How Many Stats Guru's Predicted That....

[davearm]

Okay. Let me try to explain. I was originally taken to task here by many stats people for stating before the season began that I thought Luke Scott would hit 12-14 homers at the maximum this season. I got a raft of crap from a whole bunch of people spouting, but, but, BUT he hit 18 in 359 at bats last year! To that I replied that was meaningless to me as he is changing teams and leagues and what he did last year is pretty much non-applicable as a predictor of this season. To that I got a bigger raft of crap about I was ignoring all of his minor league at bats, blah, blah, blah, and the fact he was in a less homer friendly park last season, blah, blah, blah.

My point is this, I did use his past or prior year stats to make a prediction but only in a very general sense. I took what most "experts" here were proclaiming for his homer output 24-32 range and cut it in half because of the change of teams and leagues. To this, I was laughed at and derided. Yet, my guess is just as good as anyone else's and I have just as much a chance as being right as anyone. So far, I am actually looking pretty good on my guess. The point I am trying to make here is there is no absolute right or wrong way to make an "educated" guess in predicting stats. People just need to concede this point and I am finished here with this thread.

There are better ways and worse ways, though.

So while your point is taken that there is no absolute right and wrong prediction method, your point that "my guess is just as good as anyone else's and I have just as much a chance as being right as anyone" is patently incorrect.

You may wind up being right, and the stat gurus may wind up being wrong, but the odds are against it.

[Spoonless]

What's ironic is that OldFan is arguing that the past doesn't project the future when it comes to baseball, but it sure as hell does with him.

Except that in order to come up with the theory that Scott wouldn't hit as many due to adjusting to the league, he'd have to have prior knowledge that players sometimes have issues adjusting, which is based on past performance trends of players who have switched before. OldFan is using past performance in his predictions, he's just doing it mentally instead of using information written down.

[Old#5fan]

Who in their right mind said there's an absolute way of predicting future performance??? There's several ways including wildly guessing, basing predictions on what you've seen on TV, or statistics regarding previous performance. The latter, while clearly not absolute, is the most reliable.

I disagree. No singular method is more reliable. A combination of several methods is more reliable. For example, using your own weatherman analysis which I happen to believe was very short-sighted, where the weatherman says there is an 80% chance of rain today, and apparently a stats enamored person would automatically grab their umbrella. Myself, on the other hand would look outside first before doing anything. If I see the sun is out without a cloud in the sky, I personally will leave the umbrella at home and ignore what seems simpley to be a wrong forecast. I believe observation is equally if not more important than stats.

However, to tell you the truth, I always have a small umbrella either in my vehicle on in my bag when I travel anywhere.

[ScottieBaseball]

I disagree. No singular method is more reliable. A combination of several methods is more reliable. For example, using your own weatherman analysis which I happen to believe was very short-sighted, say the weatherman says there is an 80% chance of rain today, and apparently a stats enamored person would automatically grab their umbrella. Myself, on the other hand would look outside first before doing anything. If I see the sun is out without a cloud in the sky, I personally will leave the umbrella at home and ignore what sees to be a wrong forecast. I believe observation is equally if not more important than stats.

However, to tell you the truth, I always have a small umbrella either in my vehicle on in my bag when I travel anywhere.

Yep...someone asked earlier and the answer is 'intentionally obtuse'.

[DrungoHazewood]

The prediction may be "credible" based on what is viewed as normal methodology but that doesn't mean it will be of any value in correctly predicting the player's future stats for any particular year. In other words nobody has anyway of knowing whether Jay Payton will hit more homers than Luke Scott this season. No way, no how. They can guess at it based on past stats but that is no better a guess than anyone else who is not a stats guru can make. That is my point here. Any guess is a crap shoot. Past stats are simply not reliable enough to predict anything in the future any more reliably than a simple guess or range as to what 'might' occur. Stats experts simply need to sincerely admit this instead of acting like they have more insight to the future than the average observant fan does, and I will move on. In other words past stats don't translate into future stats automatically as some here chose to believe.

I don't know why I'm even trying, but here goes.

Nobody, and I mean nobody, with any credibility ever says that past results are a perfect indicator of future performance. Of course projections just give a range of possible outcomes. That's the whole point of a system like PECOTA.

The only folks who say they think otherwise are people like you. Those who intentionally misrepresent the argument of so-called statheads to win a strawman debate. You're arguing a point you've made up, to win a debate with yourself.

But if you believe that just watching players and taking a swag at what they'll do in the future is as accurate as a detailed analysis involving numbers and scouting reports you're completely delusional and have no connection to reality.

[ScottieBaseball]

I don't know why I'm even trying, but here goes.

Nobody, and I mean nobody, with any credibility ever says that past results are a perfect indicator of future performance. Of course projections just give a range of possible outcomes. That's the whole point of a system like PECOTA.

The only folks who say they think otherwise are people like you. Those who intentionally misrepresent the argument of so-called statheads to win a strawman debate. You're arguing a point you've made up, to win a debate with yourself.

But if you believe that just watching players and taking a swag at what they'll do in the future is as accurate as a detailed analysis involving numbers and scouting reports you're completely delusional and have no connection to reality.

Pecota...is that the new tea leaf blend your wife used this morning at breakfast? I'm excited...did the leaves indicate how many innings Daniel Cabrera is going to last tomorrow? Please answer in the form of a haiku.

[Old#5fan]

I don't know why I'm even trying, but here goes.

Nobody, and I mean nobody, with any credibility ever says that past results are a perfect indicator of future performance. Of course projections just give a range of possible outcomes. That's the whole point of a system like PECOTA.

The only folks who say they think otherwise are people like you. Those who intentionally misrepresent the argument of so-called statheads to win a strawman debate. You're arguing a point you've made up, to win a debate with yourself.

But if you believe that just watching players and taking a swag at what they'll do in the future is as accurate as a detailed analysis involving numbers and scouting reports you're completely delusional and have no connection to reality.

Isn't a scouting report considered "watching" a player?

So in effect you are agreeing with me on this in that I have already espoused that the best way to try to predict the future is a combination of things, and not just any one method?

[Flip217]

I disagree. No singular method is more reliable. A combination of several methods is more reliable. For example, using your own weatherman analysis which I happen to believe was very short-sighted, where the weatherman says there is an 80% chance of rain today, and apparently a stats enamored person would automatically grab their umbrella. Myself, on the other hand would look outside first before doing anything. If I see the sun is out without a cloud in the sky, I personally will leave the umbrella at home and ignore what seems simpley to be a wrong forecast. I believe observation is equally if not more important than stats.

However, to tell you the truth, I always have a small umbrella either in my vehicle on in my bag when I travel anywhere.

Heh heh...you're all wet!

apologies, that was ridiculous...I'm getting light-headed from this thread...

[Spoonless]

Pecota...is that the new tea leaf blend your wife used this morning at breakfast? I'm excited...did the leaves indicate how many innings Daniel Cabrera is going to last tomorrow? Please answer in the form of a haiku.

Daniel Cabrera:

Eight innings, three hits, two runs;

Two walks, four strikeouts.

[CrimsonTribe]

Pecota...is that the new tea leaf blend your wife used this morning at breakfast? I'm excited...did the leaves indicate how many innings Daniel Cabrera is going to last tomorrow? Please answer in the form of a haiku.

PECOTA is not

A reliable method

Opines Old Fan 5

[BaltimoreTerp]

Pecota...is that the new tea leaf blend your wife used this morning at breakfast? I'm excited...did the leaves indicate how many innings Daniel Cabrera is going to last tomorrow? Please answer in the form of a haiku.

Arguing with him

Is as though you're banging your

Head against a wall

Or, simply put...

[ChaosLex]

Arguing with him

Is as though you're banging your

Head against a wall

Or, simply put...

I don't know about you, but I've learned a lot...

Kyle Boller sucks just because.

Hayden Penn is stupid.

Mel Kiper can't be trusted because he has bad hair.

etc.

[glenn__davis]

I don't really think

Old fan number 5 believes

Most of what he writes.

[ScottieBaseball]

Isn't a scouting report considered "watching" a player?

So in effect you are agreeing with me on this in that I have already espoused that the best way to try to predict the future is a combination of things, and not just any one method?

Of course it's a combination of things...stats being numero uno and scouting somewhere behind it. Think of it this way:

FAN: Hey, my MLB scout friend, I saw a kid pitch in a high school game yesterday and he was phenomenal! That means a TON because I've been watching baseball for over 40 years!

SCOUT: That sounds great! Maybe I'll check him out. What's his K:BB ratio?

FAN: I don't know, but I watched him and he's awesome! I think he can pitch at the big league level.

SCOUT: Really? What about splits? How's he do with righties vs. lefties?

FAN: Not a clue. He's good, though.

My point? Somebody saying "I watched him and I think he'll do well" might raise an eyebrow, but the real substance comes along in the numbers. Yes, if in that scenario said scout actually checked out the pitcher, he'd also be interested in mechanics and mound presence, but typically at higher levels, (and I know it's going to pain you to hear it said) shortcomings in either of those areas will be evident in that pitcher's statistics.

[El Gordo]

I don't really think

Old fan number 5 believes

Most of what he writes.

my so-called friends

send in my sister

to pinch-hit for me

or

useless web of numbers

still free

the sublime suchness

of a crisp line drive! :wedge: