PAP

Posted by David Wade on Thursday, July 9, 2009 at 5:59 pm

In my last article, I wondered aloud if Dusty Baker risked Aaron Harang’s future effectiveness (not to mention injury) by sending him out for one batter after a long rain delay so the big guy could qualify for a win.  That question prompted me to look over Harang’s 2009 gamelog and to revisit some articles critical of Baker that surfaced last year, when Harang suffered a decline in performance after pitching several innings on short rest in a marathon game early in 2008.  There was also one other thing I knew going in that I would have to check out, and that is the Pitcher Abuse Points for Harang under Baker’s turn as manager of the Reds.  I mentioned PAP in the last article, but I’m going to summarize it completely in this one with regards to high pitch counts’ effect on a pitcher’s future performance.

As this article explains in detail over at Baseball Prospectus, Rany Jazayerli came up with PAP as a way to measure a pitcher’s workload. The idea was to have a tool that could serve as an indicator for a pitcher being overworked and therefore more likely to suffer injury or an ensuing decline in performance. Keith Woolner wrote the article linked above and provided a sample study in hopes he could statistically demonstrate a positive correlation between high pitch outings and subsequently diminished effectiveness.

One thing Woolner tweaked after his initial findings was to focus on pitchers that he considered ‘high endurance’, to try to keep from skewing the results. That more specific data set consisted of guys who threw more than league average pitch counts per outing. He figured by focusing on them, he could exclude below average pitchers that were allowed a long outing or two here or there and instead look at above average pitchers who throw higher pitch counts (as above average pitchers are more accustomed to doing).  Basically, the hope was that he could filter out guys who might stink because they were pushed too hard because they went deep into one good game.

Woolner concluded that there was a decline in performance for the pitchers in his study over a three-week period.  However, that decline was fairly slight (3-8%) and was just as bad after a 90 pitch outing as it was a 130 pitch outing.  Where he found a more convincing and steep fall off was in outings around 140 pitches or more.  This led him to modify PAP so that by cubing the original PAP point system, the score for high pitch outings rises significantly once reaching and surpassing the 130 pitch marker.  This enabled him to match more closely the findings in his sample, and left him with the following formula-

 PAP= 0 for pitch outings less than 100 pitches.

(#of pitches – 100)^3

By cubing the number of pitches above 100, Woolner was able to match what he found in his sample- a slight decline in subsequent performance that increases quite a bit for three weeks following an outing that goes 30-40 pitches over the baseline of 100.    While even the highest decline is less than 10%, that means a heck of a lot in Major League Baseball.  As Woolner points out, you’re talking about a few games in the standings and that could be the difference in making the playoffs.  This measuring of abuse is what we’ll eventually look at as BP has rankings by pitcher for all seasons since its invention.

Woolner offers some caveats to those ready to harangue Baker. 

(Baker has had some questionable PAP numbers on his pitchers in the past- to be examined in a later installment of this series, which is looking right now as it may rival- in length- if certainly not quality- Victor Hugo’s original 18 installments of “The Count of Monte Cristo” before I come to any reasonable conclusion.)

First, we haven’t even looked at injuries yet.  That’s coming up next.  Secondly, Woolner explains that his data set encompasses all sorts of successful pitchers possessing varying body types and pitching styles.  He also stresses that while he believes this study shows a chance for decline in starts following long appearances, the time of year certainly plays a role.  I.E., if you’re in the playoffs, all bets may be off, as there won’t be a 21 day period following for your hypothetical pitcher.

As I wrote earlier, PAP and injuries are the next subject and it will focus on another article from Woolner, plus some criticism of his findings from other sources.  I also had the pleasure of sitting down with an assistant coach for the University of Kentucky’s baseball team this week and will use his opinions and thoughts (which were most appreciated) to try and present at least one example of how someone who has worked with high level pitchers feels about protecting those valuable arms.