You’ve been a bunch of assholes. The lot of you. Assholes. Even me, I’ll admit it. There seems to be a lack of understanding about these parts, and I think it’s time to do something about it. It seems to me that a lot of the arguments that take place down there revolve around statistics, as is, regrettably, so often the case these days. I think I’ve made it clear over my time writing here that I fall on the side of sabermetrics rather than what I like to call “baseball card stats” (pitcher wins, RBI, batting average, and other stats that are, you know, found on the back of a baseball card) in that debate, but I don’t take any sort of normative view on people who don’t. What gets me is that it’s turned into an orgy of ad hominem attacks and evolved into the kind of passionate, dogmatic shouting match that has made the discussion of religion and politics into a chore.
I’ve tried to address this issue before, but I think I’ll give it another crack. I think it might be useful to try to explain where those of us who prefer advanced statistics are coming from. If you prefer to be more old-school, I hope that, if you’re not converted, you’ll at least understand why we approach baseball, and evaluating it, the way we do. I know that some of my more devout brethren might take exception to this olive branch, but screw ’em. I’m sick of the crap that goes on in this comment section.
Why We Disagree
Believe it or not, what the baseball card set and sabermetricians spend their lives yelling at each other about is often surprisingly subtle. Most of the time, a good player is a good player by any measure, whether it’s FIP or pitcher wins, RBI or wOBA. If a player hits .300 or drives in 120 runs in a season, that usually (but not always) means he’s had a good year by advanced stats as well. It’s also difficult (but not impossible) for a pitcher to win 20 games without posting good advanced numbers. Where we wind up screaming at each other is when we argue about whether CC Sabathia was the best pitcher in the AL in 2010, or, like, the fourth-best pitcher. In the grand scheme of things, it doesn’t matter a whole lot. It’s not like we’re making up stats as we go along.
We Don’t Make Up Stats As We Go Along
No self-respecting sabermetrician cares about “batting average vs players from Cleavland born on Wednesdays with a last name starting with K.” That’s not a straw man–someone actually said that in yesterday’s comment section. Things like WAR and win expectancy not only have to pass a smell test before gaining widespread acceptance, but they are created and examined with much the same rigor required by social scientists: does it tell us something useful? Is it a fluke, or is it repeatable? Ridiculous characterizations like the one above don’t hold water because they have no bearing on actual performance and that is why statheads don’t actually make those characterizations.
Of course, I love stats like that (I used to bandy about a theory that tied the Phillies’ playoff success to the height of the opposing shortstop until Troy Tulowitzki ruined it in 2009), but only because I find them personally intriguing and not because I find any larger scientific significance. A player’s batting average against pitchers born in Cleveland whose names start with K is a trivia question, not an instrument of sabermetric precision.
If someone references K/9 ratio as an important factor in pitching success or talks about a prospect’s age in relation to his expected performance in the future, it’s because past observations have proven those to be relevant indicators about a player’s performance. Wilson Valdez grounded into a lot of double plays in the past, for instance, and all things being equal, he’ll ground into a lot more in the future. Real, useful statistics don’t come out of nowhere. Although, funnily enough, throughout baseball history, pitchers have had a lot of trouble against left-handed outfielders born in Donora, Pennsylvania on November 21 (see here and here). Though that could be a fluke.
What We Evaluate
In the comment section to the post that finally set me off, there was a debate about Wilson Valdez, a player about whom I’ve gotten into more shouting matches over the past year or so than any other. The argument was, more or less, whether Valdez’s performance in Wednesday night’s game was important to the Phillies’ success. Valdez was hugely important, singling in the bottom of the eighth and scoring the winning run in a 2-1 game through some smart baserunning and a little bit of luck. Did he contribute to the Phillies’ win on Wednesday? Enormously, and that’s not in dispute. But that’s not really the kind of thing we’re after with sabermetrics. Advanced stats are better-suited to 1) explaining long-term trends in the past, 2) predicting performance of players and teams in the future, and 3) identifying optimal strategy and tactics. Like I said, it’s kind of like social science. When the likes of Wilson Valdez help your team win a game, you cheer and order another round of drinks. It doesn’t mean that, after more rigorous examination, it looks like Exxon will come through in the clutch tomorrow.
The Myth of the “Winning Player”
Scott Brosius got a reputation as a “winning player” because, in 1998, he showed up at third base for the New York Yankees and they won 114 games (then an American League record) and the World Series. In Brosius’ four years in New York, the last four of his career, the Yankees won four straight pennants and three World Series. Don’t get me wrong–in 1998, Brosius was a very good player: .300 batting average, 98 RBI, 121 OPS+ and 5.7 WAR, good numbers by any measure. However, in the next three years, 1999 to 2001, Brosius hit a combined .254 with a meager .316 OBP, less than one win above replacement per year. How “winning” was Brosius in those years? Well, not enough to hurt the Yankees’ chances much, but they would have won more games if their third baseman in those years was, for instance, Chipper Jones. Notice also that in the first seven seasons of Brosius’ career, when he was in Oakland, he didn’t have a reputation as a winning player. Know why? Because he had bad teammates and in that time, the A’s made the playoffs once, when Brosius only had 92 plate appearances and didn’t make the playoff roster.
One of the favorite whipping boys around here is Bobby Abreu. Did Abreu look lackadaisical from time to time? Sure. But he is the Phillies’ all-time leader in on-base percentage, a statistic no more advanced than batting average, and hit 195 home runs and stole 254 bases in his eight years and change here. Of course, Abreu doesn’t get a reputation as a “winning” player, because the Phillies didn’t make the playoffs at all when he was here, and started their run of dominance almost immediately after he left. Of course, that may have something to do with the fact that he arrived on a team with one other good position player (Scott Rolen), one good starting pitcher (Curt Schilling), and 22 players who ranged from Biblically awful to more or less average. The Phillies finished six games under .500, and if Barry Bonds had been playing right field for the Phillies in 1998, they may just have finished with a winning record. The trend continued for Abreu’s entire stay, and the team improved after his departure not because Abreu was some sort of karmic albatross, but because Cole Hamels, Ryan Howard, Ryan Madson, and Chase Utley matured into all-star-quality players, and the team started acquiring the likes of Roy Halladay, Jayson Werth, and Brad Lidge. So what does the label “winning player” mean? Good taste in teammates.
One more note, about wins as a statistic. Those who evaluate players by baseball card stats often say that wins are the most important statistic, and to oversimplify a little, that’s absolutely right. When evaluating a team holistically. But when evaluating players, the question gets more complicated. The job of a pitcher is not to win games, per se. The job of a pitcher (and his fielders) is to prevent batters from reaching base. Of course, if the pitcher does his job well, the team ought to win, because you can’t win games if you don’t score runs, and you can’t score runs if you don’t reach base. Of course, that all depends on his batters, whose job is the opposite, to score runs by reaching base. That’s why stats like wOBA aren’t so much based on team performance as by how often a player reaches base and how many bases he gains when he does.
So when we say that wins aren’t important, don’t misunderstand. They’re supremely important, and any argument to the contrary is silly. But one player is only a part of a team, so it’s unfair to judge the man by the outcome of a game in which he only plays a small part. It’s more useful to judge him by something over which he has more control: reaching base and preventing the opposition from doing so.
Above All, We Enjoy the Game as Much as Anyone
This one you’re just going to have to take my word on. If sabermetrics were Christianity, Bill James would be its St. Paul and Joe Posnanski would be its C.S. Lewis. Both of them are rabid Kansas City Royals fans. Mark Simon of ESPN’s Stats and Info blog lives and dies with the New York Mets and has the same kind of irrational likes and dislikes about the team as a grocer from South Orange who can barely add up Francisco Rodriguez‘s assault arrests. So people like me, or Bill Baer, or Paul Boye, or the Seidman brothers, who piss people off on the internet by throwing around baseball-related acronyms that would make the United Nations blush, are perfectly capable of having an emotional attachment to a team. We go to games, we swear at the TV, we buy jerseys and hats, we rejoice when the Phillies win and we despair when they lose.
What advanced statistics offer the average fan, then, is a deeper understanding of the game. It’s more information, and that means more ammunition in arguments about whose team is better, whose cleanup hitter you’d rather have up with the game on the line, and what free agent you wish your team would sign in the offseason. These are the kind of arguments that make up the lion’s share of being a baseball fan. For me, arguing about baseball makes up 60 percent of what I talk and think about, so qualitative concepts like “Wilson Valdez sure has a great throwing arm” and “Ryan Howard strikes out a lot” get a little worn out and left me looking for something deeper.
That’s really, at the risk of sounding needlessly grandiose, what sabermetrics are about: finding truth, understanding, and meaning in baseball. That’s all. It’s not some conspiracy by a bunch of sports-hating nerds who are trying to exact revenge on you for the way you treated them in high school by stealing the beauty from the game you love. That’s really what bothers me most of all about this argument: that someone’s going to accuse me of wanting anything from baseball statistics other than to understand the game better.
To sum, because I’ve gone on for nearly 2,000 words about an argument on the internet, here’s a novel idea: if you don’t understand a statistic, don’t reject it out of hand without understanding it; you’re only exposing your own ignorance and immaturity. I’m sure if you ask, someone will explain it. I’ve never seen a serious question get mocked by a stathead, even on the internet. So try out some of these stats for size–they’re not actually all that complicated to understand because numbers are numbers, whether it’s RBI or VORP–and if you don’t like them, you can say you tried it and didn’t like it, and everyone will leave you alone.
Or you can keep making dick jokes and referencing a misstatement of gambler’s fallacy when someone who’s done his homework makes an argument you don’t understand. But if you do, you’re on your own–I just can’t be bothered to put up with that kind of nonsense anymore.