THE BOOK--Playing The Percentages In Baseball

Would it be feasible to use statistical methods to see how many pitches it takes for a pitcher to become ineffective, and then enforce that as a pitch count on that individual pitcher?  The theory behind PAP is “Pitching is not dangerous, pitching while tired is dangerous.” I agree with this, but they’re idiots if they think that every pitcher magically starts to get tired at precisely 100 pitches, which is more or less the mathematical basis for PAP.  Wouldn’t it be better to look at each pitcher and say “pitches 110-120 are 30% (or whatever) less effective than his first 10 of a game” or something similar and then measure fatigue starting from pitch #110?  I like that the Rangers are taking guys off of pitch counts and instead pulling guys when they get tired, but I would like to see a more scientific way of measuring tiredness than “The Manager thought he looked tired” or “he’s at 100 pitches therefore he’s tired.”

First of all, it is simply not true that pitchers routinely get pulled after 100 pitches.  Second of all, if you are going to make sure that pitchers don’t throw a lot of pitches, you might as well have some magic number and use that as much as possible. If the media and all the other “pitch count critics” want that magic number to be 103, I don’t think any teams would have a problem with that. Third of all, and most importantly, since we now know that relievers are substantially better than all but the very best starters (and then, even closers and some set-up men are better than the best starters) after several times through the lineup, why the heck is there even a discussion about this in the first place? Pitch count or no pitch count, I generally want my starter out of the game as soon as possible in order to bring in a better pitcher.  If that also helps keep the starters healthy, which it may or may not, that is a bonus.  In fact, since managers would keep bad and mediocre starters in the game if they were pitching well, I would make up stories about pitch counts just to get my manager to take these pitchers out of the game and get a better pitcher in the game (and pinch hit for my pitcher in the NL).

I mean, before you start going off on how “pitchers are not trained to go deep into games,” or before you go complaining about the lack of complete games, don’t you have to show that leaving starters in for 120 and 130+ pitches is actually GOOD for a team, other than the injury thing?  Well, it ain’t (good for your team) because pitchers suck the 4th or 5th time through the lineup.  If pitch count is a proxy for “times through the lineup” which it essentially is, then it is damn straight that you want pitchers on a pitch count and it has nothing to do with risk of injury or subsequent ineffectiveness.

I agree with mgl says about when the reliever becomes a better option for the team.

In response to #1, I agree that each pitcher has a different limit, and teams should work on strengthening the pitcher in order that he is capable of going longer. To do that, you need to know when a pitcher is tired. When watching games, expecially when scoring and tracking pitches, I look for

1. fewer pitches in strike zone
1a. loss of location, such as higher on fastballs
2. more swings putting balls in play

pfx might detect a change in release point, if a team can access it in real time.

I wouldn’t generally look for a loss in velocity, as some pitchers can keep it to the end, but others might lose.

I’m curious.  Has there been a change in relative effectiveness of starters the 4th time thru a lineup vs releivers over time?

I know there is a study about this phenomena in The Book (and elsewhere?), but many of those studies looked at a recent time interval.  I forget if it was 1998-2002 or 2002-2006, but something like that.  Either way it was centered in the post-LaRussa bullpen management era.

The average releiver from 1975 (or whenever men were men and starters pitched deep into games) is different than the average releiver today, but it’d be worth investigating whether or not those 70s era pitchers “who were trained to go deep into games” were able to maintain more effectiveness thier 3rrd and 4th time thru a lineup in comparison to their own effectiveness the 1st and 2nd time and in comparison to thier contemporary reliever pool.

Establishing that today’s releivers are better than “untrained” starters isn’t quite the same as establishing the fact that the “training” of starters from the 70s (or whenever) wasn’t effective then and if it was than perhaps could at least be somewhat effective now.

Since we don’t know how much of the “times through the order” phenomenon is fatigue and how much is familiarity by the batter, it is true, as Philly says, that we don’t know whether training might mitigate the effect.  I doubt it would mitigate it very much, however.  One reason is that we see a gradual “times through the order” effect.  Certainly pitchers today can pitch through 2 or 3 times through the order as well as pitchers of the past, so if we see a drop-off at 2 and 3 times through the order, which we do, I doubt it has anything to do with pitchers today not having the same stamina as pitchers of the past.

I firmly stand by my point, which is, “Why would we even want starting pitchers to go deep into games?” Don’t tell me, “To save the bullpen from fatigue.” If we know that relievers are better than starters, what are we saving them for?  Yes, it is true that the fewer inning the bullpen has to pitch per game, the better the reliever we can use (if we knew that relievers only had to pitch 1 inning per game, you could use your best 2 relievers every game).  But, if you can use other relievers - other than your 2 best - for another 2-3 innings per game and they, on the average, are better than starters, which I think they are, then you STILL don’t want your starters going deep into games, as a general rule.

If the M’s are out of it by Sept 1 (when rosters expand), I will be emailing them every day to beg them put in a new pitcher every inning when Felix is not pitching.  “You’ve got one inning… show me what you’ve got.”