Clutch Hitting and. . . The Purple Heart?

In my last post I looked at the game and how (and if) it's played on paper. One of the more controversial aspects of this new era of statistical obsession baseball has undergone is the issue of clutch hitting. Let's begin there.

Or rather, let's begin this discussion of clutch hitting with a bit of a digression. Once upon a time, there was an era of baseball "fandom" where people's views on baseball were absolutist and concrete (this is an exaggeration, but close enough to the truth). If a player was a .300 hitter he was a good hitter regardless of if he played in Fenway Park or Candlestick. If a pitcher won 20 games he was more valuable than a 12 game winner, no matter what offensive support he received. If first base is open in a late inning and the cleanup hitter is up, you walk him. No matter what. These were hard and fast truths and nearly everyone believed them.

Then Bill James (and the rest of the SABRmetricians came along) and changed all that. Instead of accepting as truth that these long "tested" and accepted truths were--for lack of a better word--true, James began by asking the fundamental questions of human inquiry. What is the purpose of the game: to get hits or get on base? How does a ballpark affect a hitter's ability to hit? What control does a pitcher have on his won-loss total at the end of the year? When is it statistically advisable to give a batter an intentional walk? Before James if you asked a manager why he bunted a runner over to third in the 5th inning with none out, the manager would tell you he was playing the percentages. If you asked a manager why he was keeping a 13th reliever on the roster to exploit the platoon advantage, he would tell you he was playing the percentages. If you asked a manager why Carney Lansford was less valuable than Jim Rice he would point to the numbers and tell you the stats showed that Rice was a better hitter.

What James did was expose, through better statistics, that the old statistics were faulty. That seems to me to be the fundamental misunderstanding of SABRmetrics today. That is to say--people think that "statheads" are attempting to make the game about stats. Clearly the game has always been about stats--when someone attempts to decry Defense Independent Pitching Stats they usually point to won-loss record. When someone attempts to decry EQA they usually point to batting average. What Bill James said is, "this is a game that will always be about stats. So why not find the best stats we can."

This is a mode of thought that can be applied to aspects of the game beyond stats--so long as we understand what this mode of thought is. It is not "let's reduce the game to statistical models and then we'll have the advantage." The mode of thought is a more Socratic, "We don't know everything about the game and we never will, so let's try to see which of our most closely held fallacies is most easily correctible." The eye sees and the mind perceives, but what it sees and perceives is fashioned by our prejudices. It is best not to avoid that those prejudices are there, but rather to acknowledge them and then attempt to move beyond them.

The reason I mention this is that there is a new packet of sports fans who are no different from the old packet--rigidly set in their ways, unequivocally convinced that they are right, and derisive of those who diverge from their POV's. The only difference between the new faction and the old is that this new faction's conviction comes from statistics and research. This is not. . . necessarily. . . a bad thing. But it's something worth noting.

I talked in a previous post about the mindset of certain atheists and I suppose it's worthwhile to return to that idea for this discussion. I wrote this about religion:

If you want to believe in God, you absolutely should. If you want to believe there's no evidence to suggest there's God, you absolutely should believe that. If you want to take that to the Richard Dawkins extreme and say that because there's no evidence to suggest there's God, believing in Him is superstition, that's fine too.

I was talking to a friend about my phrasing here the other day, and he was a bit--let's say--upset that I referred to the Dawkins POV as an "extreme." That word obviously has a number of important connotations within the theological field of discourse and was perhaps poorly chosen (although I think perhaps I did mean those connotations to some extent). My point, at the time, was merely that on one pole of the "God question" you have people who are convinced that anyone who disbelieves in God is undoubtedly wrong and foolish. On the opposite pole of the discussion, Dawkins and those who think like him believe the opposite--that anyone who DOES NOT disbelieve in God is superstitious, wrong, and foolish.

I called these two POV's extreme because they define the shades of belief between them. Any room for tolerance in someone's opposing view falls between these two extremes.

So maybe you see where I'm going here. There are a number of followers of statistics on these here interwebs who have little tolerance for the idea that stats don't tell everything there is to know about baseball. The most obvious example of this is in the Hall of Fame voting every year. When the ballots come out, there's always some loudmouthed, old school baseball writer who attacks the career and defenders of Bert Blyleven as idiotic, explaining in no uncertain terms that he doesn't "feel" like a HOFer. That he just "compiled" stats and that his "stats" don't "measure" his "ability" to "win." All these scare-quotes refer to liberties said author takes with reality, where he redefines terms that he will later use in the same article to defend the HOF credentials of some other candidate (Jack Morris usually). "Jack Morris had tons of opening day starts and complete games!" the man will say, somehow distinguishing Morris's compiling from Blyleven. "Morris had a long run of seasons with 200 IP or more," the man will say, somehow distinguishing Morris's "stats" "ability" to "measure" his quality from the same in Blyleven. And so on.

Obviously as a thinking sports fan I find this mode of thought offensive, not only for its logical fallacies. The HOF is a completely subjective concept (though one with very real consequences for those included and excluded) and the thought that anyone is silly, stupid, or foolish for their opinions is patently ridiculous (I for one believe that Dante Bichette is a no-doubt first ballot guy). The problem is, though, that there is a knee jerk response from the opposite "extreme." While the windbag old-school "proven winner not compiler" faction are furiously pecking away at their typewriting machines, all the while some statistic devotees are composing parallel but opposite columns, deriding Morris (or Dawson or Mattingly) backers for the opposite reasons. Wielding numbers like torches and pitchforks, this vocal online contingency often characterizes supporters of their most hated candidates as the Dan Shaugnessy's of the world treat their pet projects.

Understand, I am not referring to those who question the HOF, MVP, and CY Young votes they disagree with--especially if said inquiry addresses specific fallacies a voter used to justify his vote. My issue is not that the methods these statheads use are wrong. They're not. They're much better, more refined, and more intelligent than the traditional stats. Unlike the Blyleven decrier in the above example, these articles don't bother me on all accounts. Only one--this singleminded misunderstanding of the principle that Bill James used to bring advanced statistics to baseball's foreground. And that is the constant and overarching understanding that even with the help of statistics and intelligent inquiry, that you can not be certain you understand baseball. That is the fundamental misunderstanding, I think, in statistical discourse in baseball journalism today, and one I think that leads to such heated debate over the clutch question.

What is the controversy over clutch hitting? Simply the fact of its existence. That's all. Of course, the debate over whether something which may be intangible and is certainly difficult to define is the most difficult kind of debate. Pornography, God, theoretical economic models--this kind of shit brings out the worst in people. And so does debate on clutch hitting. On the one hand you have those in the "Blyleven sucks" camp ridiculing those who believe anything but absolutely that Joe Carter, Derek Jeter, and David Ortiz have some ineffable attribute of clutch virtuosity and that Alex Rodriguez, Ted Williams and Ty Cobb had the opposite inability not to choke when the chips were down. On the other hand, you have the faction who decries the idea of clutch hitting and anyone who embraces it as witchcraft of the silliest order.

The arguments on both sides are convincing, and that's what makes the singleminded exclusionary discourse so frustrating. Bill James wrote an excellent article on clutch hitting call "Underestimating the Fog" which compares the search for the sustained ability to hit in the clutch with sentries searching via beacon through thick fog for enemy forces beyond their battlements. James argues that, no, statistics do not as of now show that though clutch hits clearly exist--though the definition of clutch is difficult to pin down around its edges--the ability to hit in the clutch is at all sustainable. However, James continues, just because we can not see those statistical proofs now does not mean they don't exist. Like the sentries searching through the fog, we would be silly to conclude that just because we don't see the enemy, that does not mean they don't exist--especially if we have a strong suspicion they are out there.

Of course on the other hand the argument that no one can make the major leagues without being clutch is a strong one. By the time a player makes the major leagues he will have played in thousands of baseball games. Advancement is always heavy on the mind of the high school, college, and minor league player--and at no point is the pressure any less. Fail to perform as a high school player and you will not be drafted or asked to play for a college ball team. Fail to perform in college and you will not advance to professional ball. Fail to perform as a minor leaguer and you won't get called up to the major leagues. At any point, constant failure leads to the end of a career, leaving only the "clutch" performers in the pool of major league ball players.

Of course this doesn't take into account players with scads of talent (like Javier Vazquez, an oft-accused "unclutch" player) who teams are willing to take a chance on because of raw stuff irrespective of "performance" and players who are supposedly "compilers" (Alex Rodriguez, for example) who are supposedly good enough to put up huge numbers against lousy teams in low-pressure situations to make up for their failures "in the clutch."

It all sounds well and good, as long as you refuse to listen to the opposing argument. And the stats are little help. What is a valid statistical model? Much like the concept of "heart" any attempt to define "clutch" can be disarmed. Is clutch measured by the statistic "close and late"? Close and late is defined as  an at bat in the 7th, 8th, or 9th inning when the deficit is 3 runs or less in either direction. The problem with using this to define clutch hitting are legion enough to make this statistic laughable. According to close and late, an April at bat to lead off the 7th inning with the score 3-0 in your favor is valued equally to a Setpember at bat with two out in the ninth, a man on second, your team down by one, played against the first place team in your division while you trail them by one in the standings. Everything in between that is defined as close and late. Obviously this is a wildly imperfect measure of clutch hitting (and yet people still use it to prove whatever they hope to prove about clutch hitting, proving time and time again that people with an axe to grind will use whatever idiotic methods they can to quote/unquote prove their point)

But even more adequate definitions of clutch are assailable. Is clutch any at bat in a pennant race? Clearly must be more specific than that, as a third inning at bat in a blowout shouldn't count for anything w/r/t clutch. Is it an at bat when the game's outcome is in question? The opposite problem here, as games in April or May or while in 4th place clearly shouldn't count the same as games in a pennant race. And what about the playoffs? How do those at bats compare in clutchness to regular season games? To pennant races? And what about pitching? Do we only count games as a whole? Or "clutch" pitching in leveraged situations (like with the bases loaded and no out)? And how do we account for the fact that the pitcher himself created those leveraged situations?

Perhaps we could define a system where each individual outcome were rated on a scale of 0-100 for its "clutchness". In such a system an at bat in the bottom of the 9th in the 7th game of the World Series with the tying run aboard, down by one, and two outs would rate something like a 98 (because it could always be in extra innings, or your job might be in jeopardy, or your dying father might be in the stands, or Robert Deniro might have kidnapped your son and threatened to kill him unless you came through. . . or didn't come through. All those things would probably be required for a rating of 100). But even if you came up with such a system and everyone agreed upon its merits and you went about scoring players in terms of their clutch value, you'd run into a serious problem: inclusiveness.

You could, theoretically prove that a certain player has a certain sustainable clutch ability. And you could conceivably prove that his clutch ability is superior to another player's clutch ability. But you could not come up with a realistic standard of measurement. Why? Because a lot of players simply do not get clutch opportunities. There are two many teams in baseball who are out of contention before the first pitch is thrown on opening day to get a standard of what is "good" on average and what is "bad" on average. For nearly every other stat (even "Runners In Scoring Position") every regular player in baseball will get a comparable amount of opportunities. But even if, say, Derek Jeter seems to be a monster when hitting with the game on the line in playoff and pennant race situations year in and year out, how can we compare him to a norm when guys like Adrian Gonzalez, Hanley Ramirez, Ian Kinsler, Nick Markakis, and many others haven't sniffed a playoff situation in years. And then there are the bit players, the ones who help us establish a mean.

If there were only 5 or 6 everyday players in major league baseball, only 5 or 6 players getting 600 PA's a season, and the rest of baseball were platoons, how would we know how difficult it were to hit .300? We could point to the fact that only 1 out of 6 players who got 600 PA's hit .300, but what would that really tell us? It would tell us they were better than the other 5. Even if we expanded that number to 30 players with 600 PA's, and 6 out of 30 hit .300, that still wouldn't tell us much about what hitting .300 over 600 PA's means.

You get the point. So what can we say about clutch? Is it worth talking about at all? I would say yes. Here's my argument: we may not be able to define clutch, but much like pornography we know it when we see it.  In Game 5 of the 2001 World Series, with the favored Yankees down to the Arizona Diamondbacks two games to one, with the shadow of 9/11 looming, in extra innings with the score tied and a full count, Derek Jeter hit a walkoff home run against D-Backs closer Byung-Yum Kim.

That is clearly a clutch moment. There is no doubt in the world that it is a clutch moment. And in that clutch moment, Jeter hit a home run to win the game. What does that mean? Does that mean Jeter is a clutch player? Probably not. But it does mean he came through in the clutch. And Yankees fans can name any number of similarly clutch performances by Jeter. Clutch performances mean more than regular performances by definition, and it's something to consider when looking at a player. Does everyone get a chance to have a clutch moment? Probably, but some have significantly more chances than others. Does everyone get a chance to have a moment like Jeter's World Series home run? Obviously not. But Jeter did and he came through.

You might be wondering why "The Purple Heart" is in this post's title. It's because in this respect being a clutch hitter seems to me a lot like being a war hero. I don't mean to trivialize war by comparing it to baseball. I don't mean to compare the two in any meaningful way. But I will say this: not everyone gets a chance to go to war, and not everyone who gets a chance to go to war has an opportunity to rise to a challenge and become a war hero. In fact, it's hard to define the gray area between what is an exceptional act of heroism and what is just the ordinary duty of a soldier. However, there are people who have undoubtedly had the opportunity for heroism of an exceptional order. Does this mean that they are of greater personal fiber than those who have not had a chance to rise to such an occasion? Perhaps or perhaps not. Does it mean they have some sort of sustained ability to be a hero in every occasion which calls for it? Perhaps or perhaps not. Does it mean that a failure to become a war hero is a deficiency in a person? Absolutely not.

What it means, and all it means, is that on at least one occasion, this particular person had a particularly exceptional challenge and on this particular occasion he rose to it exceptionally. That is something that may be of small importance in the grand scheme of things (and obviously of smaller importance in baseball than in war) and may be an achievement surrounded by hoopla and exaggeration and used to extrapolate things about the person involved which are likewise surrounded by hoopla and exaggeration, but it is something that happened and means something.

That shouldn't be a radical point to make, but sometimes it seems like the contemporary discourse makes the obvious and simple seem radical, and so--there it is.