THE BOOK--Playing The Percentages In Baseball

I am not, in general, a big fan of these, “Let’s look at his BABIP, line drive rate, swings and misses, swing rate at pitches in or out of the zone, etc., in order to see where a batter REALLY is.” Not a fan at all.

It is OK to dig deeper into a player’s stats in order to filter out some of the noise, but there is absolutely nothing you can do to “filter out” sample error.  And a big part of what makes players’ performance fluctuate is pure sample error.  And there is nothing you can look at that will help you filter out the “noise” in pure sample error. It is not like looking at a batter’s triples rate in order to estimate his speed (lots of noise) and then timing him around the bases (almost no noise).

For example, let’s say that a batter’s walk and/or K rate has changes significantly from his expected rate.  And let’s say that we look at his swings at pitches out of the zone and find out that it is or it is not the same as it was before?  So what?  That really does not help us ALL that much. It helps us a little (we want to know, for example, whether he has received a lot of pitches in or out of the zone, or whether he has had a worse or better “eye” than in the past. But the problem is that there is plenty of sample error in how often a batter swings at pitches in or out of the zone (or anything else you want to look at).  In fact, the uncertainty around that swing rate is a binomial, distribution, and there ain’t nuttin’ you can do about that (I suppose you can try and get even more granular than that by testing his eyesight, looking more closely at what counts he swings and misses, what counts he gets into, etc., but there is lots of uncertainty in almost anything a batter (or pitcher) does, no matter how granular you get in the data and there ain’t nothing you can do about it.

Anyway, to recap, while it definitely helps to look at more granular data when analyzing player performance, we have to be careful about assuming that we have more certainty than we actually have in that data, as far as an estimate of a player’s true talent is concerned.

And I would like to see more “hard numbers” provided by these guys who do the analysis.  For example, let’s say we have a player who has an OPS of .900 and his projection was .800.  So he has overperformed for some period of time. Now we hear all the time how since he has a BABIP that is high, that his .900 is not “sustainable” but that if his BABIP is normal, it is sustainable. I’m afraid that is not good enough for me.  I want to know what that high or low BABIP does in terms of our new projection. “Sustainable,” “not sustainable” means nothing to me.

And BTW, I don’t know what to make of the data in this article.  Ramirez is walking more and striking out more. OK.  He is swinging at fewer pitches out of the strike zone.  Well, no shizit.  I would expect that from almost everyone who is walking more.  He is striking our more even though he is walking more too and swinging at fewer pitches out of the zone.  OK, that is fairly interesting.  Maybe he is swinging harder when he does swing. Maybe he has just gotten a lot more tough pitches to hit so far this year. Maybe he has just swing and missed a lot this year for no apparent reason.  I just don’t know what to make of the data in the article, and I am not sure what we are supposed to make of it.

It would be one thing if you used a “FIP for hitters” sort of thing, and actually tested this sort of estimation over a large number of players. But a lot of these kind of analysis, you never see the track record of how well they predict the future.

Just from watching Ramirez play, it doesn’t look like he’s changed his approach. He’ll swing if he things he’s got a pitch he can hit. Of course, I’m not a scout and I shouldn’t pretend to be.

Realistically, I wasn’t suggesting his approach has changed or trying to prove anything concrete.  This was more of an example of a stream of consciousness type post wherein I investigated some of his numbers as I went on, using the numbers at Fangraphs.

It all just started with somewhat of an “anomaly.” Of all the guys with a 1.84+ WPA, everyone had 14+ HR, except Ramirez, who had 9.  So I wanted to check to see if he had done anything, such as an inordinate amount of doubles, or something along those lines, that might make up for the dropoff in HR and enable him to have a WPA that high at this juncture of the season.

What jumped out was his OBP of .417; I’m not a Cubs fan or Ramirez follower but I never knew him to be on-base that much, which was then confirmed by his much lower OBPs in past seasons.  Why is he getting on base?  Of course it’s an increase in walks since he’s hit .300+ before, but I didn’t realize his BB% is almost double what it’s been in the past.

Since his BB% is so much higher, I thought he must be either more selective or getting more pitches out of the zone; his pitches in/out of the zone were virtually the same so it wasn’t that, but he was swinging almost 7% less at pitches out of the zone, therefore being more selective (whether he realizes it or not, which wouldn’t necessarily be indicative of a change in approach).

Then I noticed his K% was higher, too, which seemed odd/interesting.  I figured he was making less contact, which he was, by about 6% from a year ago.  So, put together, he was swinging less at pitches out of the zone while getting virtually the same amount of pitches out of the zone, and making less contact overall.  His BABIP was still .331, so it didn’t seem as though the dropoff in contact had affected his overall statline.  Not to say he will or won’t sustain anything, but rather just noting what has happened, not necessarily what will happen.

His % of flyballs is virtually identical to a year ago, as well, yet his HR/FB has dropped 3% so less of the flyballs are leaving the yard.  I think the end of the article blurs it together and tries to predict future performance incorrectly, but otherwise, the data in the article is really just there to investigate the questions that stemmed from the initial question I had.  That’s all.

I know you (Eric) weren’t really saying anything about Ramirez’ true talent or his projection.  I was ranting a little about the articles, and there are a lot of them these days, that discuss whether batters’ (or pitchers’ to a lesser extent) performances are “sustainable” or not, based on things like GB/FB ratios and BABIP.  I have some issues with them.

When you talk about pitches out of the zone, how many are we talking about?  What about contact rate on pitches in the zone?  How many?

When you say, “7% less”, do you mean, like, 10% versus 17%, or like 15.8% versus 17% (15.8% is 7% ‘less’ than 17% also).

I hear you on that.  When I said 7% less, it’s more 10% vs 17% than the other, which is a very good point you raise.  So if Ramirez was at 30.76% last year and is at 23.54% this year, the difference in his rates is ~7% but it’s actually closer to a 31% dropoff.

As far as pitches, Ramirez has seen 1137, 50.05% of which have been in the zone.  So:

IZ: 569 pitches
OZ: 567 pitches

IZ Contact - 86.35% = 491 pitches w/contact
OZ Contact - 54.50% = 308 pitches w/contact

I was trying to get a handle on how statistically significant 6 or 7% was.  It is quite, for 500 some odd pitches.  Then again, we don’t know whether last year’s OOZ pitches were more “out of zone” than this year’s.  What happens of course, is that that when batters show a propensity to swing at pitches out of the zone, pitchers throw more of those pitches. “More” also means more pitches way out of the zone.  Than the batter stops swinging at so many of those pitches.  Then the pitcher starts throwing fewer pitches out the zone.  Etc.

So we have to be careful about concluding or suspecting that a batter has reached a new level of true talent when some of these numbers (like swings on OOZ pitches) change drastically.  These drastic changes could be a normal part of the constant adjustments and re-adjustments that pitchers and batters make.

I am just thinking out loud.

Ramirez has been swinging at pitches out of the zone since 1998 so it’s unlikely that what he swung at last year had more of an impact than the previous 10 years combined.  He’s always chased balls out of the zone until this year.  Are they more out of the zone?  That’s probably not likely since pitchers have been pitching him mostly the same way for the last 4 or 5 years and he didn’t chase any more pitches out of the zone last year than he did the years before that. 

I’ve never seen Ramirez take so many close pitches.  He’s consistently laying off pitches just an inch or two out of the zone and these pitches he would swing for the fence on in previous years.  The low and outside offspeed pitch has been chased by Ramirez very few times this year and I’d bet that most of the times he’s done so it’s been in a 2-strike count where he’s protecting the plate. 

He definitely has had a different approach this season.  Is this approach new or just a sampling issue?  I don’t know, but so far he has tightened up his zone considerably over the last several years. 

He’s had about 280 plate appearances this year so the chances that he has established some kind of new true talent level are starting to be pretty good.  Does this mean his 2008 level is the new level?  I’d guess it’s in between his previous years and what he’s done this year.

#7, those are all good points.

One thing I have always wondered is if batters who “chase” have worse WPA or WPA/LI than their context-neutral stats would suggest.  It would seem intuitively that they should.

For example, in today’s ATL game, Francouer was up with the bases loaded and one out and the pitcher was trying for a K.  I thought to myself, there are not that many batters whom I would like to see less up there (as a Braves fan).  Sure enough Francouer struck out after swinging for the fences (a bad approach) on 2 fastballs and then striking out on a check swing on a slider in the dirt, a typical Frenchy AB in that situation.  Sure enough, the announcers said that he was 2 for 17 with the bases loaded this year, with zero sac flies.  I did not know that.  It just seems that when a pitcher throws fat pitches to Frenchy he hammers them, but when the pitcher throws good pitches out the zone, Frenchy is dead meat.  There are lots of batters like that of course.

Anyway, I wonder if anyone has looked at the correlation between WPA/LI and context-neutral difference (clutch differential?) and batting eye as measured by something like percent of pitches out the zone swung at and missed.

I’m not aware of any research that’s been done in that area.  It would be interesting to see. 

Since you brought up Francouer, I think there might be some similarities between this discussion and Francouer in 2007.  I recall the announcers and even reading some stuff from the sports writers about how he’d improved his plate discipline.  They’d point to the 40 walks (not sure how many exactly). Upon closer look though, it appeared that Frenchy increased his walks in only 2 months and the other 4 months were right where you’d expect him to be.  Of course the announcers didn’t put 2 and 2 together and surely Frenchy improved somewhat from 2006 to 2007 in this area, but not to the extent these people made it sound.  Many people were insistent that he had improved his approach at the plate, but the 2 months where he was actually much better at taking walks was the first and last month of the year so it wasn’t even 2 months in a row. 

I wasn’t optimistic that there had been any real change from Francouer, but I am more optimistic about Ramirez, mostly because I am a Cubs fan and have seen him play most every game over the last several years.  There is a difference this season.  It could still be a number of things, which doesn’t allow us to come to very good conclusions, but I’m beginning to lean towards Ramirez having a new true talent level.  It’s been his ability this year to lay off the close pitches that has gotten my attention.  He’s always been the type of hitter where you can throw something with in 3 or 4 inches of the plate and he’s hacking.  Another thing I’ve noticed with Ramirez this year is that he previously had a tendency to go after the high fastball, but several times this year he’s laid off and several other times he’s started to swing, but checked it.  It seems to me there’s some conscious effort on his part to not swing at balls out of the zone this year. 

That may be completely untrue.  As I said before, I just don’t know. I think there might be something there.

Speaking of players who have no idea of the strike zone, why does Wily Mo Pena have a job?

He has no defensive value, can’t run the bases (too slow), and can’t hit.  I keep having a decent hitting projection for him because of his size and his age, but he keeps disappointing.  When I watch him, he seems to have absolutely no clue as to what the pitcher is going to throw, no pitch recognition skills or strike zone judgment, and the only way he hits an off-speed pitch is when the bat accidentally makes contact with the ball.