Is Confidence In Winning A Barrier To Sucess?

[SilentJames]

But that is a generalist statement you don't need stats to predict. What these stats gurus try to claim is they can predict the win totals and none of them can nor do to any great degree of accuracy. They aren't any more accurate than Joe Fan sitting at home watching the games and making a guess. Its a joke if you ask me.

So you have a problem with ONE FACET OF STATISTICAL ANALYSIS. Yet you indict the entire field.

I don't like Glenn Beck,

Glenn Beck is a human -

Therefore ALL HUMAN BEINGS ARE HORRIBLE!

[Old#5fan]

So you have a problem with ONE FACET OF STATISTICAL ANALYSIS. Yet you indict the entire field.

I don't like Glenn Beck,

Glenn Beck is a human -

Therefore ALL HUMAN BEINGS ARE HORRIBLE!

I plead guilty to thinking all baseball stats gurus are a joke. Sorry, I probably wouldn't think that way if most every one I have ever encountered wasn't so arrogant yet blatently wrong in virtually everthing they try to predict. I don't see how anyone can be arrogant when they never get anything right.

[SilentJames]

I plead guilty to thinking all baseball stats gurus are a joke. Sorry, I probably wouldn't think that way if most every one I have ever encountered wasn't so arrogant yet blatently wrong in virtually everthing they try to predict. I don't see how anyone can be arrogant when they never get anything right.

Can you point to any specific time when these "stats gurus" were so blatantly wrong as you put it?

[BRobinsonfan]

I do get what you are saying and give you props for this post but there is still a lot more that goes into who would you rather have up with the game on the line. Your example is about as extreme you can get (Jeter versus Hernadez). Lets say though it was between Markakis and Gary Matthews Jr. GMattJr I know has hit two walk off homers as an Oriole (if not both homers at least two walkoff game winning hits. Markakis probably has at least one, maybe two. Yet say Markakis has say for example 200 close and late ABs as an Oriole in Close and Late Situations and GMattJr 45 AB's and has two game winning walkoff clutch hits. I personally would probably want G-MattJr up there and not Markakis. However, that is not etched in stone because perhaps Markakis is hitting against Riveria and he has a 300 lifetime BA against him and G-Matt 190. In that case I would probably want Markakis. There are just two many variables to be able to predict realistically what is "likely" to happen.

OldFan, I admit to using extreme examples to make the point. I think, however, its where the examples are not extreme that statistics really shine.

In your examples - I may disagree with your choices - but you're still using statistics to make your choice! You're choosing G-Matt because two game winning hits in 45 AB's vs. Nick's 2 in 200. I would say "small sample size." You would say "But he's got as many game winning hits in this situation and he's done it with fewer at bats."

We could argue who's making the best choice - but like it or not my friend - you're still using past performance (i.e. Statistics) to make your choice.

Again, I think you're hung up on this idea that we're saying we can predict everything that a player is going to do down to the 5th decimal point - when all we're doing is using past performance to choose how a player is more likely to perform.

Remember Brady Anderson's 50 HR season? Based oh his passed performance no statistical model would have anticipated that - but that doesn't invalidate statistics. That season was an aberration for him - or an "outlier."

If Brady and Frank Robinson played at the same time and you were betting a friend who was more likely to hit 50 homeruns in a year - you'd certainly put your money on Frank even after Brady's 50 homer season and even though Frank never hit more than 49 in a season - simply because the past performance of the two demonstrates clearly, irrefutably, that Frank was a much better home run hitter.

When you see Bill James or some other system "predict" performance for the 2010 season it doesn't mean that they mean Player A will actually hit exactly .294 with 18 home runs - it means that that's the mean performance of what they expect based on past performance and all the various metrics they include in their analysis. Anything within a given range on either side of those numbers is what's most likely to happen but not certain to happen.

I think if you could let go of your preconceived notions of how we all use statistics versus how you think we use them - I really do believe you'd find a more welcoming environment here.

[Sports Guy]

I plead guilty to thinking all baseball stats gurus are a joke. Sorry, I probably wouldn't think that way if most every one I have ever encountered wasn't so arrogant yet blatently wrong in virtually everthing they try to predict. I don't see how anyone can be arrogant when they never get anything right.

You prove everyday that you don't have to be a stat guru to be that way.

[BRobinsonfan]

But that is a generalist statement you don't need stats to predict. What these stats gurus try to claim is they can predict the win totals and none of them can nor do to any great degree of accuracy. They aren't any more accurate than Joe Fan sitting at home watching the games and making a guess. Its a joke if you ask me.

OldFan, I've given you a lot of grief over the years, and I'm really trying to be patient and give you the benefit of the doubt here and not think that you're simply pretending to be obstinate.

When a system predicts the Yankees will finish 98 - 64 it's not saying that it will absolutely be 98 - 64... It's simply saying that is about what you expect going into the season given the players past performance. It can't predict some key player is going to have an injury or that someone is going to be picked up in a trade mid season - it simply saying "all other things being equal, this is about what should happen. It gives you a point at which to begin the discussion about the upcoming season. Again, you are really not understanding the meaning or use of statistics at all.

[Old#5fan]

OldFan, I admit to using extreme examples to make the point. I think, however, its where the examples are not extreme that statistics really shine.

In your examples - I may disagree with your choices - but you're still using statistics to make your choice! You're choosing G-Matt because two game winning hits in 45 AB's vs. Nick's 2 in 200. I would say "small sample size." You would say "But he's got as many game winning hits in this situation and he's done it with fewer at bats."

We could argue who's making the best choice - but like it or not my friend - you're still using past performance (i.e. Statistics) to make your choice.

Again, I think you're hung up on this idea that we're saying we can predict everything that a player is going to do down to the 5th decimal point - when all we're doing is using past performance to choose how a player is more likely to perform.

Remember Brady Anderson's 50 HR season? Based oh his passed performance no statistical model would have anticipated that - but that doesn't invalidate statistics. That season was an aberration for him - or an "outlier."

If Brady and Frank Robinson played at the same time and you were betting a friend who was more likely to hit 50 homeruns in a year - you'd certainly put your money on Frank even after Brady's 50 homer season and even though Frank never hit more than 49 in a season - simply because the past performance of the two demonstrates clearly, irrefutably, that Frank was a much better home run hitter.

When you see Bill James or some other system "predict" performance for the 2010 season it doesn't mean that they mean Player A will actually hit exactly .294 with 18 home runs - it means that that's the mean performance of what they expect based on past performance and all the various metrics they include in their analysis. Anything within a given range on either side of those numbers is what's most likely to happen but not certain to happen.

I think if you could let go of your preconceived notions of how we all use statistics versus how you think we use them - I really do believe you'd find a more welcoming environment here.

I think your explanation is superb here and your last sentence is no doubt spot on in your take on how I view things. I just find that Bill James stuff to be much ado about nothing and a huge waste of time to waste one's brain cells on. I find a "mean" performance useless and cannot really see what others find value in such or a range as it is still no more than an educated guess. I mean I know Frank Robinson is superior to Vada Pinson. I don't need some sabermetric B.S. to tell me that Frank adds 4 more wins per year or will have an ops in such and such range and Pinson's will be in another such and such range. I already know all I need to know without all that baloney.

[SilentJames]

I think your explanation is superb here and your last sentence is no doubt spot on in your take on how I view things. I just find that Bill James stuff to be much ado about nothing. I find a "mean" performance useless and cannot really see what others find value in such or a range as it is still no more than an educated guess.

So you really don't believe in the scientific method of hypothesis, observe, analyze?

[Old#5fan]

So you really don't believe in the scientific method of hypothesis, observe, analyze?

I won't say I don't believe in the scientific method. In certain situations I find those things to be very valuable. I will give you an example. Say a guy is in the woods and sees a Sasquatch and is scared out of his mind. He knows what he saw and it wasn't a bear or a guy in a suit but the real deal bigfoot.

Along comes a scientist who casts the footprints, finds some hair and scat samples, analyzes all of this and the results are inconclusive. The hair matches no known DNA samples, nor the scat, the footprints are shown though that it would be nearly impossible to fake them, etc.

Now, that scientific study might have some meaning to a lot of folks including creating doubt as to what he saw, but do you think it means squat to the guy who actually saw the damn Bigfoot? He sure as hell knows what he saw!:laughlol:

(This is the only problem I have with sole reliance on a scientific method).

[Old#5fan]

OldFan, I've given you a lot of grief over the years, and I'm really trying to be patient and give you the benefit of the doubt here and not think that you're simply pretending to be obstinate.

When a system predicts the Yankees will finish 98 - 64 it's not saying that it will absolutely be 98 - 64... It's simply saying that is about what you expect going into the season given the players past performance. It can't predict some key player is going to have an injury or that someone is going to be picked up in a trade mid season - it simply saying "all other things being equal, this is about what should happen. It gives you a point at which to begin the discussion about the upcoming season. Again, you are really not understanding the meaning or use of statistics at all.

I understand this. I don't really have a problem with anyone guessing anything. I do have a problem when it is portrayed like it is fact or an absolute instead of an educated guess. I guess it must be me and my interpretation but I really don't find any great meaning or value in this type of stuff.

[BaltimoreTerp]

I won't say I don't believe in the scientific method. In certain situations I find those things to be very valuable. I will give you an example. Say a guy is in the woods and sees a Sasquatch and is scared out of his mind. He knows what he saw and it wasn't a bear or a guy in a suit but the real deal bigfoot.

Along comes a scientist who casts the footprints, finds some hair and scat samples, analyzes all of this and the results are inconclusive. The hair matches no known DNA samples, nor the scat, the footprints are shown though that it would be nearly impossible to fake them, etc.

Now, that scientific study might have some meaning to a lot of folks including creating doubt as to what he saw, but do you think it means squat to the guy who actually saw the damn Bigfoot? He sure as hell knows what he saw!:laughlol:

(This is the only problem I have with sole reliance on a scientific method).

Does that mean if I believe you are really as ignorant as you sound, that it is true no matter what you say to try and convince otherwise?

[BaltimoreTerp]

I understand this. I don't really have a problem with anyone guessing anything. I do have a problem when it is portrayed like it is fact or an absolute instead of an educated guess.

The only one doing that is you, which tells us everything we need to know.

[SilentJames]

I won't say I don't believe in the scientific method. In certain situations I find those things to be very valuable. I will give you an example. Say a guy is in the woods and sees a Sasquatch and is scared out of his mind. He knows what he saw and it wasn't a bear or a guy in a suit but the real deal bigfoot.

Along comes a scientist who casts the footprints, finds some hair and scat samples, analyzes all of this and the results are inconclusive. The hair matches no known DNA samples, nor the scat, the footprints are show that it would be possible to fake them, etc.

Now, that scientific study might have some meaning to a lot of folks including creating doubt as to what he saw, but do you think it means squat to the guy who actually saw the damn Bigfoot? He sure as hell knows what he saw!:laughlol:

I don't even know where to begin with this argument. The method of statistical analysis is sound.

Let me give you a real world example.

Carlos Pena largely came out of nowhere. He was a decent player for most of his pre-Tampa career. Then, BAM out of nowhere 1.000+ OPS. No one saw it coming or what to make of it. It was, if you will, a statistical "bigfoot".

So predictions for him were largely wrong that year, because he bucked the trend. That doesn't make the analysis wrong. Just because your hypothesis was proven wrong doesn't mean your methods were faulty - it just means that there was something you didn't account for or understand occurred.

Pena has since returned back to something more akin, but still a bit higher, than his pre-Tampa norms. The question then becomes, what happened to Carlos Pena? The same could be said for Beltre, another statistical sasquatch who vanished in Seattle. Many are guessing that Fenway will allow him to return to something closer to his big LA year, I am not one of those people.

Jsut because the results were inconclusive doesn't mean the guy saw an actual bigfoot it means that more data is needed. But if the guy is claiming he saw a bigfoot even still then no amount of data will prove him wrong, even if the guy wearing the suit comes to his door.

[ledzepp8]

So you would choose Frank because he was a better player in the past than Pinson?

Haha...well played. I'm sure he'll ignore it and go off on some other tangent.

[ledzepp8]

If you choose on a reward system it would be the kids with the higher grades. If you choose on a potential system it could be the kids with all C's as they may not be getting challenged enough, hence their lower grades.

Oh good God...