PDA

View Full Version : Pitcher Abuse Points - Predictive Validity?



WHIP
05-23-05, 05:21 PM
Of much interest is, obviously, the idea of pitch-counts. Of even more interest is knowing at what point pitchers can hurt themselves through over-use. I understand that BP has a little formula (or simple arithmetic) to decide on just how 'abused' pitchers may actually be.

Now the obvious strength of any theory is predictive validity. In short, how useful are the abuse points for predicting a pitcher's injuries? I know that Mark Prior and Kerry Wood almost always top the list (lending credence to the formula), but they really aren't that interesting to me because they're under Butcher Baker. I'm more interested in hearing about other pitchers that may have climbed the abuse point list one year and, the next, suffered some arm trouble. And if that is indeed the case, what percentage of pitchers near the top of the list do indeed suffer injuries? Or at what point in the PAP hierarchy does it make sense to say that a pitcher's abuse points are 'normal' - that is, what would be about 'league average' and, by implication, 'safe'? In other words, I'd like some analysis and criticism of BP's formula.

I don't subscribe to BP, so I don't think I could find this stuff out on my own. Thanks y'all.

Prickly Pete
05-23-05, 06:07 PM
Interesting question. There have been studies done on the correlation of workloads to injuries with pitchers, but the problem with those studies always seems to be the same thing: Pitching is such an unnatural motion that many, many pitchers get hurt even with a very light workload. So it's almost impossible to develop a direct link.

Plus, reduced effectiveness, rather than injury, is also a possible side effect of overwork.

As for BP's "Pitcher Abuse Points," it's useful but not terribly sophisticated. It mainly counts pitches and the number of starts a pitcher throws over a certain threshold. Starts are classified as Category 1 for throwing less than 100 pitches and Category II for throwing 101-109 pitches and so on. The more starts you have in the higher categories, the more "abused" you're considered.

FWIW, here are the top 5 most-abused pitchers, according to BP, for the last two seasons:

2004
Livan Hernandez
Jason Schmidt
Carlos Zambrano
Victor Zambrano
Al Leiter

2003
Javier Vazquez
Kerry Wood
Livan Hernandez
Mark Prior
Mark Redman

Prickly Pete
05-23-05, 06:13 PM
This might be of interest to you as well.


PAP was introduced by Rany Jazayerli in the summer of 1998 on the Baseball Prospectus website. At least in print or online there was no study made to discuss its validity at predicting injuries or whether it was better than simple pitch counts at measuring pitcher overuse. It was hailed by many people as a good metric, but I personally was disappointed that no study appeared in the 1999 book and again in the 2000 book backing up the claims its authors and others had made for it.

In the 2001 book, Rany essentially retracted it as a method and admitted there was no evidence that PAP told us anything other than a pitcher's PAP


Pitcher Abuse Points (http://www.baseball-reference.com/otb/pitcher_usage_old.php)

Prickly Pete
05-23-05, 06:19 PM
And this.....

BP Explains Pitcher Abuse Points (http://www.baseballprospectus.com/article.php?articleid=2633)

WHIP
05-23-05, 07:05 PM
Thanks for the links!

Prickly Pete
05-23-05, 07:38 PM
Thanks for the links!
No problem. I hope you find them useful, although I feel like you asked a decent question and I gave you a reading assignment... :doh:

pedromartinezfan
05-23-05, 07:56 PM
Here is a study I did on this stat a while ago.

2003 Pitcher Abuse Points Leaders
1. Javier Vazquez
2003: 230.7 IP, 154 ERA+
2004: 198.0 IP, 92 ERA+

2. Kerry Wood
2003: 211.0 IP, 133 ERA+
2004: 140.3 IP, 122 ERA+

3. Mark Prior
2003: 211.3 IP, 175 ERA+
2004: 118.7 IP, 118 ERA+

4. Mark Redman
2003: 190.7 IP, 112 ERA+
2004: 191.0 IP, 99 ERA+

5. Livan Hernandez
2003: 233.3 IP, 155 ERA+
2004: 255.5 IP, 115 ERA+

6. Woody Williams
2003: 220.7 IP, 107 ERA+
2004: 189.7 IP, 100 ERA+

7. Al Leiter
2003: 180.7 IP, 106 ERA+
2004: 173.7 IP, 133 ERA+

8. Russ Ortiz
2003: 212.3 IP, 109 ERA+
2004: 204.7 IP, 104 ERA+

9. Jason Schmidt
2003: 207.7 IP, 182 ERA+
2004: 225.0 IP, 139 ERA+

10. Carlos Zambrano
2003: 214.0 IP, 136 ERA+
2004: 209.7 IP, 165 ERA+

Zambrano pitched better, but that is about it. Everyone pitched alot less or a lot worse. Leiter may be the exception.

11. Joel Pineiro
2003: 211.7 IP, 117 ERA+
2004: 140.7 IP, 92 ERA+

WHIP
05-23-05, 08:34 PM
No problem. I hope you find them useful, although I feel like you asked a decent question and I gave you a reading assignment... :doh:

You can't expect to ask a question in the sabermetrics forum and not receive any literature from someone's trunk.

WHIP
05-23-05, 08:37 PM
Here is a study I did on this stat a while ago.

2003 Pitcher Abuse Points Leaders

Thanks for the numbers...if you know of any more comprehensive studies I'd like to hear about 'em. Like at what point, by and large, does the accumulation of abuse points generate injuries. There probably aren't enough data for that kind of study though.

Prickly Pete
05-23-05, 08:58 PM
Thanks for the numbers...if you know of any more comprehensive studies I'd like to hear about 'em. Like at what point, by and large, does the accumulation of abuse points generate injuries. There probably aren't enough data for that kind of study though.
There's enough data; there's just not a direct link. You really can't look at pitch counts (which is all abuse points measures) and conclude that a certain number of pitches is more likely to lead to injury.

The main conclusion of some of the pitch-count studies is that a pitcher's injury risk goes up very quickly with each pitch he throws after reaching some fatigue point. The problem is that those fatigue points surely differ for different pitchers.

BP's Pitcher Abuse Points system can't -- and doesn't try to -- account for that. It simply assumes that 100 pitches is the beginning of the fatigue point for all pitchers, and pitchers begin accumulating "abuse points" on an accelerating scale once they go above 100.

There's no way that can have a clear and direct relationship to future injury. There are just too many other variables, chief among them being age and how many pitches have been thrown in years prior, especially at a young age.

It's a fascinating topic, though.

WHIP
05-23-05, 09:17 PM
There's enough data; there's just not a direct link. You really can't look at pitch counts (which is all abuse points measures) and conclude that a certain number of pitches is more likely to lead to injury.

The main conclusion of some of the pitch-count studies is that a pitcher's injury risk goes up very quickly with each pitch he throws after reaching some fatigue point. The problem is that those fatigue points surely differ for different pitchers.

BP's Pitcher Abuse Points system can't -- and doesn't try to -- account for that. It simply assumes that 100 pitches is the beginning of the fatigue point for all pitchers, and pitchers begin accumulating "abuse points" on an accelerating scale once they go above 100.

There's no way that can have a clear and direct relationship to future injury. There are just too many other variables, chief among them being age and how many pitches have been thrown in years prior, especially at a young age.

It's a fascinating topic, though.

True enough. Too many confounding variables.

rkh5donkey
05-23-05, 11:00 PM
Has anyone ever done such a study that factors in the type of pitches the player throws? One might imagine that, for example, a hotshot fastball pitcher might suffer more abuse than a knuckler.

Prickly Pete
05-24-05, 12:54 PM
Has anyone ever done such a study that factors in the type of pitches the player throws? One might imagine that, for example, a hotshot fastball pitcher might suffer more abuse than a knuckler.
I don't know if there's been a study like that specifically relating to pitch counts, but I think it's fairly commonly accepted that certain breaking pitches -- curveballs and sliders -- are the most harmful, especially if thrown at a young age.

Some organizations are not allowing their top prospects to throw sliders at all. I think the Yankees are doing that with Philip Hughes, and I know the Mariners are not letting Felix Hernandez, considered the best pitching prospect in all of baseball, throw his slider.