THE BOOK--Playing The Percentages In Baseball

Upon reading this, the first question that came to my mind was, what effect does mechanics have on this? Pitchers with bad mechanics could have more problems with injuries that could result in fewer pitches thrown after age 26. Over on The Pitching Mechanic web site, Roger Clemens and Greg Maddux are given as examples of players with good mechanics. Both first pitched significant innings in the majors at age 21. Maddux pitched 1174 innings before age 26, Clemens 1031. It would be hard to claim that it knocked pitches off of the latter part of their careers. I realize that they are obvious outliers, but the point remains. How much of a role does pitching mechanics play in this effect?

I had put this comment on the old thread, but I’ll repeat it here:

Interesting findings.  David shows that hi-K pitchers (pre-age-26) improve their walk rates quite a bit, as I thought might be the case, but this is basically offset by their declining K rate. 

His finding that every extra pitch thrown before age 26 reduces future IP by .5 innings is very important if it holds up.  That means a pitcher essentially loses 8 pitches later for every extra one thrown pre-26!  But I think some more work is needed to be sure this is right.  Some of the stats experts here should comment, but I think controlling for Ks, BBs and FIP-ERA simultaneously may be problematic (getting a positive coefficient for BBs is certainly odd).  It also seems to me that his Pitch variable — once you control for IP, BBs and Ks — may be measuring BABIP to a significant extent.  And while much of that is luck, some of it isn’t, and so the regression may just be telling us that hi-BABIP pitchers have shorter careers.

Pitch count is meaningful only after holding BB rate, K rate, ERA, and IP constant.  What that means is that if you have two guys with identical stats (probably including number of batters faced), but one uses a lot more pitches to get the equivalent result, that one will have a shorter career.

But why would one pitcher use more pitches to get identical results?  Maybe, on a 1-2 count, he’s more likely to waste a pitch.  Pitchers who do that are probably different in other ways from pitchers who don’t do that.  And so pitch count is probably significant for reasons that *correlate* with pitch count, rather than because the extra pitches are hurting the guy’s arm.

For instance: maybe 1-2-wasting pitcher gets the Ks he does because batters fall for the outside pitch.  Eventually, they’ll figure out that they have to stop doing that, and the Ks will drop.  And the pitcher is out of baseball not because his arm suffers, but because, his arm wasn’t as good as his stats to begin with.

Just an example, but I do think something like this is what’s going on.

Phil:  My assumption would be that batters faced is NOT identical, that in fact that’s the main thing being measured here.  If a pitcher uses more pitches per inning, controlling for Ks and BBs, then he’s probably facing more batters.  And if FIP-ERA is controlled for, that probably means he’s giving up more hits on BIP.  Just a guess....

I’m not even sure that it makes sense to control for Ks and BBs, since those are major causes of a pitcher throwing a lot of pitches (though I understand that David’s trying to control for pitcher quality).  Perhaps it would be better to just use P/IP and FIP-ERA as the independent variables?  That is, given X level of talent, does throwing a lot of pitches matter?

Guy: David controls for IP, so you have identical IP, BB, K, and ERA.  The only think that could lead to signficantly more batters faced is H, and since the ERAs are the same, any variation in H is likely to be small.

Phil:  he uses FIP-ERA, not actual ERA.  So H variation could be large.  (This also means he’s controlling for BBs and Ks twice.)

Right, I stand corrected.  Still, it would be difficult for H variation to be *that* large, wouldn’t it, since BABIP variation is low?

Well, back of envelope calculations: 
Two pitchers with 500 IP before age 26, only difference is .020 spread in BABIP (.290, .310).  The .290 pitcher will need about 3 fewer pitches per 27 outs, or about 165 over 500 IP.  So David’s model estimates 90 fewer IP after age 26. 

Let’s assume the true talent BABIP difference is 10 points.  So the lo-BABIP pitcher will have an ERA about 0.25 better after age 26, and was about 0.50 better before age 26.  Is it plausible this pitcher will last 90 more innings, based on performance (and perception of his performance)?  I’d think so....

Guy, excellent!  I think you’ve found the answer.