ERA in pre-season and looked like a future Cy Young. I commented at the time that I thought this guy looked like there was something wrong with him (based on observation) and he was damaged goods when the Orioles got him.

I was pretty much derided for that view because of his great "stats." Well, we all know who turned out to be right on that now don't we? I don't care about being right or wrong, but I do think way too much emphasis is placed on stats which overall are not reliable enough to accurately predict much of anything other than general possibilities. In other words they are a better history of a players performance than a future indicator, hence not very reliable regarding future performance?

Would you agree with my take expressed thusly?

[DrungoHazewood]

I'd be glad to pay him 4.75 next year solely as insurance because I wouldn't ever base my playing time arrangements on financial concern.

You have a luxury that the Baltimore Orioles most certainly don't.

[tim27]

[QUOTE=tim27;1034377]How many staffs have more than 2 guys who throw 200 innings a year.

I pretty much disagree with your entire post but it is a waste of my time to get into it....Just answer that question.

I'd say most good staffs have more than two guys throw 200 innings. The Red Sox this year will have 3, and that's with nominal ace Schilling hurt. The LAA will have three. The Tigers will have three, or be close. I think you need AT LEAST 850-900 innings from your starters, ideally another 100.

I love it how it's a waste of time for you to explain yourself. If it's such a waste of time then why do you have 45,000 posts?

[Sports Guy]

So what I take from this is stats really just aren't very reliable to predict what will occur in the future as there are just too many variables. I distinctly recall a few years ago on this forum many posters were all hyped up about Kurt Ainsworth because he had a great ERA in pre-season and looked like a future Cy Young. I commented at the time that I thought this guy looked like there was something wrong with him (based on observation) and he was damaged goods when the Orioles got him.

I was pretty much derided for that view because of his great "stats." Well, we all know who turned out to be right on that now don't we? I don't care about being right or wrong, but I do think way too much emphasis is placed on stats which overall are not reliable enough to accurately predict much of anything other than general possibilities. In other words they are a better history of a players performance than a future indicator, hence not very reliable regarding future performance?
You were right because he got hurt...Had nothing to do with his talent.

[Sports Guy]

[QUOTE=Sports Guy;1034385]

I'd say most good staffs have more than two guys throw 200 innings. The Red Sox this year will have 3, and that's with nominal ace Schilling hurt. The LAA will have three. The Tigers will have three, or be close. I think you need AT LEAST 850-900 innings from your starters, ideally another 100.

I love it how it's a waste of time for you to explain yourself. If it's such a waste of time then why do you have 45,000 posts?

No, it is a waste of time for me to explain myself to YOU...Big difference.

And you'd say most staffs?

Well, if Guthrie ends up being as good as many think he will, we should have 3 next year who can throw 200 innings...So, you are wrong there.

And you gave me the example of the best staffs in the game....How does that equal MOST?

[tim27]

You have a luxury that the Baltimore Orioles most certainly don't.

Why don't the O's have that luxury? They'll be paying Gibbons more than that to sit at home. They'll be paying Baez more than that to get scorched in the 8th. They'll be paying Payton that to have 400 abs with a .700 ops. They'll have a starting rotation making under 10 mill next year. They could most certainly carry Trasch next year and not bust the budget- it's not as if they're going to sign any big-time FA.

[Old#5fan]

You were right because he got hurt...Had nothing to do with his talent.

Technically true, but posters here were deriding me for making that observation based on how great his stats were in meaningless spring training games (another topic that I am sure has been argued here previously). I was basing it on watching him pitch in the regular season the previous year. He just did not look like a good major league pitcher to me and no doubt it was due to his being damaged goods.

So my point is made that you cannot soley rely on stats. The same argument was made with Tejada as posters were saying his field (stats-wise) was not a problem but many of us saw that it was by observation. Now the stats are reflecting it moreso but the problem was visible previously as well.

The point I am making is over-reliance on stats is a faulty stance to take. Watching a player's on-field performance will tell an experienced baseball fan of many years much more about a player. I agree if you are young and have not played the game or followed it very long you probably would get caught in the trap of placing your views of a player disproportionally based on stats. I just wish to caution against this. That is all, I am through with my "preaching" on this for now.;)

[tim27]

[QUOTE=Sports Guy;1034402][QUOTE=tim27;1034396]

No, it is a waste of time for me to explain myself to YOU...Big difference.

And I'm the one that trolls your posts? O-KAY. If it's a waste of time to explain yourself to someone who won't neccessarily agree with what you say then I guess it is a waste. I'll give you the advice you give everyone who has a problem with you: Put me on ignore.


And you'd say most staffs?

Most good staffs. And, despite our pathetic bullpen, we have a good starting rotation and its only getting better.



Well, if Guthrie ends up being as good as many think he will, we should have 3 next year who can throw 200 innings...So, you are wrong there.

But you don't believe that do you? You don't think we can count on him to finish next year in the rotation let alone with 200 innings. So that's a pretty hollow argument coming from you.

[tim27]

[QUOTE=tim27;1034396]



And you gave me the example of the best staffs in the game....How does that equal MOST?

Didn't I say "most good staffs"? I know reading comprehension is tough for you SG. You just can't wait to make you point no matter what the other guy is saying. It must be nice living in a world where all dicussions equate hollering at a brick wall.

[RShack]

Furthermore, the best predictor of an individual's performance is the past performance of that very individual, not normative data about everybody. And even that is a poor predictor.
This isn't really true.
Actually, it is... whether it's baseball or anything else about people...


Sports Guy]
Let's say a player is 35 years old....They were very good up until age 32(10 year period)...In the last 3 seasons, they have been on the decline....What do you think happens...They revert back to the form of their 20s and their prime or they keep declining as they get into their mid to upper 30s?
In your example, what about referring to the 3-year trend in the past performance of that individual is NOT looking at that individual's past performance?

Reading Comprehension 101? Or Logic 101? Hmmm, I'm not sure...

[Sports Guy]

But you don't believe that do you? You don't think we can count on him to finish next year in the rotation let alone with 200 innings. So that's a pretty hollow argument coming from you.

Doesn't matter if i believe it or not...You do and you are the one sayign we need Trax for insurance despite the fact that i would think you ebleive we have 3 200IP pitchers(or at least 2 and one who will go 180+)

So really, what do we need Trax for?

The idea of needing him for insurance is bogus.

[NewMarketSean]

Technically true, but posters here were deriding me for making that observation based on how great his stats were in meaningless spring training games (another topic that I am sure has been argued here previously). I was basing it on watching him pitch in the regular season the previous year. He just did not look like a good major league pitcher to me and no doubt it was due to his being damaged goods.

So my point is made that you cannot soley rely on stats. The same argument was made with Tejada as posters were saying his field (stats-wise) was not a problem but many of us saw that it was by observation. Now the stats are reflecting it moreso but the problem was visible previously as well.

The point I am making is over-reliance on stats is a faulty stance to take. Watching a player's on-field performance will tell an experienced baseball fan of many years much more about a player. I agree if you are young and have not played the game or followed it very long you probably would get caught in the trap of placing your views of a player disproportionally based on stats. I just wish to caution against this. That is all, I am through with my "preaching" on this for now.;)

Old#19Fan, when have you ever been wrong in your 40 years of astute TV watching and ignoring stats? When have your eyes ever deceived you, and a player you liked never panned out? That of course is if they ever have deceived you, oh wise and mighty TV watcher.

[Sports Guy]

Actually, it is... whether it's baseball or anything else about people...


In your example, what about referring to the 3-year trend in the past performance of that individual is NOT looking at that individual's past performance?

Reading Comprehension 101? Or Logic 101? Hmmm, I'm not sure...

But a 10 year performance should be looked at before a 3 year performance, right?

Shouldn't we ignore those trends of the last 3 years??

[osfan83]

Why don't the O's have that luxury? They'll be paying Gibbons more than that to sit at home. They'll be paying Baez more than that to get scorched in the 8th. They'll be paying Payton that to have 400 abs with a .700 ops. They'll have a starting rotation making under 10 mill next year. They could most certainly carry Trasch next year and not bust the budget- it's not as if they're going to sign any big-time FA.

This is why they don't have that luxury. They already will have too much money tied up in players that will probably contribute very little to great success. Replace the contracts of Gibbons, Payton and Baez with league minimum contracts, and you might then have the luxury to spend 5Mil on an over 35, mediocre at best, insurance policy.

[RShack]

But a 10 year performance should be looked at before a 3 year performance, right?

Shouldn't we ignore those trends of the last 3 years??
SG, what are you doing? You can do better than this...

[Mackus]

I'd be glad to pay him 4.75 next year solely as insurance because I wouldn't ever base my playing time arrangements on financial concern.If we had to have him back, I agree completely that we shouldn't base his playing time on his salary.

But, since it is our option to pay him that, it'd be silly to pick it up. We can find a guy just as likely to help us out for far less. The Mets got Jorge Sosa this year for about a million or so. Guys like that are usually available if you want an insurance policy. Then take that extra $4M and spend it on something actually useful, like perhaps another reliever or a decent DH.

[Witchy Chick]

[QUOTE=tim27;1034396]

No, it is a waste of time for me to explain myself to YOU...Big difference.

And you'd say most staffs?

Well, if Guthrie ends up being as good as many think he will, we should have 3 next year who can throw 200 innings...So, you are wrong there.

And you gave me the example of the best staffs in the game....How does that equal MOST?

This guy (http://www.baseball-reference.com/k/kazmisc01.shtml) will hit 200 IP. And so will this guy (http://www.baseball-reference.com/s/shielja02.shtml). And this one (http://www.baseball-reference.com/j/jacksed01.shtml) will probably be within 30-40 IP (depending on # starts, IP/game, injury-avoidance). And that's not a good staff.

And who the friggin' hell are the three 200 IP pitchers with LAA? Top three (at the moment) are Lackey (181 1/3), Escobar (172), and Weaver the younger (only 133). Weav' ain't pitching 67 innings in the final month of the season. :confused:


Witchy

[Sports Guy]

SG, what are you doing? You can do better than this...

Just trying to figure out what little logic you have.

Basically, you are saying that stats mean something if you want them to mean something...That is how i am taking it.

That while things will likely normalize, you don't really know.

You also said past performance is the biggest key...I gave you an example and yet you sit there and don't really answer me.

So, which is it? When does past performance matter? How far back do you go?

And if you do it over a few years(like in my example), aren't you then looking at trends, both historical and current? And isn't this going against what you are saying?

[Mackus]

Why don't the O's have that luxury? They'll be paying Gibbons more than that to sit at home. They'll be paying Baez more than that to get scorched in the 8th. They'll be paying Payton that to have 400 abs with a .700 ops. They'll have a starting rotation making under 10 mill next year. They could most certainly carry Trasch next year and not bust the budget- it's not as if they're going to sign any big-time FA.

I've never thought that past mistakes were good justification for making more mistakes, but thats just me.

[Sports Guy]

I've never thought that past mistakes were good justification for making more mistakes, but thats just me.

Well, if you are an Orioles fan, they are.

[Flosman]

But a 10 year performance should be looked at before a 3 year performance, right?

Shouldn't we ignore those trends of the last 3 years??

SG you are just being hard headed. Shack is not making the point you are attributing to him. Part of looking at an individuals historic stats would be the trends reflected in the stats. So expected age/performance trends that were actually reflected in the data would have to be considered. No where did Shack say that you look at the stats over the players entire career, solely, to make your assessment. Shacks point is correct in that the actual players stats are the best indicator of future performance. Jon is correct that more information can be obtained by looking at a players numbers compared to similiar players at the same point in thier careers with similiar numbers and trends. Shacks main point is that none of this is a certainty.

[DrungoHazewood]

So what I take from this is stats really just aren't very reliable to predict what will occur in the future as there are just too many variables. I distinctly recall a few years ago on this forum many posters were all hyped up about Kurt Ainsworth because he had a great ERA in pre-season and looked like a future Cy Young. I commented at the time that I thought this guy looked like there was something wrong with him (based on observation) and he was damaged goods when the Orioles got him.

I was pretty much derided for that view because of his great "stats." Well, we all know who turned out to be right on that now don't we? I don't care about being right or wrong, but I do think way too much emphasis is placed on stats which overall are not reliable enough to accurately predict much of anything other than general possibilities. In other words they are a better history of a players performance than a future indicator, hence not very reliable regarding future performance?

Would you agree with my take expressed thusly?

I disagree with your assessment that stats should take a backseat to scouting when analyzing players. Scouting reports are crucial, but they're also very subjective. A player with a long track record is likely to repeat that performance, and that's measured with stats. You need to verify those stats with scouting regarding health and level of competition and weaknesses.

But the numbers are vital. GMs who ignore them do it at their own peril.

[RShack]

Just trying to figure out what little logic you have.

Basically, you are saying that stats mean something if you want them to mean something...That is how i am taking it.

That while things will likely normalize, you don't really know.

You also said past performance is the biggest key...I gave you an example and yet you sit there and don't really answer me.

So, which is it? When does past performance matter? How far back do you go?

And if you do it over a few years(like in my example), aren't you then looking at trends, both historical and current? And isn't this going against what you are saying?
I realize that you don't like uncertainty, but it's a very real and important part of life, including baseball. Sadly, the most useful perspective does not reduce to a simple soundbite, because it it multi-dimensional. It is not reducible to one bullet point, there are several:
Stats are a valuable tool. It would be dumb to not consider what they suggest.
Previous performance is the best predictor of future performance, but it is nonetheless an unreliable predictor.
For a given individual, the previous performance of that individual is a better indicator than normative data for a group, but even that is still an unreliable predictor (i.e., low r-value)
The sensible thing to do is to consider multiple factors, including past performance and detailed observations of current performance by knowledgeable observers/evaluators.
Applying the proper weight to the various factors is something that is still more art than science, and is likely to continue that way for the foreseeable future. The error that is routinely made around here is to pretend otherwise.

[Frobby]

You have a luxury that the Baltimore Orioles most certainly don't.

So the Rangers can pay $10 mm a year for Vicente Padilla and the Orioles can't afford $4.75 mm for Trachsel?

I'm far from blind to Trachsel's faults, his age, and the statistics you cite that indicate he could fall off the end of the earth. But I also have a hard time ignoring the fact that this guy is (probably) about to complete yet another season with a sub-5.00 ERA -- for the 13th time in 15 major league seasons and the 8th season in a row. He's going to pitch enough innings to qualify for the 11th time in the last 12 seasons. If he was 32 and had good peripherals, he'd be making Padilla/Meche/Lilly money for that. The $4.75 mm is a salary that takes into account all of the risks you mentioned.

I'm not yet advocating that the Orioles should exercise the option. First, there is a month to play, let's see how he's looked after 162 games have been played. If he's really been as lucky as you think, maybe he'll get bombed in his 6 remaining starts.

Second, it depends on the overall plan for next year. If we are in rebuilding mode, and aren't even trying to compete, then go ahead and let Olson and Penn take their lumps with no safety net other than a bunch of other unproven pitchers. That's not a bad strategy in the right circumstances.

[tywright]

If we're talking about bullpens blowing leads, Milwaukee takes the cake especially since they are in a division race.

http://www.baseballprospectus.com/unfiltered/?p=506


The Brewers have blown a lot of leads this year — 13 times they have been ahead by at least three runs, and gone on to lose the ballgame. Even more remarkably, nine of those losses have come since July 28, which is the sort of thing that gets managers and pitching coaches fired.



If we're talking about why teams are trading for Trachsel...be patient

[DrungoHazewood]

So the Rangers can pay $10 mm a year for Vicente Padilla and the Orioles can't afford $4.75 mm for Trachsel?

That's worked out well for them, hasn't it? In fact, you could make the argument that giving $10M to Vincente Padilla is one of the major reasons the Rangers are 60-72, 18.5 games out of first place.

One of my pet peeves is paying $millions for players who're not going to push you towards more wins. You don't pay $5M a year for placeholders. You don't pay $5M for people with 50% or 75% collapse rates. You can get placeholders for $500k, and spend the $4.5M that's left over on something that'll help you win more ballgames.

[BigDawg819]

Wait now we want to be on the same scale of the Rangers in terms of overpaying for mediocre talent? Hope is no longer fading, its gone............

[tim27]

That's worked out well for them, hasn't it? In fact, you could make the argument that giving $10M to Vincente Padilla is one of the major reasons the Rangers are 60-72, 18.5 games out of first place.

One of my pet peeves is paying $millions for players who're not going to push you towards more wins. You don't pay $5M a year for placeholders. You don't pay $5M for people with 50% or 75% collapse rates. You can get placeholders for $500k, and spend the $4.5M that's left over on something that'll help you win more ballgames.

You could make that argument but you'd be very, very wrong. With or without Padilla the Rangers are 60-72; hell, without him they're probably worse as hard as that is to believe. They stink this year because Millwood has been terrible, because McCarthy's been hurt a lot and stunk when he did pitch, because Robinon Tejada has stunk since April, because Mike Young has had a subpar season, because Blalock's been hurt, because Kinsler hasn't developed quite as they hope, and, imo, the biggest reason, because they went into the season with a 39 yo Kenny Lofton as their best outfielder. It had nothing to do with Padilla's struggles. He could be on pace for 200 inn and a 3.50 era and they'd still be right where they are. You do realize that Traschel would be FAR and AWAY the best starter in their rotation this year? FAR and AWAY.
So, sure, we could not exercise Traschel's option and save a few million. Big F'In deal. We still aren't going to sign A-Rod or Hunter or Jones (Not that you would advocate that anyway) and we run the risk of JJ Johnson throwing 150 innings next year at a 6.5 era.

[Frobby]

That's worked out well for them, hasn't it? In fact, you could make the argument that giving $10M to Vincente Padilla is one of the major reasons the Rangers are 60-72, 18.5 games out of first place.

One of my pet peeves is paying $millions for players who're not going to push you towards more wins. You don't pay $5M a year for placeholders. You don't pay $5M for people with 50% or 75% collapse rates. You can get placeholders for $500k, and spend the $4.5M that's left over on something that'll help you win more ballgames.

Like who, Jay Payton? :cool:

I don't know what a "placeholder" is. Every player has an expectation, an upside, a downside, and risk. I think you and I probably agree on Trachsel's upside (a repeat of this year) and his downside (Chen 2006). We probably differ significantly on expectation (I'd say, 5.00-5.25 for 2008) and risk (you say 50-75% collapse rate, I say 25-50%).

Where I think we differ more is the risk associated with the $500k "placeholders" as you put it. Every year several of these guys come through and have seasons like the ones Trachsel typically has had maybe better. But every year a lot of them have seasons like Chen. Russ Ortiz or worse.

[tim27]

I've never thought that past mistakes were good justification for making more mistakes, but thats just me.

Is it a mistake to pay a guy 4.5 million for 160 innings of 4.5 era? I don't think so- not in today's market.

Now you can tell me he's going to fall off but I bet you said the same thing six months ago, right? You can't be certain that he will and if he is capable of reproducing his year this year next year then it will be a mistake to let him go over 4.5 mill.

[DrungoHazewood]

Now you can tell me he's going to fall off but I bet you said the same thing six months ago, right? You can't be certain that he will and if he is capable of reproducing his year this year next year then it will be a mistake to let him go over 4.5 mill.

Certain? No. Pretty darn sure? Yep.

[tim27]

Certain? No. Pretty darn sure? Yep.

Were you or were you not pretty damn sure he'd be terrible for the O's in April 07?

[DrungoHazewood]

Like who, Jay Payton? :cool:

I don't know what a "placeholder" is. Every player has an expectation, an upside, a downside, and risk. I think you and I probably agree on Trachsel's upside (a repeat of this year) and his downside (Chen 2006). We probably differ significantly on expectation (I'd say, 5.00-5.25 for 2008) and risk (you say 50-75% collapse rate, I say 25-50%).

Where I think we differ more is the risk associated with the $500k "placeholders" as you put it. Every year several of these guys come through and have seasons like the ones Trachsel typically has had maybe better. But every year a lot of them have seasons like Chen. Russ Ortiz or worse.

I'll take a supposedly risky $350k placeholder over a Payton or his ilk all day long. If Payton fails... well, we've seen what happens. He just keeps getting plugged into the lineup all year and he'll probably get 300 PAs next year, and no matter what we're out $10M.

If you instead went with Jon Knott, and he fails you release him or send him to the minors and plug in Jason Dubois. If Dubois fails you dump him and plug in Tike Redman. If Tike Redman fails... you get the point. You've saved $millions that can be spent elsewhere, and you've gotten the same performance in the field.

[DrungoHazewood]

Were you or were you not pretty damn sure he'd be terrible for the O's in April 07?

I thought it was very unlikely he'd pitch as well as he has.

[DrungoHazewood]

You could make that argument but you'd be very, very wrong. With or without Padilla the Rangers are 60-72; hell, without him they're probably worse as hard as that is to believe. They stink this year because Millwood has been terrible, because McCarthy's been hurt a lot and stunk when he did pitch, because Robinon Tejada has stunk since April, because Mike Young has had a subpar season, because Blalock's been hurt, because Kinsler hasn't developed quite as they hope, and, imo, the biggest reason, because they went into the season with a 39 yo Kenny Lofton as their best outfielder. It had nothing to do with Padilla's struggles. He could be on pace for 200 inn and a 3.50 era and they'd still be right where they are. You do realize that Traschel would be FAR and AWAY the best starter in their rotation this year? FAR and AWAY.
So, sure, we could not exercise Traschel's option and save a few million. Big F'In deal. We still aren't going to sign A-Rod or Hunter or Jones (Not that you would advocate that anyway) and we run the risk of JJ Johnson throwing 150 innings next year at a 6.5 era.

You're actually making the argument that giving 18 starts and $10M to a pitcher with a 6.24 ERA isn't hurting the Texas Rangers?

[Flosman]

I'll take a supposedly risky $350k placeholder over a Payton or his ilk all day long. If Payton fails... well, we've seen what happens. He just keeps getting plugged into the lineup all year and he'll probably get 300 PAs next year, and no matter what we're out $10M.

If you instead went with Jon Knott, and he fails you release him or send him to the minors and plug in Jason Dubois. If Dubois fails you dump him and plug in Tike Redman. If Tike Redman fails... you get the point. You've saved $millions that can be spent elsewhere, and you've gotten the same performance in the field.

I agree with your theory, too bad the O's don't. The O's would be a much better team without guys like Payton, Mora, Gobbons type guys committed to multiyear deals. Save the money and sign difference makers. If you are going to settle for a hopefully league average guy make it a one year deal. It is what I would do.

[Fan4Life]

I thought it was very unlikely he'd pitch as well as he has.

Which should make you somewhat less than pretty darn certain he'll callapse next year. :p

[Fan4Life]

I agree with your theory, too bad the O's don't. The O's would be a much better team without guys like Payton, Mora, Gobbons type guys committed to multiyear deals. Save the money and sign difference makers. If you are going to settle for a hopefully league average guy make it a one year deal. It is what I would do.

What difference makers have we passed on that were worth the money they eventually got? Not since Delgado and Vlad(assuming there was a number he would come for)has there been a signing we passed on that I felt was a mistake. Can you think of any, with regard to being difference makers?

[tim27]

You're actually making the argument that giving 18 starts and $10M to a pitcher with a 6.24 ERA isn't hurting the Texas Rangers?

No, I'm making the argument that giving 18 starts to a 10mil pitcher with a 6.24 era is one of the "Major" reasons the Rangers are terrible. With or without Padilla they suck.

[tim27]

I thought it was very unlikely he'd pitch as well as he has.

Are you more certain that he'll stink in 08 as opposed to 07?

[DrungoHazewood]

What difference makers have we passed on that were worth the money they eventually got? Not since Delgado and Vlad(assuming there was a number he would come for)has there been a signing we passed on that I felt was a mistake. Can you think of any, with regard to being difference makers?

If you don't want to spend the savings on free agents, pour it all into scouting and development. I'd rather do that than sign most free agents. In any case, there are always better uses of money than $10M for random 30-something players who might be average if they overachieve.

[Since'54]

I realize that you don't like uncertainty, but it's a very real and important part of life, including baseball. Sadly, the most useful perspective does not reduce to a simple soundbite, because it it multi-dimensional. It is not reducible to one bullet point, there are several:
Stats are a valuable tool. It would be dumb to not consider what they suggest.
Previous performance is the best predictor of future performance, but it is nonetheless an unreliable predictor.
For a given individual, the previous performance of that individual is a better indicator than normative data for a group, but even that is still an unreliable predictor (i.e., low r-value)
The sensible thing to do is to consider multiple factors, including past performance and detailed observations of current performance by knowledgeable observers/evaluators.
Applying the proper weight to the various factors is something that is still more art than science, and is likely to continue that way for the foreseeable future. The error that is routinely made around here is to pretend otherwise.

One of the best and funniest parts of Moneyball is the description of pre-draft meetings Billy Beane was having with his scouts. Beane wanted to draft a catcher who had good stats but his "knowledgeable observers/evaluators" couldn't get past the fact that the prospect had a big rear end and didn't look like a ballplayer. Hence Beane's remark, "We're not selling blue jeans here".

To me the stats, imperfect predictors that they are, are going to be better to go on than observations from baseball lifers. Too many built-in prejudices
and long-held opinions that, when analyzed, are pretty much baseless.

[DrungoHazewood]

Are you more certain that he'll stink in 08 as opposed to 07?

Yes. As I've said before, pitchers sometimes have a decent season with awful periperials. They almost never repeat it. That goes double for 37-year-olds.

[DrungoHazewood]

No, I'm making the argument that giving 18 starts to a 10mil pitcher with a 6.24 era is one of the "Major" reasons the Rangers are terrible. With or without Padilla they suck.

And they would suck slightly less if they'd made a better use of the resources used to sign Padilla.

[tim27]

I'll take a supposedly risky $350k placeholder over a Payton or his ilk all day long. If Payton fails... well, we've seen what happens. He just keeps getting plugged into the lineup all year and he'll probably get 300 PAs next year, and no matter what we're out $10M.

If you instead went with Jon Knott, and he fails you release him or send him to the minors and plug in Jason Dubois. If Dubois fails you dump him and plug in Tike Redman. If Tike Redman fails... you get the point. You've saved $millions that can be spent elsewhere, and you've gotten the same performance in the field.

This is the fallacy of your argument and I'll try to make you see it one time. With Payton we know what we're going to get, granted it costs more money than it's worth. So we bring up Knott and he sucks over a month. Then they'll be some knucklehead saying, "Well, you only gave him a 100 abs how do you know he really sucks?" Ditto for Dubois and Redman. You'll save some $ undoubtedly- not that it's getting spent anywhere else mind you- but you'll most likely get an inferior performance and you still won't know what you've got, short of a breakout by a career minor leaguer.

[tim27]

And they would suck slightly less if they'd made a better use of the resources used to sign Padilla.

Yeah, they could have signed Trascel for half that.

[Sports Guy]

Is it smart to bank 4.75 million dollars that a player will literally be one of the only players in ML history to do something?

[Sports Guy]

This is the fallacy of your argument and I'll try to make you see it one time. With Payton we know what we're going to get, granted it costs more money than it's worth. So we bring up Knott and he sucks over a month. Then they'll be some knucklehead saying, "Well, you only gave him a 100 abs how do you know he really sucks?" Ditto for Dubois and Redman. You'll save some $ undoubtedly- not that it's getting spent anywhere else mind you- but you'll most likely get an inferior performance and you still won't know what you've got, short of a breakout by a career minor leaguer.Yea, a bad player.

[tim27]

Yes. As I've said before, pitchers sometimes have a decent season with awful periperials. They almost never repeat it. That goes double for 37-year-olds.

It's pretty bold to sit here and harp on Traschel for a year with the certainty that he'll suck, be proven wrong, and then say, "Wait till next year because he'll definitely suck then." Like I said, keep predicting his downfall and it'll eventually come but you and all your periperials aren't telling anybody something they don't know already.

[tim27]

Is it smart to bank 4.75 million dollars that a player will literally be one of the only players in ML history to do something?

Smarter than banking 14million on a player in a career free fall.

[Sports Guy]

Smarter than banking 14million on a player in a career free fall.

What the hell does this matter?

You shouldn't really bank on either.

[tim27]

What the hell does this matter?

You shouldn't really bank on either.

You're the one that wants Sexson. Yeah, Payton sucks at 4.5 mil, but he's better than Sexson and he makes three times as much.

[Sports Guy]

You're the one that wants Sexson. Yeah, Payton sucks at 4.5 mil, but he's better than Sexson and he makes three times as much.
You are so clueless it is unreal.

What have i said about a Sexson trade?

That i would want Balentien back with him...Do you understand that that is the key or is that concept above your head?

BTW, in an awful year for Sexson, he has a higher OPS and has shown a lot more power than Payton.

Also, Sexson is a one year gamble to see if he can bounce back....He was solid just last year so it is a gamble that could pay off big time for us and if it fails, then we cut him and have Balentien to fall back on, which is the whole point of trading for Sexson.

[tim27]

You are so clueless it is unreal.

What have i said about a Sexson trade?

That i would want Balentien back with him...Do you understand that that is the key or is that concept above your head?

You also want to give up Millar, a superior player at this point to Sexson, and D-Cab. I understand the concept it's just a terrible idea, that's all.


BTW, in an awful year for Sexson, he has a higher OPS and has shown a lot more power than Payton.

He's a terrible defensive 1st baseman; Payton is a good defensive lf than can run. Give me Payton any day of the week if their salaries are equal.


Also, Sexson is a one year gamble to see if he can bounce back....He was solid just last year so it is a gamble that could pay off big time for us and if it fails, then we cut him and have Balentien to fall back on, which is the whole point of trading for Sexson.

So a one year gamble on Trasch- who's actually performed well this year- is a bad idea, but a one year gamble on Sexson who's getting worse every year and makes three times as much is a good idea?
Makes a lot of sense.:rolleyes:

[Old#5fan]

I realize that you don't like uncertainty, but it's a very real and important part of life, including baseball. Sadly, the most useful perspective does not reduce to a simple soundbite, because it it multi-dimensional. It is not reducible to one bullet point, there are several:
Stats are a valuable tool. It would be dumb to not consider what they suggest.
Previous performance is the best predictor of future performance, but it is nonetheless an unreliable predictor.
For a given individual, the previous performance of that individual is a better indicator than normative data for a group, but even that is still an unreliable predictor (i.e., low r-value)
The sensible thing to do is to consider multiple factors, including past performance and detailed observations of current performance by knowledgeable observers/evaluators.
Applying the proper weight to the various factors is something that is still more art than science, and is likely to continue that way for the foreseeable future. The error that is routinely made around here is to pretend otherwise.

This is a money post! Especially your last bullet and final two sentences. Truer words have never been written here (IMHO)!;)

[DrungoHazewood]

This is the fallacy of your argument and I'll try to make you see it one time. With Payton we know what we're going to get, granted it costs more money than it's worth. So we bring up Knott and he sucks over a month. Then they'll be some knucklehead saying, "Well, you only gave him a 100 abs how do you know he really sucks?" Ditto for Dubois and Redman. You'll save some $ undoubtedly- not that it's getting spent anywhere else mind you- but you'll most likely get an inferior performance and you still won't know what you've got, short of a breakout by a career minor leaguer.

I really, really, really, really doubt that Knott, Dubois, or even Tike Redman would have been meaningfully worse than Jay Payton's .259/.299/.368 line. No left fielder in the majors is more than a win worse than Jay Payton.

[Sports Guy]

So a one year gamble on Trasch- who's actually performed well this year- is a bad idea, but a one year gamble on Sexson who's getting worse every year and makes three times as much is a good idea?
Makes a lot of sense.:rolleyes:

Big difference is and what i would bank on is that Sexson would be coming from a pitchers park to a hitters park and he was still pretty good last year.

Trax, while he was decent with the Mets last year, was coming to the AL, with poor peripherals, a worse team and to a better hitters park.

Plus, offense is down across the league...This has helped Trax and hurt Sexson.

[Old#5fan]

Yes. As I've said before, pitchers sometimes have a decent season with awful periperials. They almost never repeat it. That goes double for 37-year-olds.


This may be off topic if so I apologize (didn't want to start a thread on this) but do you think Mussina is done? His recent poor outings and his age would seem to indicate so.

[DrungoHazewood]

It's pretty bold to sit here and harp on Traschel for a year with the certainty that he'll suck, be proven wrong, and then say, "Wait till next year because he'll definitely suck then." Like I said, keep predicting his downfall and it'll eventually come but you and all your periperials aren't telling anybody something they don't know already.

You're going much further out on a limb predicting that a guy who's clearly walking a tightrope should be around next year, and be paid $5M.

If you roll a pair of dice and hit snakeyes twice in a row, am I being bold by predicting that the third roll isn't going to follow the first two?

[ScottieBaseball]

So what I take from this is stats really just aren't very reliable to predict what will occur in the future as there are just too many variables. I distinctly recall a few years ago on this forum many posters were all hyped up about Kurt Ainsworth because he had a great ERA in pre-season and looked like a future Cy Young. I commented at the time that I thought this guy looked like there was something wrong with him (based on observation) and he was damaged goods when the Orioles got him.

I was pretty much derided for that view because of his great "stats." Well, we all know who turned out to be right on that now don't we? I don't care about being right or wrong, but I do think way too much emphasis is placed on stats which overall are not reliable enough to accurately predict much of anything other than general possibilities. In other words they are a better history of a players performance than a future indicator, hence not very reliable regarding future performance?

Would you agree with my take expressed thusly?

Stats, while flawed, are an EXACT measurement of PAST performance, but also a pretty good indicator of future performance. Yes, there are variables, but the evolution of sabremetrics has done a great deal in accounting for them and factoring them into an equation. To say, "Not very reliable" would be incorrect. Statistics simply need to be taken in context.

Trust me...I'm far from a stathead. I believe that just because you can't put it in a spreadsheet and sort it doesn't mean it's not tangible. A few months ago someone (don't remember who) argued that there's no statistical evidence that protection in the lineup exists. Having caught until I was 16 years old, I beg to differ. There were plenty of times that I asked the umpire for time out, tucked my skull cap and mask under my arm, and walked out to the mound to ask the pitcher, "How do you think we should pitch to this guy?" Followed closely by, "You see who's in the on-deck circle, don't you?" Stats might not support it, but I can assure everyone that protection in the lineup exists.

[tywright]

You are so clueless it is unreal.

What have i said about a Sexson trade?

That i would want Balentien back with him...Do you understand that that is the key or is that concept above your head?

BTW, in an awful year for Sexson, he has a higher OPS and has shown a lot more power than Payton.

Also, Sexson is a one year gamble to see if he can bounce back....He was solid just last year so it is a gamble that could pay off big time for us and if it fails, then we cut him and have Balentien to fall back on, which is the whole point of trading for Sexson.


I can't believe you just called him clueless. Actually it doesn't surprise me

[Frobby]

This may be off topic if so I apologize (didn't want to start a thread on this) but do you think Mussina is done? His recent poor outings and his age would seem to indicate so.

Yeah, could be. I've remained a big Moose fan all these years, but it looks to me like he's toast. But, he's a resourceful and intelligent piece of toast, so maybe he has one more adjustment to make to compensate for his ever-worsening fastball.

[Sports Guy]

I can't believe you just called him clueless. Actually it doesn't surprise me
In this situation, with the way he is implying the whole Sexson thing and what i wanted to accomplish, yes he is clueless.

[Old#5fan]

So a one year gamble on Trasch- who's actually performed well this year- is a bad idea, but a one year gamble on Sexson who's getting worse every year and makes three times as much is a good idea?
Makes a lot of sense.:rolleyes:

No, it doesn't. I have never liked Sexon. He has a huge strike zone due to his size and very overated on many message boards. I personally am very glad the Orioles never wasted good money on him.

[Sports Guy]

Yeah, could be. I've remained a big Moose fan all these years, but it looks to me like he's toast. But, he's a resourceful and intelligent piece of toast, so maybe he has one more adjustment to make to compensate for his ever-worsening fastball.yea, Moose could be done but he also may be hurt...He wasn't pitching that bad before his recent terrible stretch.

He certainly deserves another year to see if he is done or not IMO.

[DrungoHazewood]

This may be off topic if so I apologize (didn't want to start a thread on this) but do you think Mussina is done? His recent poor outings and his age would seem to indicate so.

Hard to say, but the trends aren't optimistic. His ERA by half since the beginning of 2006 is 3.24, 3.96, 4.62, 7.09. His strikeout rate has declined a lot between this year and last. It wouldn't surprise me if he's done, but he might just be having a bad year or some nagging injuries.

[Roy Firestone]

As ugly as Trachsel's peripherals have been this year, the records of players like him in year X+1 are downright hideous.

Here's everyone in the expansion era over 35, with 140+ IP, < 3.5 K/9, and > 3.5 BB/9:


1961-2006

AGE >= 35
INNINGS PITCHED >= 140
STRIKEOUTS/9 IP <= 3.5
WALKS/9 IP >= 3.5

RSAA YEAR RSAA IP SO/9 IP BB/9 IP
1 Phil Niekro 1986 -4 210.1 3.47 4.06
2 Claude Osteen 1975 -10 204 2.78 4.06
3 Bob Knepper 1989 -30 165 3.49 4.09

Their ERAs the next year were 6.30, out of majors, and 5.68 (in 44 innings). They were all retired before year X+2.

If you lower the innings threshold to 100 you get Mike Cuellar's last year in Baltimore. The next season he pitched three innings, allowed seven runs, and retired. You also get Bob Forsch in '89, and he retired immediately after the year. And you get Wilbur Wood in 1977. In '78 he actually pitched 168 innings, but to a 5.20 ERA in a league with a 3.68.

What Trachsel is doing right now is at the very edge of what's possible. Continuing that next year would be historically unique. If they pick up his option I'll be bitterly disappointed.


Hey Jon...I have an idea. Can you check and see if Niekro is available?
Hes probably only 63 or 64 and I bet can pitch one inning every other night!

[BigDawg819]

Hey Jon...I have an idea. Can you check and see if Niekro is available?
Hes probably only 63 or 64 and I bet can pitch one inning every other night!

He can't be worse then Baez..... ;)

[Frobby]

I'm annoyed that this thread has gotten so far off-topic. There are plenty of other threads about how Trachsel will do next year and whether we should keep him.

Back to the original topic -- the bullpen has pulled the rug out from under him SEVEN TIMES! That is rather incredible. I don't remember any Oriole pitcher ever being victimized that frequently. DCab, meanwhile, only has been vicitimized one time.

Part of this is that Trax rarely goes beyond 6 IP. That multiplies the chances that the bullpen can screw it up. He's also left with runners on base quite a few times. Still -- SEVEN TIMES?

[Baltimoron]

I realize that you don't like uncertainty, but it's a very real and important part of life, including baseball. Sadly, the most useful perspective does not reduce to a simple soundbite, because it it multi-dimensional. It is not reducible to one bullet point, there are several:
Stats are a valuable tool. It would be dumb to not consider what they suggest.
Previous performance is the best predictor of future performance, but it is nonetheless an unreliable predictor.
For a given individual, the previous performance of that individual is a better indicator than normative data for a group, but even that is still an unreliable predictor (i.e., low r-value)
The sensible thing to do is to consider multiple factors, including past performance and detailed observations of current performance by knowledgeable observers/evaluators.

Care to quantify these ideas?

What do you mean by unreliable? What r value are you thinking off?



Applying the proper weight to the various factors is something that is still more art than science, and is likely to continue that way for the foreseeable future. The error that is routinely made around here is to pretend otherwise.

This is simply not true for hitters and to a lesser degree for pitchers as well.

[Old#5fan]

Hard to say, but the trends aren't optimistic. His ERA by half since the beginning of 2006 is 3.24, 3.96, 4.62, 7.09. His strikeout rate has declined a lot between this year and last. It wouldn't surprise me if he's done, but he might just be having a bad year or some nagging injuries.

I heard his top fastball now only clocks at 87 mph. That isn't going to get major league hitters out very often I don't care how good his control may be.
I heard some analyst on sports center claim that the only comparable long time top level pitcher pitcher with three terrible outings as Mussina has had at the same age was Jack Morris, and after falling off so badly he was done.

[Sports Guy]

No, it doesn't. I have never liked Sexon. He has a huge strike zone due to his size and very overated on many message boards. I personally am very glad the Orioles never wasted good money on him.

Overrated? How do you mean overrated?

He started getting good playing time in 1998....Since then, he has had 7 years of 29 homers or more...He has had an OBP of 340 or more 6 times and just missed out on that in 2 other seasons. He has driven in 100 or more runs 6 times and anoher time he was over 90. He has also hit 28 or more doubles in 5 different years and has had a slugging % over 500 nine times.

He was hurt and missed most of 2004 or his stats would be even better than this IMO.

He has had an OPS of 850 or more 8 times and 4 times it was over 900. Just missed 850 last year.

I am not sure why you say he is overrated or has been overrated unless people are saying he is a HOFer.

He has a career 860 OPS, which is taking a hit this year.

[Tx Oriole]

IMO i would not mind having ST around another season. Can we be sure how any of the young pitchers will work out? Can't be sure of Loewen or Penn or others. SG would it really hurt to ahve ST as the number 5?

[DrungoHazewood]

Hey Jon...I have an idea. Can you check and see if Niekro is available?
Hes probably only 63 or 64 and I bet can pitch one inning every other night!

He was in Cooperstown, you should have grabbed him!

[Old#5fan]

Overrated? How do you mean overrated?

He started getting good playing time in 1998....Since then, he has had 7 years of 29 homers or more...He has had an OBP of 340 or more 6 times and just missed out on that in 2 other seasons. He has driven in 100 or more runs 6 times and anoher time he was over 90. He has also hit 28 or more doubles in 5 different years and has had a slugging % over 500 nine times.

He was hurt and missed most of 2004 or his stats would be even better than this IMO.

He has had an OPS of 850 or more 8 times and 4 times it was over 900. Just missed 850 last year.

I am not sure why you say he is overrated or has been overrated unless people are saying he is a HOFer.

He has a career 860 OPS, which is taking a hit this year.

I don't think anyone I have seen has ever said he is HOF material but I have seen a slew of people on various message boards (not necessarily this one) drooling over this guy and I have never been impressed with him at all. He has too many wholes in his swing and an enormous strike zone. Furthermore, a guy his size should be hitting upper thirty's into the 40's as as far as homers go, not in the twenty's. Heck Tony Batista hit that many homers and had 90 rbi's every year as an Oriole. Not that impressive to me.

[Sports Guy]

I don't think anyone I have seen has ever said he is HOF material but I have seen a slew of people on various message boards (not necessarily this one) drooling over this guy and I have never been impressed with him at all. He has too many wholes in his swing and an enormous strike zone. Furthermore, a guy his size should be hitting upper thirty's into the 40's as as far as homers go, not in the twenty's. Heck Tony Batista hit that many homers and had 90 rbi's every year as an Oriole. Not that impressive to me.

TBat didn't have the OPS or OBP to go along with it like Sexson.

Sexson has hit 30 homers or more 6 times...In other season, he is either not played a full season or was hurt(in 2004). So really, he is pretty much doing what you would hope every full season he plays.

[hoosiers]

You're misunderstanding the predictive value of individual numbers. No one says they're a crystal ball. They just give us a likely range of what's going to happen.

Can you please explain the "predictive value" of individual numbers and how they may provide a "likely range of what's going to happen"?

And show me how that relates to predicting Steve Traschel's numbers in 2007? Or Eric Bedard's? Or Danys Baez? Or Jim Hoey? Or Cory Doyne? Or Aubrey Huff? Or Kevin Millar?

Thank you.

[DrungoHazewood]

Can you please explain the "predictive value" of individual numbers and how they may provide a "likely range of what's going to happen"?

And show me how that relates to predicting Steve Traschel's numbers in 2007? Or Eric Bedard's? Or Danys Baez? Or Jim Hoey? Or Cory Doyne? Or Aubrey Huff? Or Kevin Millar?

Thank you.

No. I'm tired of playing these games and splitting hairs. If you don't think past performance gives you a clue into future results, that's your prerogative.

[Sports Guy]

Can you please explain the "predictive value" of individual numbers and how they may provide a "likely range of what's going to happen"?

And show me how that relates to predicting Steve Traschel's numbers in 2007? Or Eric Bedard's? Or Danys Baez? Or Jim Hoey? Or Cory Doyne? Or Aubrey Huff? Or Kevin Millar?

Thank you.

Off of the top of my head, Bedard's DIPS and FIP era's from last year told you he should be better this year than last....And he has.

[RShack]

To me the stats, imperfect predictors that they are, are going to be better to go on than observations from baseball lifers. Too many built-in prejudices
and long-held opinions that, when analyzed, are pretty much baseless.
Who's talking about baseball lifers with bad judgment? Many jobs are important. Many jobs are done crappily by some people. That doesn't mean you eliminate the job. It means you use people who have a clue. You need clueful people who know how to use stats properly, and you need clueful people who are knowledgeable observers/evaluators of how a guy performs. The entire "either-or" argument is vacuous. Discounting either role is just dumb.

[Fan4Life]

For me, stats are only as good as the consistency level they represent. I'm not sure there is a way to evaluate that other than looking at the stats in a more micro level. As a hitter, going 12-40 can be broken down into a 1-20 slump followed by an 11-20 streak... or a 4-14, 4-13, 4-13, etc etc.... their stats will look the same in the aggregate but they are much different hitters if those types of numbers are typical of each player. Frobby had a post about Hernandez earlier remarking on how terrible he was last year in July and August I think but that in September last year, he had an OPS over 1.1. That hot September last year, and perhaps again this year, would likely bring his numbers up to a reasonable level so that when looking at his overall 2007 season, his stats won't reflect how terrible he was when we needed him. The question I would have is: Is this type of hitter (2-15 and then 10-25) more common than the more consistent 4-12, 4-13 guy?

[Old#5fan]

No. I'm tired of playing these games and splitting hairs. If you don't think past performance gives you a clue into future results, that's your prerogative.

I would say "don't give a clue" is similar to a "crapshoot" would you not agree? As such, I am on that side of this argument. In fact, I bet if you took a poll of who believes that past statistics (especially minor league) will be of signicifant use to predict future major league performance that most people (over 50%) don't believe it. (Not sure about this board, but in real life).

[Sports Guy]

I would say "don't give a clue" is similar to a "crapshoot" would you not agree? As such, I am on that side of this argument. In fact, I bet if you took a poll of who believes that past statistics (especially minor league) will be of signicifant use to predict future major league performance that most people (over 50%) don't believe it. (Not sure about this board, but in real life).

Well, there would be a lot of wrong people out there Eli.

[DrungoHazewood]

I would say "don't give a clue" is similar to a "crapshoot" would you not agree? As such, I am on that side of this argument. In fact, I bet if you took a poll of who believes that past statistics (especially minor league) will be of signicifant use to predict future major league performance that most people (over 50%) don't believe it. (Not sure about this board, but in real life).

No.

My belief is that past performance is strongly related to future performance.

Your belief is that it's not, especially when you're talking about minor league performance.

If your supposition is true most people would be wrong.

[RShack]

No. I'm tired of playing these games and splitting hairs. If you don't think past performance gives you a clue into future results, that's your prerogative.
Come on, Drungo, that's the kind of all-or-nothing logic that Eli used: near-certain reliability or complete-and-utter crap shoot, nothing in between. Of course stats that model *some* aspects of past performance give you a clue. They can give you many clues. Sometimes they can provide clues that you wouldn't/couldn't see without them. But it's also true that there's a big diff between something that's a clue and something that's a truly reliable predictor. It's also true that stats about past baseball performance do not reflect everything that's relevant, they only reflect the subset of attributes for which data is readily available and easily acquired. This is not splitting hairs, these things are just facts.

[ScottieBaseball]

Stats are an effective tool. Stats are an effective tool. Stats are an effective tool. Stats are an effective tool. Stats are an effective tool. Stats are an effective tool...

Stats are an effective tool and a primary one at that when coupled with personal observation and the variables that can't be accounted for/captured via sabremetrics (i.e. adjusted stats). Neither stats nor "the good set o' peepers in my head" are the be-all, end-all in predicting the future. If that's what you're looking for and you're not A) God or B) holding the keys to a time machine, then you're not going to have any luck.

[DrungoHazewood]

Come on, Drungo, that's the kind of all-or-nothing logic that Eli used: near-certain reliability or complete-and-utter crap shoot, nothing in between. Of course stats that model *some* aspects of past performance give you a clue. They can give you many clues. Sometimes they can provide clues that you wouldn't/couldn't see without them. But it's also true that there's a big diff between something that's a clue and something that's a truly reliable predictor. It's also true that stats about past baseball performance do not reflect everything that's relevant, they only reflect the subset of attributes for which data is readily available and easily acquired. This is not splitting hairs, these things are just facts.

It's splitting hairs when I've said numbers aren't the be-all end-all many times over but my words gets twisted around to the point where people are challenging me to predict Corey Patterson's OPS against lefties down to three decimal points, and if I decline that's held up as proof that numbers are almost meaningless.

I've said my piece. I'm done with this thread.

[Old#5fan]

No.

My belief is that past performance is strongly related to future performance.

Your belief is that it's not, especially when you're talking about minor league performance.

If your supposition is true most people would be wrong.

Most people would be wrong where? Here or the man on the street? I am 52 years old and have been an avid Oriole, Colt, and now Ravens fan since I attended my first game at Memorial stadium in 1963. I can assure you that most common fans I know (and most of my friends, and family are sports fans) would not put all the stock into stats that is prevalent on this board.

[Sports Guy]

Most people would be wrong where? Here or the man on the street? I am 52 years old and have been an avid Oriole, Colt, and now Ravens fan since I attended my first game at Memorial stadium in 1963. I can assure you that most common fans I know (and most of my friends, and family are sports fans) would not put all the stock into stats that is prevalent on this board.Well, you and your friends would be wrong.

[Old#5fan]

Come on, Drungo, that's the kind of all-or-nothing logic that Eli used: near-certain reliability or complete-and-utter crap shoot, nothing in between. Of course stats that model *some* aspects of past performance give you a clue. They can give you many clues. Sometimes they can provide clues that you wouldn't/couldn't see without them. But it's also true that there's a big diff between something that's a clue and something that's a truly reliable predictor. It's also true that stats about past baseball performance do not reflect everything that's relevant, they only reflect the subset of attributes for which data is readily available and easily acquired. This is not splitting hairs, these things are just facts.

Yeah, but apparently these stats guys wish to use the stats after the fact but whenever you try to pin them down to predict something they are pretty much useless or as someone might say unreliable in that context.

[Sports Guy]

Yeah, but apparently these stats guys wish to use the stats after the fact but whenever you try to pin them down to predict something they are pretty much useless or as someone might say unreliable in that context.You just clearly don't understand the concept though.

You bring up CPat...I showed you that 3 years he has been good versus lefties and 3 years he has not been good against lefties.

Small sample sizes can easily change from year to year. This shouldn't be that hard to figure out.

[Old#5fan]

Well, you and your friends would be wrong.

No, they would not be. Every single time someone is asked here to solely use stats to predict anything all they do is hedge and give possibilities. So in reality just how useful are these stats? They are not much better than guessing when it comes to "predicting anything." As a measurement of what has already happened they have great value but other than that they do not stand on their own as much value at all. A observation by a scout would be of more benefit in most cases.

I will give you an example. Suppose a guy is hitting 300 in the high minors but in reality he is hitting 400 on fast balls and 200 on curves. So if you look at his stats he gets promoted to the majors. Once he gets to the majors scouts have already told other teams the guy has trouble with the curve. Now he gets a steady diet of even better quality (i.e. major league curves) and hits below the mendoza line and gets sent back down. Now how reliable were the stats in this case? I will tell you, they are not even presenting an accurate picture period, or the kid wouldn't even be called up until he learns how to hit a curve ball much better. So relying on a scout is far better than any stats.

[Old#5fan]

You just clearly don't understand the concept though.

You bring up CPat...I showed you that 3 years he has been good versus lefties and 3 years he has been against righties.

Small sample sizes can easily change from year to year. This shouldn't be that hard to figure out.

So where is the value of that in predicting what he will do next year?:confused: :rolleyes:

[RShack]

Care to quantify these ideas?

What do you mean by unreliable? What r value are you thinking off?
No, I don't care to sift thru 11 textbooks on stats and human performance and create a summary. There's just too much stuff to deal with here. Part of the problem is that it's a complicated business, and people who have access to db's want to believe there's a simple cookbook. There isn't. However, I can give you an example that might shed some light on the main difficulty. This isn't about baseball, but I believe the lessons are relevant to baseball.

For a bunch of years I taught big CS classes at Georgia Tech. In the first 2 of them, we grew to the point where we had more than 5,000 seats per year. The grading scheme was weekly quizzes and a big final exam, and it was weighted to let people who did well on the final overcome earlier weaknesses in performance. (This is because I didn't want their grade to reflect what they knew in Week 3, I cared what they knew at the end of the course.) In an effort to ID which students might most need extra work, we deployed a big stat effort aimed at early ID of people who would prove to do poorly in the end so we could try to intervene to help them succeed. (Everybody at GT is smart, but an 18-yr-old goofball is still an 18-year-old goofball regardless of his SAT scores.)

We got stat prof's and doctoral students and psychology faculty and everydamn body involved. We had tons of performance data, indexed by topic and difficulty. Nobody in academia had more data about where their many students succeeded and failed than I did. (IMO, most academic record-keeping and grading is a joke.) But despite a wealth of solid performance data, we couldn't do it. There were just too many variables in play that don't show up as nice, neat little numbers. So, after spending God-only-knows how much time and energy banging our heads against a wall, we eventually invented Plan B, which was to build a neural net and feed everything into it, including both detailed performance data and the subjective answers students gave on weekly surveys about how they thought they were doing, what they thought of the class, etc. Then, the neural net chugs away and does computations like crazy, searching for weights and paths through the data that prove to be reliable predictors. After it chewed on data from a year or two's worth of classes, it became able to make accurate predictions. By week 4 of a 15-week semester, the neural net was predicting the final grades each student would wind up with, and it did so with scary accuracy. But (here's the important part) we don't know why. When we looked at the factors it was giving weight to, it made no sense. It tended to discount seemingly solid factors about demonstrated performance, and it tended to give high weight to student responses on seemingly random subjective questions. It made no rational sense of any kind, but it worked like gangbusters.

The point here is that there are many factors in play, and we don't even know what most of them are. Many important factors are things that we don't have data for. The bottom line is that I could look at a kid's grades after 12 weeks of quizzes, and I would have a clue about if he was *likely* to do well on the final. But my clue was not a reliable predictor. We did develop a reliable predictor, but it relied on things that we don't understand.

There is a whole lot of the same stuff going on when you try to predict future human performance in many things, baseball included. If you disagree, then please show me a stats model that predicted in advance things like Finley, Kent, Clemens, Mora, Cuellar, etc., etc., etc.

Bottom line: I am *not* dissing stats. I am simply drawing attention to the fact that we do not have effective cookbook solutions to effectively predicting complex human performance. We do have things that give us clues, but making near-certain predictions based on those clues is a very risky business. Furthermore, it is least risky in precisely the many routine cases when everybody pretty much knows the answer anyway, stats or no stats, and most risky in precisely those "if-fy" cases where we most want/hope/pray that stats can tell us something we don't already know. As for Trax, we know he's gonna fall off the table sometime pretty soon, but we don't know exactly when. The fact that we know this about him is not predominantly due to great insights gained from stats.

[Sports Guy]

No, they would not be. Every single time someone is asked here to solely use stats to predict anything all they do is hedge and give possibilities. So in reality just how useful are these stats? They are not much better than guessing when it comes to "predicting anything." As a measurement of what has already happened they have great value but other than that they do not stand on their own as much value at all. A observation by a scout would be of more benefit in most cases.

I will give you an example. Suppose a guy is hitting 300 in the high minors but in reality he is hitting 400 on fast balls and 200 on curves. So if you look at his stats he gets promoted to the majors. Once he gets to the majors scouts have told other teams the guy has trouble with the curve. Now he gets a steady diet of even better quality (i.e. major league curves) and hits below the mendoza line and gets sent back down. Now how reliable where the stats in this case? I will tell, you they are not even presenting an accurate picture period.
First of all, anyone who looks at BA as a predictor is looking at the wrong thing.

Secondly, what else is that player doing? Are they walking a lot? Are they striking out at an alarming rate?

Pitchers or hitters league?

How old are they?

Are they showing power potential?

[Sports Guy]

So where is the value of that in predicting what he will do next year?:confused: :rolleyes:

Predicting splits is not something that is easy to do...No one has ever said it would and no one has ever said they would attempt to do this...However, if someone usually struggles vs lefties, it is pretty same to assume they are likely to do it again.

Look at Teixeira this year...I am pretty sure he hit better on the road than in Texas....Well, Texas' stadium is the best hitters park in the league and he has always hit better in Texas than on the road.

There is no way of accurately predicting something like that...It is just something that happens.

What you can predict for CPat is overall performance.

[Old#5fan]

Predicting splits is not something that is easy to do...No one has ever said it would and no one has ever said they would attempt to do this...However, if someone usually struggles vs lefties, it is pretty same to assume they are likely to do it again.

Look at Teixeira this year...I am pretty sure he hit better on the road than in Texas....Well, Texas' stadium is the best hitters park in the league and he has always hit better in Texas than on the road.

There is no way of accurately predicting something like that...It is just something that happens.

What you can predict for CPat is overall performance.

However, if someone usually struggles vs lefties, it is pretty same to assume they are likely to do it again.

Well, he is struggling with righties this season is he going to do that again? You just prove my point that stats generally are unreliable in predictions.

You may be able to predict that his overall stats will be similar but you have no clue what he is going to do leftie verus rightie next season do you? Just admit it. To tell you the truth even his overall stats may be noticeably different as he could have a season long slump or get hot all season and have a career year. We have seen that before with players like Mora and Huff.

[Old#5fan]

First of all, anyone who looks at BA as a predictor is looking at the wrong thing.

Secondly, what else is that player doing? Are they walking a lot? Are they striking out at an alarming rate?

Pitchers or hitters league?

How old are they?

Are they showing power potential?

Most of these things you can get from a scout.

[Old#5fan]

First of all, anyone who looks at BA as a predictor is looking at the wrong thing.

Secondly, what else is that player doing? Are they walking a lot? Are they striking out at an alarming rate?

Pitchers or hitters league?

How old are they?

Are they showing power potential?

So if the guy is hitting 400 in the high minors over a season ,you would need to look at a whole lot of other things? Pardon me, but I think that would get some attention.;) I could use the same example though only with the guy hitting 500 against fast balls and 100 against curves.

[hoosiers]

It's splitting hairs when I've said numbers aren't the be-all end-all many times over but my words gets twisted around to the point where people are challenging me

D,

You are the one pushing out raws stats, historical precedence and claiming Steve Trax will have to defy history to be productive next season. And yet, it appears, if your list of three individuals is accurate, that Steve Trax is already "defying history" with his performance this season. No? How many wins and what ERA+ did those other three achieve in their predictor seasons?

Is Trax's 2007 season "historically unique" based on your criteria of relevant peripheral stats of age, walks and Ks? Isn't isolating those particular criteria to predict his 2008 numbers somewhat subjective?

I certainly hope that your "word twisting" issue does not apply to me. You are outspoken on the likely uselessness of Steve Trax next season and I'd think you would be more willing to suffer fools like me when it is you who is making such authoritative comments - simillar to predictions of dismal results from Steve Trax for 2007.

[Sports Guy]

However, if someone usually struggles vs lefties, it is pretty same to assume they are likely to do it again.

Well, he is struggling with righties this season is he going to do that again? You just prove my point that stats generally are unreliable in predictions.

You may be able to predict that his overall stats will be similar but you have no clue what he is going to do leftie verus rightie next season do you? Just admit it. To tell you the truth even his overall stats may be noticeably different as he could have a season long slump or get hot all season and have a career year. We have seen that before with players like Mora and Huff.
When has anyone ever claimed that they could accurately predict splits?

Please show me where anyone said that.

Huff is another example...Up until like 2003 or 2004, he hit lefties really weel...Then that stoped the last few years...Why? Who knows?

But no on has ever said they could accurately predict splits.

What you can do is use common sense to come up something...IE, if player A struggles versus lefties year in and year out, they are likely to continue to struggle.

[Sports Guy]

Most of these things you can get from a scout.

What? This comment makes no sense.

Those are all statisical things..K's, walks, age, power and hitters or pitchers leagues are all stats...Has nothing to do with scouts.

[osfan83]

What are the odds that Cory patterson will hit 50 HR's next year?
What are the odds that Daniel Cabrera will have a sub 2.00 ERA and strike out 250?
What are the odds that Kevin Millar will swipe 40 bases next year?

Stats help you assign the odds to these events happening. Can I guarentee that Patterson won't hit 50 HR's next year? No. But based on his past performance it would be highly unlikley that he will hit 50 HR's next year.

Same for ST next year. I can't guarentee he won't have a 2.50 ERA, strike out 180 batters and win 25 games, but based on his past statistics I can give an educated guess on what he will do next year.

[Fan4Life]

D,

You are the one pushing out raws stats, historical precedence and claiming Steve Trax will have to defy history to be productive next season. And yet, it appears, if your list of three individuals is accurate, that Steve Trax is already "defying history" with his performance this season. No? How many wins and what ERA+ did those other three achieve in their predictor seasons?

Is Trax's 2007 season "historically unique" based on your criteria of relevant peripheral stats of age, walks and Ks? Isn't isolating those particular criteria to predict his 2008 numbers somewhat subjective?

I certainly hope that your "word twisting" issue does not apply to me. You are outspoken on the likely uselessness of Steve Trax next season and I'd think you would be more willing to suffer fools like me when it is you who is making such authoritative comments - simillar to predictions of dismal results from Steve Trax for 2007.

Drungo is correct that historical stats can HELP you determine future results.. meaning if 10 guys with Trachsel's stats are pitching this year, you can reasonably predict next year a number of them will collapse. And yet, some of them may go on to repeat the same performance from the previous year. So, currently, Trachsel is defying the odds, he could do so again next year... but what are the chances he will?

[Sports Guy]

Drungo is correct that historical stats can HELP you determine future results.. meaning if 10 guys with Trachsel stats are pitching this year, you can reasonably predict next year a number of them will collapse. And yet, some of them may go on to repeat the same performance from the previous year. So, currently, Trachsel is defying the odds, he could do so again next year... but what are the chances it will?

Well, it apparently has never happened before in the history of baseball. Is it smart to bank on it happening for the first time??

[Fan4Life]

Well, it has apparently has never happened before in the history of baseball. Is it smart to bank on it happening for the first time??

No, but you would have said that in April too. This assumes he can get through September in simlilar fashion so really, the jury is still out. He could flounder horribly in September and support the statistcal analysis of how this year would go.

[Old#5fan]

What? This comment makes no sense.

Those are all statisical things..K's, walks, age, power and hitters or pitchers leagues are all stats...Has nothing to do with scouts.

You mean to tell me a scout or anyone who pays close attention to watching a player bat four times a game is not able to tell if he is a Millar or Cust type or a wild hacker like Tejada? I can watch Bobby Abreu for example and tell you the guy gets a lot of walks just like Giambi because they take borderline pitches all the time. I don't need stats to see that. I can get that easily by watching the guy. Stats just confirm it I was right. I think you and some here operate the other way around. You look at stats soley because apparently you either don't watch the games enough to render a judgement or simply don't know enough by watching to know anything.

[Sports Guy]

No, but you would have said that in April too. This assumes he can get through September in simlilar fashion so really, the jury is still out. He could flounder horribly in September and support the statistcal analysis of how this year would go.Yea and he has been a statisical anomoly this year.

Betting on that is a bet you lose 9 times out 10.

[Old#5fan]

Well, it apparently has never happened before in the history of baseball. Is it smart to bank on it happening for the first time??

Maybe, maybe not. Baseball like a lot of sports is amazingly unpredictable in most cases. Hitting for the cycle never happened either until it happened. So under your theory it would not have been possible until it occurred?:rolleyes:

[Old#5fan]

Yea and he has been a statisical anomoly this year.

Betting on that is a bet you lose 9 times out 10.

So what people play the lottery with a million times worse odds yet there is a winner isn't there?:eek:

[Sports Guy]

You mean to tell me a scout or anyone who pays close attention to watching a player bat four times a game is not able to tell if he is a Millar or Cust type or a wild hacker like Tejada? Sure but this has absolutely nothing to do with you saying you can tell those things from a scout. I can look at a stats sheet and see 100 walks and 120 k's and know what type of player i have. I
can watch Bobby Abreu for example and tell you the guy gets a lot of walks just like Giambi because they take borderline pitches all the time. I don't need stats to see that.Of course you don't need a stat for that because well, there isn't a stat for that. You are just thinking of examples that make no sense.
I think you and some here operate the other way aroundDon't think.

[Fan4Life]

Yea and he has been a statisical anomoly this year.

Betting on that is a bet you lose 9 times out 10.

Agreed. I'm not advising anyone to take that kind of bet, merely pointing out as you have just stated, stats help predict what the outcome likely will be, but there are anomalies.

[Sports Guy]

So what people play the lottery with a million times worse odds yet there is a winner isn't there?:eek:

There are a lot more people who play every week spends hundreds of thousands of dollars in their lifetime and never win.

Your examples keep getting worse and worse.

[Old#5fan]

Sure but this has absolutely nothing to do with you saying you can tell those things from a scout. I can look at a stats sheet and see 100 walks and 120 k's and know what type of player i have. IOf course you don't need a stat for that because well, there isn't a stat for that. You are just thinking of examples that make no sense. Don't think.

There isn't a stat that shows Abreu is among the league leaders in walks? Why do you think that is the case? Uh, the fact he takes a lot of borderline pitches? Unbelieveable. It makes perfect sense if you watch each player and study how they approach an at bat!!:confused:

[RShack]

Well, it apparently has never happened before in the history of baseball. Is it smart to bank on it happening for the first time??
That's *exactly* the question I asked you about the wisdom and plausibility of your scheme for turning over 80% of the starting line-up... and your response was basically "Don't know... who cares?... why does it matter?"

[Old#5fan]

There are a lot more people who play every week spends hundreds of thousands of dollars in their lifetime and never win.

Your examples keep getting worse and worse.

Oh, I agree with you. I think it is a stupid waste of money. But don't tell that to people who have won the lottery, including some who have won it twice!:eek: :D

[Sports Guy]

Agreed. I'm not advising anyone to take that kind of bet, merely pointing out as you have just stated, stats help predict what the outcome likely will be, but there are anomalies.
Well of course...No one has ever said different.

You are trying to figure out the most likely outcome.

[Old#5fan]

Agreed. I'm not advising anyone to take that kind of bet, merely pointing out as you have just stated, stats help predict what the outcome likely will be, but there are anomalies.

And there are more bad predictions than good when using stats as the sole or primary basis.

I would love for all these stats gurus to predict what Aubrey Huff and Melvin Mora will put up stats wise next year?

I would also like to know how many of them predicted Baez would be so horrible this season?

[ScottieBaseball]

No, they would not be. Every single time someone is asked here to solely use stats to predict anything all they do is hedge and give possibilities. So in reality just how useful are these stats? They are not much better than guessing when it comes to "predicting anything." As a measurement of what has already happened they have great value but other than that they do not stand on their own as much value at all. A observation by a scout would be of more benefit in most cases.

I will give you an example. Suppose a guy is hitting 300 in the high minors but in reality he is hitting 400 on fast balls and 200 on curves. So if you look at his stats he gets promoted to the majors. Once he gets to the majors scouts have already told other teams the guy has trouble with the curve. Now he gets a steady diet of even better quality (i.e. major league curves) and hits below the mendoza line and gets sent back down. Now how reliable were the stats in this case? I will tell you, they are not even presenting an accurate picture period, or the kid wouldn't even be called up until he learns how to hit a curve ball much better. So relying on a scout is far better than any stats.

No one, not one single "stat head" in this thread said stats and stats alone are the sole predictor of the future. What has been repeated over and over again is that they are an effective tool widely used and accepted throughout baseball, from Little League all the way up to Major League Baseball. Scouts, managers, coaches, and fans alike all put them to use.

I challenge you to find one baseball professional, whether it's a scout or coach, that will unequivocally agree that statistics are the equivalent of a crapshoot. You might find an oddball or two that will they give stats very little thought, but I can assure the gross majority put them to use in analyzing a player's past as well as his future potential.

You do realize, by the way, you just used statistics in an attempt to support your position, but instead just proved their value, don't you? Your curveball-to-fastball ratio? That, my man, is a statistic. More often than not, a guy with a weakness like the one you described gets promoted with the hopes that he goes against the numbers and overcomes his shortcomings.

[Sports Guy]

There isn't a stat that shows Abreu is among the league leaders in walks? Why do you think that is the case? Uh, the fact he takes a lot of borderline pitches? Unbelieveable. It makes perfect sense if you watch each player and study how they approach an at bat!!:confused:

No, there is no stat that says how many borderline pitches someone takes.

[Baltimoron]

No, I don't care to sift thru 11 textbooks on stats and human performance and create a summary.

You don't have too. You wrote this:



Previous performance is the best predictor of future performance, but it is nonetheless an unreliable predictor.

For a given individual, the previous performance of that individual is a better indicator than normative data for a group, but even that is still an unreliable predictor (i.e., low r-value)


What do you mean by unreliable? (an r value?)

What "low r-value" are you thinking off?

--------------------------------------

Long story short, we know the theoretical limits on any projection system's accuracy for an offensive stat like OBP or OBA is an r in the .72-.73 (for 550 PAs) range:


What does this mean? You can take several hundred ballplayers, give them 550 PA one year, give them 550 PA another year, make sure that these guys’ true talent level in OBP does not change, make sure they play in the same parks, make sure they face the same quality of pitchers, and their year-to-year correlation will be 0.72. That is, the absolute maximum year-to-year r you can hope for, given a large number of ballplayers is .72.

How about instead of OBP, we look at wOBA (which is analogous to OPS)? Here, our var(true) is .036^2, var(random) is .022^2, and var(observed) is .042^2. Our r is .73.

So, when looking at forecasters, and you look at their correlation coefficient of their forecast to the actual results, anything close to .73 means that they did as good a job as possible. (They could actually go over that level, since the number of players in their sample is still small enough that the level of uncertainty of that r will be a bit high. But, given thousands of players over several years, that uncertainty level will drop quite a bit.)
link (http://www.insidethebook.com/ee/index.php/site/article/forecasters_how_accurate_can_they_possibly_be/)

Here are some r value for various projection systems from 2006:


Shandler .702
James .685
ZIPs .684
Chone .677
Marcel .664
link (http://lanaheimangelfan.blogspot.com/2006/12/more-projection-stuff.html)

Perhaps its semantics, but I don't consider these to be "unreliable predictors".

-----------------------------------------------------

Obviously, its much worse for pitchers because of all the noise in ERA, but less so if you use component ERA.

Here are some r-value for various systems projecting 2006 ERA numbers:

.459 ZIPs
.445 Baseball Info Solutions (Bill James has nothing to do with this although its in his Handbook, he claims it can't be done)
.442 Marcel
.423 Shandler

ERA itself has an r of .29
link (http://lanaheimangelfan.blogspot.com/2006/12/pitcher-projections.html)

[Sports Guy]

And there are more bad predictions than good when using stats as the sole or primary basis.

I would love for all these stats gurus to predict what Aubrey Huff and Melvin Mora will put up stats wise next year?

I would also like to know how many of them predicted Baez would be so horrible this season?
Many of us predicted Baez would be average at best....No one can really predict someone to be as bad as he has been but you can predict him to be subpar and many did.

[RShack]

Long story short, we know the theoretical limits on any projection system's accuracy for an offensive stat like OBP or OBA is an r in the .72-.73 (for 550 PAs) range:
Right. And .72-.73 is a generally a lousy predictor. The fact that it's the best we can do, and is less bad than others, doesn't change the fact that it's lousy.

[osfan83]

And there are more bad predictions than good when using stats as the sole or primary basis.

I would love for all these stats gurus to predict what Aubrey Huff and Melvin Mora will put up stats wise next year?

I would also like to know how many of them predicted Baez would be so horrible this season?

I predict Huff will have between 15-30 Hr's next year.
I predict Mora will hit between .250 and .290 next year.

If you want we can give the likleyhood of Huff hitting 15-18, 19-22, 22-25, 25-30 and 30+ HR's.

You guys are looking for guarentees, and there arn't any...but there are odds. I can't guerantee you the O's won't make the playoffs this year, they are still mathmatically alive, but I can give you the odds of it happening.

ST might be a Cy Young award winner next year, but the odds point to him being one of the worst 20 starting pitchers in the league next year.

[Baltimoron]

We have imperfect memories and are biased by stuff like primacy and recency. Stats are actual data. Think of stats as a proxy for a perfect memory.

Without any stats, how accurately could you assess the offensive performance of the Orioles this year? Even assuming you watched every game, without recorded data, how good do you think your estimates would be of who did what?

Stats aren't the problem. They are just data. Its when this data is misused or misinterpreted or mis-cited that the problem arises.

[Baltimoron]

Right. And .72-.73 is a lousy predictor. The fact that it's the best we can do, and is less bad than others, doesn't change the fact that it's lousy.

My head just asploded.

.72-.73 is essentially perfect. There is no less bad than others.

[RShack]

My head just asploded.

.72-.73 is essentially perfect. There is no less bad than others.
What do you think the implications are of a coefficient of .72-.73. How likely is that to prove to be correct? 95%? What do you think it means?

[ScottieBaseball]

No, they would not be. Every single time someone is asked here to solely use stats to predict anything all they do is hedge and give possibilities. So in reality just how useful are these stats? They are not much better than guessing when it comes to "predicting anything." As a measurement of what has already happened they have great value but other than that they do not stand on their own as much value at all. A observation by a scout would be of more benefit in most cases.

I will give you an example. Suppose a guy is hitting 300 in the high minors but in reality he is hitting 400 on fast balls and 200 on curves. So if you look at his stats he gets promoted to the majors. Once he gets to the majors scouts have already told other teams the guy has trouble with the curve. Now he gets a steady diet of even better quality (i.e. major league curves) and hits below the mendoza line and gets sent back down. Now how reliable were the stats in this case? I will tell you, they are not even presenting an accurate picture period, or the kid wouldn't even be called up until he learns how to hit a curve ball much better. So relying on a scout is far better than any stats.

No one, not one single "stat head" in this thread said stats and stats alone are the sole predictor of the future. What has been repeated over and over again is that they are an effective tool widely used and accepted throughout baseball, from Little League all the way up to Major League Baseball. Scouts, managers, coaches, and fans alike all put them to use.

I challenge you to find one baseball professional, whether it's a scout or coach, that will unequivocally agree that statistics are the equivalent of a crapshoot. You might find an oddball or two that will they give stats very little thought, but I can assure the gross majority put them to use in analyzing a player's past as well as his future potential.

You do realize, by the way, you just used statistics in an attempt to support your position, but instead just proved their value, don't you? Your curveball-to-fastball ratio? That, my man, is a statistic. More often than not, a guy with a weakness like the one you described gets promoted with the hopes that he goes against the numbers and overcomes his shortcomings.

[Baltimoron]

What do you think the implications are of a coefficient of .72-.73. How likely is that to prove to be correct? 95%? What do you think it means?

It means an omniscient being could not due better than .72-.73. How is that lousy? Without knowing the future, there can be no better projection.

[Old#5fan]

I predict Huff will have between 15-30 Hr's next year.
I predict Mora will hit between .250 and .290 next year.

If you want we can give the likleyhood of Huff hitting 15-18, 19-22, 22-25, 25-30 and 30+ HR's.

You guys are looking for guarentees, and there arn't any...but there are odds. I can't guerantee you the O's won't make the playoffs this year, they are still mathmatically alive, but I can give you the odds of it happening.

ST might be a Cy Young award winner next year, but the odds point to him being one of the worst 20 starting pitchers in the league next year.

The point I am making those predictions ( I assume you based them on past stats) are so wide in scope that me or anyone else could make them without knowing nada about their full stat history. All these stats thrown out by gurus are still no better than a guess (crap shoot) or roll of the dice when it comes to what "might" happen.

[Baltimoron]

The point I am making those predictions ( I assume you based them on past stats) are so wide in scope that me or anyone else could make them without knowing nada about their full stat history. All these stats thrown out by gurus are still no better than a guess (crap shoot) or roll of the dice when it comes to what "might" happen.

The funny thing is that there are posts in this thread that mathematically demonstrate your claim is utter nonsense.

[Old#5fan]

No one, not one single "stat head" in this thread said stats and stats alone are the sole predictor of the future. What has been repeated over and over again is that they are an effective tool widely used and accepted throughout baseball, from Little League all the way up to Major League Baseball. Scouts, managers, coaches, and fans alike all put them to use.

I challenge you to find one baseball professional, whether it's a scout or coach, that will unequivocally agree that statistics are the equivalent of a crapshoot. You might find an oddball or two that will they give stats very little thought, but I can assure the gross majority put them to use in analyzing a player's past as well as his future potential.

You do realize, by the way, you just used statistics in an attempt to support your position, but instead just proved their value, don't you? Your curveball-to-fastball ratio? That, my man, is a statistic. More often than not, a guy with a weakness like the one you described gets promoted with the hopes that he goes against the numbers and overcomes his shortcomings.

It is a stat that is not kept as an official stat but only informally by scouts.

I am being misconstrued here. I am not saying that statistics are the equivilent of a crapshoot but when trying to predict the future of minor league player translating to a big league sucess story I believe they are. For example I have seen it posted on this forum that there are no "can't miss" major league prospects in the Orioles system right now (not counting Wieters). Yet, that doesn't mean that no current Oriole minor leaguer will not advance to a sucessful major league career. Nonetheless nobody was really willing to use stats to name who would make it.

Heck, I can throw out guess that Olsen, Hoey, and Liz might make it, and I don't have any idea what their minor league stats look like. However, I have watched them pitch with the Orioles and they appear to have potential. I have also seen guys who I think have little potential and one of them is Birkins and the other is Bell. However, I don't see any stats people indicating what their position is on them.

I recall a pitcher the Orioles had for a short time named Borowski who for some reason they let get away who is now having a pretty solid career as a closer. Where there any stats gurus who thought this guy would pan out? AGain, I just don't see stats as much use as a reliable [B]predictor [/I]of anything.

[RShack]

It means an omniscient being could not due better than .72-.73. How is that lousy? Without knowing the future, there can be no better projection.
No it doesn't. You're just referring to the limits of available techniques for a certain kind of problem. In other problems, r values go way up into the mid-to-high .90's (as they approach 1.0 which is a perfect correlation). What are they? Better than omniscient? ;-)

A value of .73 is way different than omniscience. If you take something that's .72-.73 and make a prediction based on that, how likely is that prediction to be wrong? This is not a question about omniscience, it's a question about accuracy of the prediction, i.e., how much you can trust it to be right?

[Baltimoron]

Think of it this way.

We have a coin that is perfectly balanced. We know that is we flip it we have a 50/50 shot of heads or tails.

We flip it ten times and get 6 heads and 4 tails.

Do people think this proves the 50/50 split is wrong, becuase it isn't 5 and 5?

Or do people get that the 50/50 is correct and that within a given sample sizes there is a certain amount or expected random deviation, but as the sample size grows the results converge towards their expected true value (the 50/50 split) ("the law of big numbers").

[Old#5fan]

Think of it this way.

We have a coin that is perfectly balanced. We know that is we flip it we have a 50/50 shot of heads or tails.

We flip it ten times and get 6 heads and 4 tails.

Do people think this proves the 50/50 split is wrong, becuase it isn't 5 and 5?

Or do people get that the 50/50 is correct and that within a given sample sizes there is a certain amount or expected random deviation, but as the sample size grows the results converge towards their expected true value (the 50/50 split) ("the law of big numbers").

I don't know what your example has to do with predicting how the next 10 flips will go, because you have no idea. You can only guess based on odds. To me that is a whole lot different than accurately predicting something. There are way too many variables in baseball to make meaningful predictions beyond a guess regarding what will happen in advance.

[Baltimoron]

No it doesn't. You're just referring to the limits of available techniques for a certain kind of problem. In other problems, r values go way up into the .90's as they approach 1.0 which is certainty. A value of .73 is way different than omniscience. If you take something that's .72-.73 and make a prediction based on that, how likely is that prediction to be wrong? This is not a question about omniscience, it's a question about accuracy of the prediction, i.e., how much you can trust it to be right. You don't need omniscience for that.

what in the world are you talking about?

The reason the r value is low is because we are limited by sample size. You get that if we give each guy 650 PAs (instead of 550), the r goes up, right?

[osfan83]

The point I am making those predictions ( I assume you based them on past stats) are so wide in scope that me or anyone else could make them without knowing nada about their full stat history. All these stats thrown out by gurus are still no better than a guess (crap shoot) or roll of the dice when it comes to what "might" happen.


An Educated guess is better than a guess. If it were truley a crapshoot, then why not put Paul Bako in your cleanup spot. He has just as much of a chance of hitting 40 HR's as Tejada or Markakis!

[Old#5fan]

An Educated guess is better than a guess. If it were truley a crapshoot, then why not put Paul Bako in your cleanup spot. He has just as much of a chance of hitting 40 HR's as Tejada or Markakis!

True, but you don't need stats to tell you Bako isn't a very good hitter. All you need to do is watch how anemic he looks at the plate!:p


I get what you are espousing though. Stats have their place but I still think they are of limited use in trying to predict the future especially in whether a player with strong minor league stats will carry those to the majors. That is where I think it is mostly a crapshoot.

[Baltimoron]

I don't know what your example has to do with predicting how the next 10 flips will go, because you have no idea. You can only guess based on odds. To me that is a whole lot different than accurately predicting something. There are way too many variables in baseball to make meaningful predictions beyond a guess regarding what will happen in advance.

You completely missed my point. Just as divergence from the 50/50 over sample size x does not mean the 50/50 is wrong, divergence from a projection over the course of a season does not mean the projection is wrong.

The bolded part is wrong and has been clearly refuted in this thread.

The primary limiting factor is sample size, not the failure of the underlying model to account for the pertinent variables.

[Old#5fan]

You completely missed my point. Just as divergence from the 50/50 over sample size x does not mean the 50/50 is wrong, divergence form a projection does not mean the projection is wrong.

The bolded part is wrong and has been clearly refuted in this thread.

The primary liming factor is sample size, not the failure of the underlying model to account for the pertinent variables.

To me a projection is different from a predicition but we could just be arguing semantics. IMHO a projection is just a wide range of possible outcomes, where a prediction is much more specific and on target.

I don't view them as one and the same.

[Old#5fan]

We have imperfect memories and are biased by stuff like primacy and recency. Stats are actual data. Think of stats as a proxy for a perfect memory.

Without any stats, how accurately could you assess the offensive performance of the Orioles this year? Even assuming you watched every game, without recorded data, how good do you think your estimates would be of who did what?

Stats aren't the problem. They are just data. Its when this data is misused or misinterpreted or mis-cited that the problem arises.

This makes sense.

[Baltimoron]

To me a projection is different from a predicition but we could just be arguing semantics. IMHO a projection is just a wide range of possible outcomes, where a prediction is much more specific and on target.

I don't view them as one and the same.

OK, but consider:


Noun

* S: (n) projection (a prediction made by extrapolating from past observations)
link (http://wordnet.princeton.edu/perl/webwn?s=projection)

[Sports Guy]

True, but you don't need stats to tell you Bako isn't a very good hitter. All you need to do is watch how anemic he looks at the plate!:p


I get what you are espousing though. Stats have their place but I still think they are of limited use in trying to predict the future especially in whether a player with strong minor league stats will carry those to the majors. That is where I think it is mostly a crapshoot.

Again, you are just completely wrong.

This isn't an opinion...You are just incorrect.

I am convinced that you are Eli Eon...You guys say the same things.

[RShack]

what in the world are you talking about?

The reason the r value is low is because we are limited by sample size. You get that if we give each guy 650 PAs (instead of 550), the r goes up, right?
Various things change with sample size, including P values. But let's not worry about that, because stats are (if nothing else) tedious. An r value is just a statement about the degree of correlation between 2 things. If you take one thing (your prediction about how somebody does) and another thing (how that person actually does), and if you do that for a bunch of persons, and you plot them all... then if all your predictions line up perfectly with how everybody actually did, then your r value is 1.0, it's a perfect correlation between what your predictive tool said and what actually happened. That's omniscience. The fact that we can't do better than .72-.73 for this stuff is not a statement about omniscience or perfect correlations (that's 1.0). It's just a statement about the state of the art for this kind of problem.

Not that r values mean everything here, because we don't necessarily care about exact correlations. That's just what everybody turns to first. There are things going on here that really present dilemmas about whether we care about parametrics vs. non-parametrics, whether we care about correlation vs. determination, the list goes on and on. I think statistics is not a good topic for a message board. However, you seem to be confusing how one technique for baseball compares to others, when the core issue is how the best technique compares to ominiscience (aka after-the-fact inspection of reality). By that standard, .73 is not strong. For example, many cookbook summaries (which I do not recommend) say that anything below .50 is weak and anything above .80 is strong. That's way oversimplifying things. At the same time, it's not until you get well into the .90's that you can really stand by a specific correlation; until then you can just say that there's an imperfect relationship of some kind. If we were talking about predicting the behavior of building materials, you wouldn't wanna build a bridge based on that; you'd have to intentionally overbuild it to compensate.

But again, this is not a good topic for a message board. Anybody but you-and-me is either asleep by now or isn't reading it anyway...

[Fan4Life]

Various things change with sample size, including P values. But let's not worry about that, because stats are (if nothing else) tedious. An r value is just a statement about the degree of correlation between 2 things. If you take one thing (your prediction about how somebody does) and another thing (how that person actually does), and if you do that for a bunch of persons, and you plot them all... then if all your predictions line up perfectly with how everybody actually did, then your r value is 1.0, it's a perfect correlation between what your predictive tool said and what actually happened. That's omniscience. The fact that we can't do better than .72-.73 for this stuff is not a statement about omniscience or perfect correlations (that's 1.0). It's just a statement about the state of the art for this kind of problem.

Not that r values mean everything here, because we don't necessarily care about exact correlations. That's just what everybody turns to first. There are things going on here that really present dilemmas about whether we care about parametrics vs. non-parametrics, whether we care about correlation vs. determination, the list goes on and on. I think statistics is not a good topic for a message board. However, you seem to be confusing how one technique for baseball compares to others, when the core issue is how the best technique compares to ominiscience (aka after-the-fact inspection of reality). By that standard, .73 is not strong. For example, many cookbook summaries (which I do not recommend) say that anything below .50 is weak and anything above .80 is strong. That's way oversimplifying things. At the same time, it's not until you get well into the .90's that you can really stand by a specific correlation; until then you can just say that there's an imperfect relationship of some kind. If we were talking about predicting the behavior of building materials, you wouldn't wanna build a bridge based on that; you'd have to intentionally overbuild it to compensate.

Are you then agreeing the tool we use to try and predict a players outcome in the future is unreliable?

[Old#5fan]

Again, you are just completely wrong.

This isn't an opinion...You are just incorrect.

I am convinced that you are Eli Eon...You guys say the same things.


Well, maybe incorrect to you but to me I think I just view stats in a more secondary postion to you in that I think they are subordinate to experienced observation. I will give you an example. Assuming you are a guy who likes well built women if one walked by you she would catch your eye. Right away you would know that she was built like the proverbial ---- house correct? Now you later see her on tv in the Miss America competition in a swimsuit andget stats that she is built specifically as 36-24-36. Now, what did these stats do for you? If the answer is confirm what you already knew you understand how I view the majority of stats and their usage.

Another example, when you tried out for Little League as a kid and got to hit in tryouts. You see a kid mashing pitch after pitch like a young Tejada and then you see the usual chump kid who can't make contact unless the coach throws the ball and hits his bat. Do you need to look at this kid's "stats" in BP to know which one you want hitting third on your team or even playing? Then when the season starts and the Tejada like kid is second in the league in batting average and third in homers, whereas the other kid had a couple of fluke hits and hit like 100 did you really need these stats to tell you anything than what you already knew?

[Old#5fan]

Are you then agreeing the tool we use to try and predict a players outcome in the future is unreliable?

I thought I read where he was asked this and said that stats are the most reliable tool available but still overall or generally unreliable. (Correct me if I am wrong).

[Fan4Life]

Well, maybe incorrect to you but to me I think I just view stats in a more secondary postion to you in that I think they are subordinate to experienced observation. I will give you an example. Assuming you are a guy who likes well built women if one walked by you she would catch your eye. Right away you would know that she was built like the proverbial ---- house correct? Now you later see her on tv in the Miss America competition in a swimsuit andget stats that she is built specifically as 36-24-36. Now, what did these stats do for you? If the answer is confirm what you already knew you understand how I view the majority of stats and their usage.

Another example, when you tried out for Little League as a kid and got to hit in tryouts. You see a kid mashing pitch after pitch like a young Tejada and then you see the usual chump kid who can't make contact unless the coach throws the ball and hits his bat. Do you need to look at this kid's "stats" in BP to know which one you want hitting third on your team or even playing?

OK, but lets take another step.... you are the manager of a AA minor league team and you have these same two kids on your team.... you observe the same information from both that you did in little league, one is a very good hitter, and one is a rather poor hitter. Their stats obviously confirm what you already knew. In your view, if just looking at their stats, do you think that you wouldn't be able to determine which player is more likely to succeed at the MLB level than the other? Or would it be a crapshoot?

[RShack]

Are you then agreeing the tool we use to try and predict a players outcome in the future is unreliable?
Sorry to be obstuse about this, but it sorta depends on what you mean by "the tool we try to use" (stats in general? specific calculations? what?) and exactly what you mean by prediction ("here's what his OPS will be" vs. "he will do worse than this year"). In general, I think that it's fine to use stats as providing rules of thumb about what is likely. "Likely" is a fuzzy term that is treated with dubious significance. What I take issue with is people using stats as if they provide a near-certain prediction, but then absolving themsevles by saying, "hey, I never said it was certain". The problem is that people routinely treat them as if they are near certain, but then pretend they didn't. I don't think this is due to "bad will" or "dishonesty". I think it's because (a) stats are more readily available than are detailed evaluations by knowledgeable observers, (b) the fact that they are expressed as numbers somehow provides them with an aura of exactitude and legitimacy, (c) people really, really want something that they can look at to educate themselves, and (d) understanding what they can do and not do is a complex subject that requires much tedious and boring study. I think these factors conspire in a way that causes people to accidentally misuse stats.

Short answer: Rule of thumb? Sure. Most likely result? Sure. A solid basis for predicting what will actually happen? No way, especially in the cases that are the most interesting, i.e., the cases where we most want stats to help us.

[Fan4Life]

Sorry to be obstuse about this, but it sorta depends on what you mean by "the tool we try to use" (stats in general? specific calculations? what?) and exactly what you mean by prediction ("here's what his OPS will be" vs. "he will do worse than this year"). In general, I think that it's fine to use stats as providing rules of thumb about what is likely. "Likely" is a fuzzy term that is treated with dubious significance. What I take issue with is people using stat-based predictions as if they provide a near-certain prediction, but then absolving themsevles by saying, "hey, I never said it was certain". The problem is that people routinely treat them as if they are near certain, but then pretend they didn't. I don't think this is due to "bad will" or "dishonesty". I think it's because (a) stats are more readily available than are detailed evaluations by knowledgeable observers, (b) the fact that they are expressed as numbers somehow provides them with an aura of exactitude and legitimacy, and (c) people really, really want something that they can look at to educate themselves. I think these factors conspire in a way that causes people to accidentally misuse stats.

Short answer: Rule of thumb? Sure. Most likely result? Sure. A solid basis for predicting what will actually happen? No way, especially in the interesting and exceptional cases that are exactly the most interesting cases.


Excellent, now... can you tell me if there exists a way to correlate the r for a statistical prediction(perhaps the range predicted for any given category of a players recent history or history in general)? As in, has there ever been a study that looked at over a given period of time what the r has been for any given group of players? And I agree with you in that when making predictions, some people relazing that the odds are in their favor tend to espouse them as certain, rather than likely. So when the anomaly arrives, which as the coin flip post provided, can arrive at differing intervals, there they are... wondering what happened? For me, the human factor is something that can have a great affect and cannot be measured in stats. Whether your fighting with your wife, made a bad business decision, got into an arguement with your mother, your uniform doesn't feel comfortable as usual.... you didn't rest well last night, something you ate isn't agreeing with you... you met someone and you can't stop thinking about her... the media printed something about you that is bothering you....I realize these guys are professionals and generally speaking can focus where and when necessary... but they aren't robots...

[RShack]

I thought I read where he was asked this and said that stats are the most reliable tool available but still overall or generally unreliable. (Correct me if I am wrong).
I don't know if they are more or less trustworthy than the predictions of the best evaluators of talent. They are more reliable in the sense that stats will tell you the same thing, again and again, while different evaluators of talent will vary in what they tell you. For the purposes used here, using stats to predict future performance of an individual is not nearly as strong as some people would have you believe. Say it's accurate 68% of the time. That's better than random, but it's still far from something you'd bank on in any given case. Over a large number of cases, you could make a fortune betting that way, but that doesn't really tell you whether or not sign this guy or that guy.

[RShack]

Excellent, now... can you tell me if there exists a way to correlate the r for a statistical prediction(perhaps the range predicted for any given category of a players recent history or history in general)? As in, has there ever been a study that looked at over a given period of time what the r has been for any given group of players? And I agree with you in that when making predictions, some people relazing that the odds are in their favor tend to espouse them as certain, rather than likely. So when the anomaly arrives, which as the coin flip post provided, can arrive at differing intervals, there they are... wondering what happened? For me, the human factor is something that can have a great affect and cannot be measured in stats. Whether your fighting with your wife, made a bad business decision, got into an arguement with your mother, your uniform doesn't feel comfortable as usual.... you didn't rest well last night, something you ate isn't agreeing with you... you met someone and you can't stop thinking about her... the media printed something about you that is bothering you....I realize these guys are professionals and generally speaking can focus where and when necessary... but they aren't robots...
When you are talking about ranges of performance and specific subgroups of players, and you wanna do that in any mildly comprehensive way, then things get more complicated. It would take me a huge bunch of tedious and unpleasant work to answer your question. I'm sure that there are a few people around here, such as Drungo and others, who could respond to that much more quickly than I without needing near the homework I would have to do. Make no mistake, there are people here who know way more about baseball stats than I do.

You are correct in observing that baseball stats fail to capture many things that are perhaps relevant. In addition to the factors you mention, others include things like a guy changing his stance, changing how he guesses about what pitch is coming, changes in the way opposing pitchers treat him, the list is endless. This does not in any way devalue stats for reaching normative conclusions, i.e., statements about what is normal over many instances. But the absence of relevant data does seriously limit our ability to adequately model what happens. People think it's just balls and bats and numbers. It's not. It's people too. Chief among the things we don't know how to model well is almost anything that concerns specific people's behavior.

[hoosiers]

Drungo is correct that historical stats can HELP you determine future results.. meaning if 10 guys with Trachsel's stats are pitching this year, you can reasonably predict next year a number of them will collapse. And yet, some of them may go on to repeat the same performance from the previous year. So, currently, Trachsel is defying the odds, he could do so again next year... but what are the chances he will?

Well, here's the gist of my problem with those who speak so authoritatively on baseball statistics and their predictive capabilities.

If you ask me, Drungo's original assertion is based on subjective, well-intentioned, criteria - defined as Age, Innings Pitched, K/9 and W/9. I believe Drungo was trying to perform a reasonable analysis using reasonable stats and certainly do not begrudge that effort.

However, why should we accept those as the most relevant predictors, based on baseball history, of Trax's productivity next year? First, why is age a serious factor - especially when Trax already appears to be making the age comp utterly useless this year? Further, why get so excited and attempt to draw such definitive conclusions about criteria that limit the sample size to three pitchers since 1961?

Accepting the ridiculously small sample size, why not START by comparing Trax to those three pitchers in Year X? Is Trax already pitching the pants off of those guys in Year X - using quality starts or ERA+ or some accepted stat? If he is, wouldn't that make the comparative samples a LOT less relevant from which to draw the "edge of possible" comparison? Shouldn't that make one a little less quick to conclude Trax will fall into the historical pattern of those three particular pitchers in Year X+1?

Second, more importantly, why is 140 IP a factor? Of course, one can nit-pick the underlying assertions in any analysis, but I do not mean to do that here. I could stand to be enlightened on why IP is relevant as a predictor for next year's X+1 results.

I just have enormous problems with subjectively chosen stats that arrive at a three pitcher sample size in the expansion ERA and conclude "What Trachsel is doing right now is at the very edge of what's possible". I don't think that's a logical conclusion from a small sample size at all. I think a better conclusion would be to scratch one's head and say, "this doesn't really work, let's run the numbers again a different way".

[Baltimoron]

Various things change with sample size, including P values. But let's not worry about that, because stats are (if nothing else) tedious. An r value is just a statement about the degree of correlation between 2 things. If you take one thing (your prediction about how somebody does) and another thing (how that person actually does), and if you do that for a bunch of persons, and you plot them all... then if all your predictions line up perfectly with how everybody actually did, then your r value is 1.0, it's a perfect correlation between what your predictive tool said and what actually happened. That's omniscience. The fact that we can't do better than .72-.73 for this stuff is not a statement about omniscience or perfect correlations (that's 1.0). It's just a statement about the state of the art for this kind of problem.

Not that r values mean everything here, because we don't necessarily care about exact correlations. That's just what everybody turns to first. There are things going on here that really present dilemmas about whether we care about parametrics vs. non-parametrics, whether we care about correlation vs. determination, the list goes on and on. I think statistics is not a good topic for a message board. However, you seem to be confusing how one technique for baseball compares to others, when the core issue is how the best technique compares to ominiscience (aka after-the-fact inspection of reality). By that standard, .73 is not strong. For example, many cookbook summaries (which I do not recommend) say that anything below .50 is weak and anything above .80 is strong. That's way oversimplifying things. At the same time, it's not until you get well into the .90's that you can really stand by a specific correlation; until then you can just say that there's an imperfect relationship of some kind. If we were talking about predicting the behavior of building materials, you wouldn't wanna build a bridge based on that; you'd have to intentionally overbuild it to compensate.

But again, this is not a good topic for a message board. Anybody but you-and-me is either asleep by now or isn't reading it anyway...

You sound dangerously close to trolling. Here's the post you made to which I responded:


I realize that you don't like uncertainty, but it's a very real and important part of life, including baseball. Sadly, the most useful perspective does not reduce to a simple soundbite, because it it multi-dimensional. It is not reducible to one bullet point, there are several:
Stats are a valuable tool. It would be dumb to not consider what they suggest.
Previous performance is the best predictor of future performance, but it is nonetheless an unreliable predictor.
For a given individual, the previous performance of that individual is a better indicator than normative data for a group, but even that is still an unreliable predictor (i.e., low r-value)
The sensible thing to do is to consider multiple factors, including past performance and detailed observations of current performance by knowledgeable observers/evaluators.
Applying the proper weight to the various factors is something that is still more art than science, and is likely to continue that way for the foreseeable future. The error that is routinely made around here is to pretend otherwise.

The boded part is wrong.

That the r does not equal 1 does not mean the system is flawed, or that "Applying the proper weight to the various factors is something that is still more art than science"

If you had a perfect projection system (i.e an omniscient* "all knowing" being) and knew exactly how good a player was, your r, or the correlation between your projection and the player's results over 500 or 600 or whatever plate appearances is still hampered by sample size and the expected distribution over that sample size.

To work with the coin example, we know (i.e. are "omniscient") the coin is 50/50, but we also know that over x trials, we don't expect the results to be 50/50. That over x trials the "r" is less than 1 does not mean the coin is not 50/50.

There is no suggestion that we can know the future. My point is we have projection systems that rely on what a player has done that come very close to being as perfect as they can be given we can't and don't know the future. This means your contention that "Applying the proper weight to the various factors is something that is still more art than science" is wrong.

The issue is not weighing factors differently because the projection models do not accurately model reality. The primary issue is our sample size is not large enough. The models aren't prefect, but they are pretty damn close. And they are very reliable.


* Here is how I used first used omniscient in this thread:


It means an omniscient being could not due better than .72-.73. How is that lousy? Without knowing the future, there can be no better projection.

Why do you use the word differently - "ominiscience (aka after-the-fact inspection of reality)" - to suggest it means one who knows the future, when its clear I am not using it that sense?

[RShack]

You sound dangerously close to trolling.
So much for trying to put effortfull content into posts.

Believe whatever you want about stats, everybody else does (both pro and con). If you think it's a science, not an art, to know how to weigh stats vs. talent evaluation, maybe you should take a science course.

You should probably discuss it further with somebody who's not a troll.

[ScottieBaseball]

Well, maybe incorrect to you but to me I think I just view stats in a more secondary postion to you in that I think they are subordinate to experienced observation.

Despite my earlier defense of statistics and their value, I personally put experienced observation ahead of stats sometimes, too. There are some very, very intangible elements to the game of baseball and that's part of the appeal to me. Gut feelings and hunches apply more here than in any other sport in my opinion.

I applaud you for the above sentence I snipped from your post. I took it as a roundabout concession to agree to disagree, and the ability to do so and to see things from someone else's side of the table even if it's presented in an abrasive fashion is what can make a message board interesting, entertaining, and enjoyable.

[Migrant Redbird]

For me, stats are only as good as the consistency level they represent. I'm not sure there is a way to evaluate that other than looking at the stats in a more micro level. As a hitter, going 12-40 can be broken down into a 1-20 slump followed by an 11-20 streak... or a 4-14, 4-13, 4-13, etc etc.... their stats will look the same in the aggregate but they are much different hitters if those types of numbers are typical of each player.

I think that you're heading in the wrong direction if you attempt to slice and dice statistics from a relatively small sample size into even smaller samples.

Baseball statistics are not random numbers, although we often treat them that way for purposes of analysis. The outcome of each at bat is the result of a pitcher and a hitter attempting to outperform each other, and their success depends in part upon how well each performed and in part upon the chance occurrence of where the batted ball flew. Wee Willie Keeler may have "hit 'em where they ain't", but the average major league hitter rarely is able to control the vector of the batted ball anywhere near that finely (except occasionally on a bunt). A hitter might succeed in pulling the ball or hitting it the other way, but that's about the limits of his control and the rest is fortuitous circumstance.

So if we have a group of 20 at bats by an individual batter, it's entirely possible that the batter was doing something wrong on everyone of those at bats -- cocking his bat a split second early; dropping his elbow at the onset of his swing; blinking at the instant the pitcher released the ball; etc. -- which made it very unlikely indeed that any of those 20 at bats was likely to result in a base hit. Conversely, sometimes a batter will get into a zone where his timing is perfect and the baseball looks as big as a basketball as it approaches the plate, to the point that he's hitting everything that gets thrown up to the plate and even his outs are hard hit line drives that were unfortunately hit right at a defensive player. These at bats, neither the ones of the slump nor the ones of the hitting streak, are not "random" events in the sense that a coin toss has an equal chance of coming up heads or tails. The at bats are deterministic, in which a given set of inputs will produce the same outcome repeatedly, just as long as the inputs ALL remain constant, regardless of how many times you repeat the event.

Of course, the inputs not remain constant (the pitcher varies his pitch velocity, movement, and location while the hitter attempts to adjust), so the outcomes aren't really identical. Albert Pujols does not hit a home run every time Brad Lidge delivers a pitch (although maybe he does in Brad's nightmares). Not only do the inputs vary on each pitch, but most of them can't be determined to any great degree of success. Managers and coaches try (Dave Duncan is renowned for his extensive collection of books charting the result of every pitch location and type for each hitter the Cardinals face) and some tentative conclusions can be drawn from the very incomplete data, but as a rule we're dealing with a large set of unknown input factors producing an outcome which is deterministic -- predetermined -- but which is pseudo random from our perspective. When you encrypt an email, the encrypted bits which you send to the recipient appear to be pseudo random, but they're anything but, or the recipient would never be able to decrypt and read what you sent him.

Thus, a sample of at bats by a hitter is a collection of pseudo random event outcomes which are pre-ordained, but can be treated as a collection of true random events if one has a sufficiently large sample size. If you take that large sample and slice it up into smaller samples for the purpose of subjecting it to detailed analysis, you're largely defeating your purpose and any conclusions which you draw from the smaller sample sizes are likely to be quite invalid.

We call these smaller samples "splits", and there have been more silly and boneheaded conclusions drawn from splits than from any other statistical weapon in the fan's arsenal. Why don't Corey Patterson's platoon differentials for 2007 look like the ones from 2006? Well, because the sample size has been drawn too small and Corey's platoon splits for 2007 are based upon sufficiently different input data from those in 2006 that any analysis -- other than to observe that a difference exists -- is fruitless.

You want to observe how much splits meander all over the place? Well, look at the monthly splits for practically any major league hitter or pitcher -- good or bad -- and you will see variations which dwarf those which normally occur from season to season. Let's take the estimable Steve Trachsel for an example:


Split GS W L CG IP ERA H R ER HR BB SO
April/Mar. 5 1 2 0 28.1 4.13 22 13 13 2 14 11
May 6 3 1 1 38 2.84 33 13 12 3 17 7
June 6 1 3 0 26.1 8.89 40 26 26 6 14 13
July 2 0 1 0 10 8.10 16 9 9 3 6 6
August 6 1 1 0 38 2.37 40 12 10 2 18 8

The splits do capture accurately the outcomes of Trachsel's pitching during each month of the 2007 season, but I defy anyone to take one of those monthly splits and tell me what kind of a pitcher Trachsel was for the entire season or to project from any month how he would pitch in any other unknown month this season or any other. There's simply way too much variation from split to monthly split for the splits to have any use except to say that Trachsel had a lousy June and a blistering August. And you will find similar variations from month to month for most hitters. If you attempt to draw profound conclusions by evaluating such small samples, your efforts are doomed to failure. If you slice the samples up even finer, as in the example of a 1-20 slump and an 11-20 streak, you're just magnifying the folly.

How large should a hitter's at bat sample be to support meaningful conclusions? The answer, as it usually is, is "It depends." I've contended frequently that something approaching a full season is needed; you can not evaluate David Newhan's prowess as a hitter based upon the 1.324 OPS he posted in June of 2004 (as much as fans were tempted to do it); even David's .814 OPS over 412 plate appearances in the 2004 season was not really a valid indicator of his major league hitting skills, as his subsequent OPS+ of 62, 73, and 57 in the following 3 seasons have demonstrated.

So, ideally, you would accumulate thousands of plate appearances over a player's career in order to evaluate him and project his future performance. Unfortunately, you can't really do that anywhere outside a Diamond Mind simulation. Tejada's platoon differential OPS in 2006 was an incredibly balanced .878 in 533 PAs against righthanders and an identical .878 in 176 PAs against southpaws, but one must guard against drawing any conclusions from that similarity. In 2005, the splits were .877 (close!) and .828, but in 2004 they were .856 and 1.012. Over his entire career, Tejada's differential was .815/.840, and even the LH PAs appear to be an adequate sample size (over 1700 plate appearances), but keep in mind that the Tejada who was putting up those numbers back in '98 or '99 with Oakland may be a far different hitter than the one who steps up to the plate today. It does us little good to aggregate plate appearances over multiple seasons if the hitter changed a great deal from one year to the next (as in coping with a lingering injury), and we simply don't have the data available for adjustments on an individual player from year to year. (We approximate that when we apply park differentials to adjust for a player changing teams, but we should keep in mind that other factors may (and likely will) dwarf the effects of the team change.)

That's a lot of analysis, a fair amount of which may be half-baked, in an attempt to explain why I believe fans should avoid trying to analyze players definitively based upon a handful of at bats. It can be an entertaining exercise, but one should always keep in mind that it's only for fun, and one shouldn't be advocating serious decisions based upon such small samples.

[Old#5fan]

OK, but lets take another step.... you are the manager of a AA minor league team and you have these same two kids on your team.... you observe the same information from both that you did in little league, one is a very good hitter, and one is a rather poor hitter. Their stats obviously confirm what you already knew. In your view, if just looking at their stats, do you think that you wouldn't be able to determine which player is more likely to succeed at the MLB level than the other? Or would it be a crapshoot?

In that case probably so, but again, with such a wide disparity in ability a scout would easily observe it. What about if the stats were basically the same as there a lot of players with similiar to virtually identical stats? At that point the stats are not much use at all and a scout becomes crucial as to which player might have potential and which might not. So virtually under any situation other than fantasy league baseball a good scout is more important than mere stats.

[Moose Milligan]

In that case probably so, but again, with such a wide disparity in ability a scout would easily observe it. What about if the stats were basically the same as there a lot of players with similiar to virtually identical stats? At that point the stats are not much use at all and a scout becomes crucial as to which player might have potential and which might not. So virtually under any situation other than fantasy league baseball as scout is more important than mere stats.

Funny...Eli Eon goes away and Old#5Fan appears...espousing the same beliefs and typing in the same style.

Amazing.

[Old#5fan]

Despite my earlier defense of statistics and their value, I personally put experienced observation ahead of stats sometimes, too. There are some very, very intangible elements to the game of baseball and that's part of the appeal to me. Gut feelings and hunches apply more here than in any other sport in my opinion.

I applaud you for the above sentence I snipped from your post. I took it as a roundabout concession to agree to disagree, and the ability to do so and to see things from someone else's side of the table even if it's presented in an abrasive fashion is what can make a message board interesting, entertaining, and enjoyable.

Your take and mine are not completely far apart just at different stages on the scale of stats versus experienced observation. Your take is correct and you can "hold the applause.":)

[Old#5fan]

I don't know if they are more or less trustworthy than the predictions of the best evaluators of talent. They are more reliable in the sense that stats will tell you the same thing, again and again, while different evaluators of talent will vary in what they tell you. For the purposes used here, using stats to predict future performance of an individual is not nearly as strong as some people would have you believe. Say it's accurate 68% of the time. That's better than random, but it's still far from something you'd bank on in any given case. Over a large number of cases, you could make a fortune betting that way, but that doesn't really tell you whether or not sign this guy or that guy.

I get what you are saying. Its not quite a crapshoot but it is far from a viable predictor on an individual basis. I still think scouts are more important because they are the means to separate players with virtually identical stats as to which may have potential and which may not.

[Witchy Chick]

FWIW – Jeff Conine (who many “predicted” would fall off a cliff 2 years ago when he was a 40-year-old Oriole) has put up a .717 OPS (in limited ABs with Philly last year) and a .727 OPS (also limited ABs, combined with Cincy/Mets) this year. Not world-beater status, but not “falling off the cliff” either. It's certainly better than Payton's .660. :p

Without being as vehement and verbose ;) as Old#5, I see the value in using stats as “predictors” of likely outcomes. However, as someone stated, there are anomalies.

This is more a philosophical question than statistical/analytical: How many times does an anomaly have to buck the “predicted” trend before it’s no longer an anomaly?

If Trax has 3 or 4 "un-Trax-like" seasons in a row -- are those trend-bucking seasons still anomalies? Or do they become the norm? :confused:



I'd just like to add that this is a great discussion, and I am learning lots. :D Makes me want to go back to college and study statistical analysis.


Witchy

[Fan4Life]

I think that you're heading in the wrong direction if you attempt to slice and dice statistics from a relatively small sample size into even smaller samples.

Thus, a sample of at bats by a hitter is a collection of pseudo random event outcomes which are pre-ordained, but can be treated as a collection of true random events if one has a sufficiently large sample size. If you take that large sample and slice it up into smaller samples for the purpose of subjecting it to detailed analysis, you're largely defeating your purpose and any conclusions which you draw from the smaller sample sizes are likely to be quite invalid.



I wasn't trying to use 40 atbats as a sample size as much as pointing out that hitters like the one you described that can be in a funk for awhile and then in a zone for awhile can produce simliar total stats at the end of the sample size, say one season, as a guy who doesn't have quite as many peaks and valleys and when he has them, they last for lesser periods of time, but at the end of the sample size when seeing how both of them appear to be similar, one of them has stats that would more likely reflect future production and one has stats that incompass a variety of peaks and valleys which would be more difficult to use in projecting future production IMO.

I'm learning a lot too and am enjoying the brainshare going on here..

[Sports Guy]

FWIW – Jeff Conine (who many “predicted” would fall off a cliff 2 years ago when he was a 40-year-old Oriole) has put up a .717 OPS (in limited ABs with Philly last year) and a .727 OPS (also limited ABs, combined with Cincy/Mets) this year. Not world-beater status, but not “falling off the cliff” either. It's certainly better than Payton's .660. :p

Without being as vehement and verbose ;) as Old#5, I see the value in using stats as “predictors” of likely outcomes. However, as someone stated, there are anomalies.

This is more a philosophical question than statistical/analytical: How many times does an anomaly have to buck the “predicted” trend before it’s no longer an anomaly?

If Trax has 3 or 4 "un-Trax-like" seasons in a row -- are those trend-bucking seasons still anomalies? Or do they become the norm? :confused:



I'd just like to add that this is a great discussion, and I am learning lots. :D Makes me want to go back to college and study statistical analysis.


Witchy
Yes there are.....But would you build a team or acquire a player hoping for them to become an anomaly or would you go the safer route, look at the stats that best predict a players likely outcome and sign those players accordingly?

[Flosman]

The boded part is wrong.

That the r does not equal 1 does not mean the system is flawed, or that "Applying the proper weight to the various factors is something that is still more art than science"

If you had a perfect projection system (i.e an omniscient* "all knowing" being) and knew exactly how good a player was, your r, or the correlation between your projection and the player's results over 500 or 600 or whatever plate appearances is still hampered by sample size and the expected distribution over that sample size.

To work with the coin example, we know (i.e. are "omniscient") the coin is 50/50, but we also know that over x trials, we don't expect the results to be 50/50. That over x trials the "r" is less than 1 does not mean the coin is not 50/50.

There is no suggestion that we can know the future. My point is we have projection systems that rely on what a player has done that come very close to being as perfect as they can be given we can't and don't know the future. This means your contention that "Applying the proper weight to the various factors is something that is still more art than science" is wrong.

The issue is not weighing factors differently because the projection models do not accurately model reality. The primary issue is our sample size is not large enough. The models aren't prefect, but they are pretty damn close. And they are very reliable.


* Here is how I used first used omniscient in this thread:



Why do you use the word differently - "ominiscience (aka after-the-fact inspection of reality)" - to suggest it means one who knows the future, when its clear I am not using it that sense?

This as been an interesting discussion. Baltimoron deserves credit for his knowledge of "baseball Stats" and I would hire Shack to try to tackle a complex modeling problem if the need ever arose. One issue I have with the "baseball stat" guys is they seem to take offense to what Shack is saying. He is not saying anything that remotely means what Eli or Old fan are saying. He is not saying that projecting is a crapshoot. He is saying in an individual case there currently is no method to get a reliable accurate projection to a degree that you would want to bet the farm on using only baseball stats. Shack is proven to be correct in his assessment.

Another issue with Baltimoron's logic is you keep referring to a coin flip as a event that is an example. It is not a similar system, it is a known 50-50 event any variation is a product entirely of the sample size. given enough data it will prove to be the case all of the time. Human proformance in a baseball setting is a much more complicated system and each player is different and changing individually over time. The coin is a static system and really the sample size is the only reason you will have a variable result. each player is a different system unto itself and the variation from the projection and the actual results can not be explained by sample size. Sample size can be a factor but Shack is correct in that most of the variation is due to things we can not at this time, Model, measure or understand.

[Migrant Redbird]

I wasn't trying to use 40 atbats as a sample size as much as pointing out that hitters like the one you described that can be in a funk for awhile and then in a zone for awhile can produce simliar total stats at the end of the sample size, say one season, as a guy who doesn't have quite as many peaks and valleys and when he has them, they last for lesser periods of time, but at the end of the sample size when seeing how both of them appear to be similar, one of them has stats that would more likely reflect future production and one has stats that incompass a variety of peaks and valleys which would be more difficult to use in projecting future production IMO.

You need to consider what you're analyzing is really meaningful. I tend to focus in upon the largest samples available if I'm trying to evaluate what a player is likely to produce in the future. Even if one player does appear to be steadier (fewer peaks and valleys) than another, I'm not sure that disparity would stand up in the future -- which is the only thing that matters.

Nor am I sure that it really makes a difference. If there were some way to assure that a player's peaks came during the post season and in critical series in the regular season and his slumps occurred only when the team was either way ahead or way behind, then it might make a difference. Since that can't be assured, what I'm most interested in is total production, not how it's distributed. I read a lot of posts disparaging players because their home runs tend to be solo blasts or occur when the additional runs are meaningless to the outcome of a game, and I tend to discount those posts. My own opinion is that it all evens out over the long run, even if the timing gets us very frustrated in the short run. You can look at pat performance and reach conclusions that slumps by key players at critical times caused early exits from the post season, but it doesn't tell you anything useful for the next time your team is playing in October.


I'm learning a lot too and am enjoying the brainshare going on here..

I've enjoyed all the discussions, especially the posts of rshackleford, one of which I borrowed for a discussion on a Cardinals forum about whether the Cards should attempt to bring Maroth back next season. Maroth hit hard in rehab start (http://www.stlouissportsforum.com/forums/showthread.php?t=2038). My own feeling is that the Cards rotation will still be painfully thin in 2008 and there aren't likely to be any remedies available on the free agent market.

[Witchy Chick]

Yes there are.....But would you build a team or acquire a player hoping for them to become an anomaly or would you go the safer route, look at the stats that best predict a players likely outcome and sign those players accordingly?

I don’t think it’s completely detrimental to have one or two “buck-the-trend” possibilities on your team. I wouldn't "build a team" with a bunch of those guys, though. Depending on what we potentially do this off-season, I would be OK with having Trax back next year. I think the possibility of going with a Bedard, Guthrie, DCab, Loewen, Olson/Penn rotation next season is a scary one. Bedard is a proven factor at this point, Guts is still a question mark, DCab is unreliable, Loewen is coming back from a major injury, and Penn/Olson are unproven kids. One injury (to Bedard) and we’re screwed. Trax might be a stabilizing factor – and might continue to buck the trends. (Before his injury was revealed, I felt the same way about Kris Benson being here this year.)

BTW – how does Richie Sexson fit into this discussion? :cool: He’s due to make WAY more than Trax’ $4.5 million option in 2008 and he is most definitely trending downward and getting older.

2005: 39 HR, 121 RBI, .917 OPS, 140 OPS+
2006: 34 HR, 107 RBI, .842 OPS, 120 OPS+
2007: 21 HR, 63 RBI, .700 OPS, 88 OPS+

A prospect we might potentially acquire along with Sexson is worth $10 million in salary (Sexson vice Trax)?


Witchy

[Sports Guy]

. I think the possibility of going with a Bedard, Guthrie, DCab, Loewen, Olson/Penn rotation next season is a scary one.Less scary than Trax in there IMO.

BTW – how does Richie Sexson fit into this discussion? :cool: He’s due to make WAY more than Trax’ $4.5 million option in 2008 and he is most definitely trending downward and getting older.

2005: 39 HR, 121 RBI, .917 OPS, 140 OPS+
2006: 34 HR, 107 RBI, .842 OPS, 120 OPS+
2007: 21 HR, 63 RBI, .700 OPS, 88 OPS+

A prospect we might potentially acquire along with Sexson is worth $10 million in salary (Sexson vice Trax)?

I haven't really talked about a straight up deal for Sexson though.

[RShack]

One issue I have with the "baseball stat" guys is they seem to take offense to what Shack is saying. He is not saying anything that remotely means what Eli or Old fan are saying. He is not saying that projecting is a crapshoot. He is saying in an individual case there currently is no method to get a reliable accurate projection to a degree that you would want to bet the farm on using only baseball stats. Shack is proven to be correct in his assessment.

Another issue with Baltimoron's logic is you keep referring to a coin flip as a event that is an example. It is not a similar system, it is a known 50-50 event any variation is a product entirely of the sample size. given enough data it will prove to be the case all of the time. Human proformance in a baseball setting is a much more complicated system and each player is different and changing individually over time. The coin is a static system and really the sample size is the only reason you will have a variable result. each player is a different system unto itself and the variation from the projection and the actual results can not be explained by sample size. Sample size can be a factor but Shack is correct in that most of the variation is due to things we can not at this time, Model, measure or understand.
Thank you, kind sir. You get it.

I can't believe anybody actually read all that stuff ;-)

There's a woman who I worked with at Stanford named Daphne Koller who is doing some amazing things. The over-simplified story is that she's basically using stats to generate useful hypothesis to test, i.e., she's not using stats to answer questions, she's using them to have computation figure out the right questions to ask. When I was there, she was using software systems to come up with scarily well-focused questions along the lines of "Does this piece of DNA control this specific trait?" Once the right question was asked, it was fairly straightforward to find an answer. The hard part was finding the right question to ask. The impressive thing was that her systems effectively ID'd what certain aspects of DNA really do, by asking the right question which others could then answer. I would love it if she suddenly became a baseball freak. However, I think she's kinda busy on other things at the moment.

I think that's the future value of stats: using them to help us ask the right questions. One place where stats have the longest history of use is in psychology research. Over a 100 years, that research has generated tons of research conclusions about things that don't matter at all. It's because prof's come up with hypotheses to test that, once they are tested, don't answer any important questions. In short, most psych research asks really dumb questions, but it can get published. Once we shift our heads around so that we use stats to help us know what questions to ask, then I think human knowledge will explode even more.

To me, it's the difference between "research" and "re-search". What most people do with baseball stats is "re-search". That's understandable, simply because the Daphne Koller's of the world are rare and special. But that's not what's gonna teach us highly useful things. We need to just face the fact that using stats to learn important things is a very hard problem. It's nobody's fault, it's just the nature of the beast. Along the way, we should at least try to be mindful of what re-search can and cannot tell us. Around here, that point gets lost routinely. That's all I was really trying to say.

[RShack]

This is more a philosophical question than statistical/analytical: How many times does an anomaly have to buck the “predicted” trend before it’s no longer an anomaly?
No, that is not just a philosophical question. To the contrary, it is a very important question. It's precisely the kind of issue that I wish stats guys around here were more focused on.

It comes up alot around here. It came up with the whole "let's play Luis at SS everyday" debacle. While I agree with the stats guys that it's dumb to think he's gonna do better in the bigs than he did in the MiL, it does beg the question, "Has anybody noteworthy ever done just that: been a better ML'er than he was a Mil'er?" But nobody seems to care. Same thing with age. It always happens that stats guys say that so-and-so is toast because of his birthday, yet we know that anomalies about that happen all the time. (My hunch is that the answer to the age thing lives in places that available stats can't see, which exposes the flaw in the myth that "stats EXACTLY model what happens". They certainly don't do that; they only model the specific subset of phenomena that they represent. )

Anomalies are important for a few reasons. One of the most important reasons is that they tell us what we do NOT know or understand. They should not be dismissed as freak-show trivia that almost never happen. The simple fact is that we know anomolies happen alot, and we do not understand why. They are what we are guaranteed to miss when using stats as predictors. Therefore, anomalies belong at the top of 2 lists:

1. Things we don't understand, and therefore should humble-up about (rather than ignore);
2. Things that smart stat people should focus on. (And maybe many smart stats people are, I'm not saying they aren't.)

Maybe you should go back to school and study this stuff ;-)

[Sports Guy]

No, that is not just a philosophical question. To the contrary, it is a very important question. It's precisely the kind of issue that I wish stats guys around here were more focused on.

It comes up alot around here. It came up with the whole "let's play Luis at SS everyday" debacle. While I agree with the stats guys that it's dumb to think he's gonna do better in the bigs than he did in the MiL, it does beg the question, "Has anybody noteworthy ever done just that: been a better ML'er than he was a Mil'er?" But nobody seems to care. Same thing with age. It always happens that stats guys say that so-and-so is toast because of his birthday, yet we know that anomalies about that happen all the time. (My hunch is that the answer to the age thing lives in places that available stats can't see, which exposes the flaw in the myth that "stats EXACTLY model what happens". They certainly don't do that; they only model the specific subset of phenomena that they represent. )

Anomalies are important for a few reasons. One of the most important reasons is that they tell us what we do NOT know or understand. They should not be dismissed as freak-show trivia that almost never happen. The simple fact is that we know anomolies happen alot, and we do not understand why. They are what we are guaranteed to miss when using stats as predictors. Therefore, anomalies belong at the top of 2 lists:

1. Things we don't understand, and therefore should humble-up about (rather than ignore);
2. Things that smart stat people should focus on. (And maybe many smart stats people are, I'm not saying they aren't.)

Maybe you should go back to school and study this stuff ;-)

Nobody cares because you don't hand an important spot on the team to a guy that has a 2% chance of making it.

Yea, maybe he bucks a trend but why gamble on that when you can get more "sure things"??

We are a pathetic team and franchise...Whether one wants to admit that or not, it is true.

And on pathetic teams, we should be looking for the best possible players for now and in the future to help us get to where we want to be.

This really shouldn't be difficult to figure out.

You say stat guys ignore this but it isn't true at all...Stat guys just prefer to go with the better odds and the better choice.

[osfan83]

Wow, there are some very bright people that post on this board! But I think this discussion is over-thinking the issue. I don't think anyone believes past statistics can accuratly predict what will exactly happen in the future. It's a case of odds. Sure ST may buck the trend next year, but that is exactly what he would be doing, BUCKING A TREND. Spend the league minimum in hopes of a player deviating from the norm? OK? Pay a player $5 in hopes he beats fairly large odds? Not a good plan.

[RShack]

I think this discussion is over-thinking the issue.
Depends on what you find interesting. No right or wrong to it, just a matter of what you find interesting.



I don't think anyone believes past statistics can accuratly predict what will exactly happen in the future. It's a case of odds.
That's the standard claim, but that's not how such things are often treated around here. Some folks routinely act as if cookbook stats predict stuff a lot more reliably and accurately than they do. And if/when you question that, some folks start getting all snippy, taking pot shots, and acting as if you're anti-stat or something. Not everybody, but some folks do.

[Migrant Redbird]

Nobody cares because you don't hand an important spot on the team to a guy that has a 2% chance of making it.

I don't know. You've got guys wanting to preserve a 2nd round draft pick (& some cash) rather than expend it upon a proven major league closer. I'm not trying to argue that going after Cordero would be a good idea (I doubt if it would), but that 2nd round draft pick is probably less than a 2% chance of becoming more than a marginal major league player. Even if your scouts have accurately assessed his potential (difficult given his amateur competition level and the aluminum bats), injuries or unpredictable development are still likely to thwart his professional career.

We gamble on 2% (and longer chances) all the time.


We are a pathetic team and franchise...Whether one wants to admit that or not, it is true.

There are worse, and the Marlins have demonstrated twice that, whether through great management or great luck, it's possible to turn from being a pathetic franchise into a world champion.


You say stat guys ignore this but it isn't true at all...Stat guys just prefer to go with the better odds and the better choice.

I doubt if most of the "stat guys" actually understand the "better odds" they claim to prefer. I don't. I realize that I don't have sufficient data to develop accurate probabilities for sports events, and that all the "odds" I throw out there are little more than WAGs. I don't think that any of the "stat guys" do that much better.

I can say with some degree of confidence that Albert Pujols is more likely to hit a game winning home run in a given situation than Steve Trachsel is, but I'd be very hard pressed to quantify the edge that Albert has. Twice? Ten time? A hundred times? It's probably in there somewhere, and it's large enough that I'd pinch hit Albert for Steve if the situation arose, but don't ask me to put any numbers on it.

[Sports Guy]

I don't know. You've got guys wanting to preserve a 2nd round draft pick (& some cash) rather than expend it upon a proven major league closer. I'm not trying to argue that going after Cordero would be a good idea (I doubt if it would), but that 2nd round draft pick is probably less than a 2% chance of becoming more than a marginal major league player. Even if your scouts have accurately assessed his potential (difficult given his amateur competition level and the aluminum bats), injuries or unpredictable development are still likely to thwart his professional career.

We gamble on 2% (and longer chances) all the time.These situations aren't even remotely the same.




There are worse, and the Marlins have demonstrated twice that, whether through great management or great luck, it's possible to turn from being a pathetic franchise into a world champion.
Not when you take into account rescources and ballpark.


I doubt if most of the "stat guys" actually understand the "better odds" they claim to prefer. I don't. I realize that I don't have sufficient data to develop accurate probabilities for sports events, and that all the "odds" I throw out there are little more than WAGs. I don't think that any of the "stat guys" do that much better.Player A has a 600 MiL OPS....Player B has a 900 MiL OPS...They are the same age and have been at the same levels every year...Who has a better chance at success?

[Fan4Life]

These situations aren't even remotely the same.



Not when you take into account rescources and ballpark.

Player A has a 600 MiL OPS....Player B has a 900 MiL OPS...They are the same age and have been at the same levels every year...Who has a better chance at success?

If I understand what's been said, the 900 OPS (Player B) guy has the better chance of success, but the remaning question that cannot be answered is, how much greater are those odds that player B has at success than player A? If we took 50 Player A's and 50 player B's, what percent of Player B's would actually have greater success? We don't know... We only know it is more likey for Player B's to succeed based on what we have measured.

[Sports Guy]

If I understand what's been said, the 900 OPS (Player B) guy has the better chance of success, but the remaning question that cannot be answered is, how much greater are those odds that player B has at success than player A? If we took 50 Player A's and 50 player B's, what percent of Player B's would actually have greater success? We don't know... We only know it is more likey for Player B's to succeed based on what we have measured.

Isn't that all you need to know?

[RShack]

If I understand what's been said, the 900 OPS (Player B) guy has the better chance of success, but the remaning question that cannot be answered is, how much greater are those odds that player B has at success than player A? If we took 50 Player A's and 50 player B's, what percent of Player B's would actually have greater success? We don't know... We only know it is more likey for Player B's to succeed based on what we have measured.
I think you are asking the right questions.

As for whether or not we can or cannot say what percentage of Player B's would have greater success than Player A's, I'm not sure whether or not that can be answered. It wouldn't be surprised if it could be answered... for a largish group of Player B's anyway. It certainly seems plausible that it could, and it's certainly the right kind of question to ask. If it can, we don't read that around here.

Your basic question, "If the odds are better, well, how much better are they?", is a great question and is exactly the kind of stuff I wish we heard more about. There is a tendency for some to just say, "The odds are better, and that's all we need to know, end of discussion", but that's just wrong. Unless every other single factor is equal, it's not all we need to know, and thus it should not be the end of discussion. Even a pure Vegas game player who cares nothing about baseball will want to know more than which side the odds favor, he'll want to know by how much.

Statistical projections are inherently about probabilities. That's all they are, nothing more, nothing less. Therefore, it certainly makes sense to want to know more about the probabilities than we get when somebody just makes up a number like 2% or 7% (or whatever) in the middle of an argument and then runs with it.

[Fan4Life]

I think you are asking the right questions.

As for whether or not we can or cannot say what percentage of Player B's would have greater success than Player A's, I'm not sure whether or not that can be answered. It wouldn't be surprised if it could be answered... for a largish group of Player B's anyway. It certainly seems plausible that it could, and it's certainly the right kind of question to ask. If it can, we don't read that around here.

Your basic question, "If the odds are better, well, how much better are they?", is a great question and is exactly the kind of stuff I wish we heard more about. There is a tendency for some to just say, "The odds are better, and that's all we need to know, end of discussion", but that's just wrong. Unless every other single factor is equal, it's not all we need to know, and thus it should not be the end of discussion. Even a pure Vegas game player who cares nothing about baseball will want to know more than which side the odds favor, he'll want to know by how much.

Statistical projections are inherently about probabilities. That's all they are, nothing more, nothing less. Therefore, it certainly makes sense to want to know more about the probabilities that somebody just making up a number like 2% in the middle of an argument, then running with it.

I present exhibit A....



We only know it is more likey for Player B's to succeed based on what we have measured.

Isn't that all you need to know?