THE BOOK--Playing The Percentages In Baseball

Friday, May 23, 2008

Size of parks and talent: Be careful about what you are pretty sure is true…

Almost no one, even analysts you read on THT, BP, and other sabermetric columns, blogs, and web sites, does not think that fly ball pitchers or pitchers who give up a lot of HR’s (pretty much the same animal, I think) fare better in stadiums in which it is difficult to hit a home run, and have no business pitching in a stadium that allows a lot of HR (present company excepted).  Remember the hullabaloo when Cincinnatti signed Eric Milton, a fly ball and home run machine?

I have for years tried to prove that that CW is true and have thus far failed.  In fact, I have never subscribed to that theory.  I don’t subscribe to any theory unless there is some (credible) evidence that it is true.  Nor have I ever subscribed to the commonly accepted (again, even in analytical circles) theory that fast or good outfielders far better in larger outfields, and ditto for small outfields and slow defenders (you know, the old axiom that in a large OF, you need fast outfielders and in a small outfield, you don’t).

Well, these things appear to make intuitive sense, but like a lot of things in baseball and in life, appearing to make sense does not the truth make.  If that were the case, all of the talking heads would be geniuses and we, the skeptics and seekers of the truth, would just be drinking diet coke in our mother’s basements.

Anyway, has anyone ever (I mean ever) seen any proof or evidence that either of these propositions is true?  I haven’t, and I have read as much sabermetric research as anyone I would think (since I don’t really have anything better to do).  That is especially true of proposition #2, that of the fast and slow outfielders.

I did a study looking at the issue, using outfield size in square feet, speed ratings on OF’ers, and UZR.  I did a preliminary study which seemed to show that CW was right and that speedy outfielders did “more better” than slower ones in large outfields, as compared to small ones.  However, I looked at entire outfields and did not break them down into left, center, and right (so that left field at Fenway was treated the same as CF and RF (it is a small outfield, all together) and I did not look at that many years (I used a small sample).

Later on, I increased the sample size and broke down each outfield into left, center, and right.  A much better study of course.  I don’t think I ever published the results, although I might have mentioned it, but there was some evidence that a faster outfielder did comparatively better in a small outfield.  I don’t know why, or if the results were significant, but there was definitely no evidence that a large outfield requires a speedy outfielder.  No evidence at all.  In fact, as I said, the evidence, for some reason, suggested the opposite.

Getting back to the first proposition, that you don’t want a flyball/home run prone pitcher in a “small” stadium (one with a high HR PF), that would probably be true if HR park factors were multiplicative.  In other words, if a batter or pitcher who allowed 6 HR per 500 PA pitched in a park that had a HR PF of 1.5, he would give up or hit 9 HR, and if a player had a HR rate of 20, he would allow or hit 30 in this same park.  But that is apparently not the case.  Tom Tippett in a SABR convention a few years ago, did a nice presentation in which he showed that HR PF are likely a combination of multiplicative and additive.  IOW, the 6 HR player might get a 25% boost in that 1.5 HR PF park and an additional 2 HR, for a total of 9.5, and the 20 HR player would also get a 25% boost plus an additional 2 HR, for a total of 27.  So the 6 HR guy’s HR rate increases by 58% while the 20 HR guy has only a 35% increase.  I made up the numbers, BTW.  I don’t know the proper percentages and additions.

In other words, it is probably NOT true that a high HR pitcher fares really poorly in a high HR park.  In fact, it might be the other way around.  The low HR pitcher might do worse, comparatively speaking.

What brought this whole thing up was that I came across an old and very good article/analysis on BP by James Click that looked at extreme GB and FB pitchers and how they were affected by various component park effects, including HR.  The hypothesis was that since most park factors have to do with fly balls, that the fly ball pitchers should be affected more than the ground ball pitchers in many of these component PF’s. Guess what he found?  Virtually no difference in how extreme FB and GB pitchers are affected by component park factors.  He was surprised.  I was not.  As I said, I have been trying to find out what types of players (other than L/R), batters and pitchers, do better in the various parks, and I have pretty much struck out across the board.

Towards the end of the article, Click says this:

With GB/FB ratio appearing to have no effect on park factors, we must concede that the other unique aspects of each park--weather, hitting background, surface, altitude, etc.--cause similar amounts of variance for both types of pitcher.

Anyway, the moral of the story is to be really careful about accepting as the gospel that which we hear all the time, even when it comes from reliable sources whose intelligence and integrity we respect, and even if it really, really, seems to make sense.