Monday, March 18, 2013
The meaning of baseball statistics
By David Schoenfield
In writing about stealth MVP candidates earlier today, I'm reminded of something Bill James wrote 25 years ago. In 1987, Andre Dawson won the National League MVP Award. Playing for the last-place Cubs, Dawson led the league in home runs and RBIs, but hit .287 and drew just 32 walks, leading to a mediocre .328 OBP. That was the crazy rabbit-ball year and the overall NL OBP was .328 -- including pitchers. Dawson was not a good choice, not one of the 10 or 15 most valuable players in the league, despite the home run and RBI totals. In the "1988 Baseball Abstract," James wrote:
There are occasions in your professional life that make you think you're not making any progress. The election of Andre Dawson as the National League's MVP is one of mine. One of the ironies of editing the Baseball Abstract is that I am occasionally attacked for placing too much faith in baseball statistics and missing the other elements of the game, the things that "statistics can't measure." The reality is that the essential work of this book is to try to teach people not to trust statistics. It was never my idea that we needed to look more carefully at baseball statistics because statistics are the best way to look at baseball. It was my point, rather, that people do make judgments about baseball players primarily by statistics, not should but do, and because they do they need a better understanding of what those statistics really mean.
We've made a lot of progress in understanding baseball in the past 25 years, although last year's Miguel Cabrera-Mike Trout MVP debate was framed, to a large extent, around the same philosophical argument James made a quarter-century ago. Trout failed to win the MVP Award, not so much as a rebellion against sabermetrics and WAR, but because of Cabrera's statistics -- namely, his batting average, home runs and RBIs. (The fact that the Tigers made the playoffs and the Angels didn't should have been a non-starter, since the Angels won more games.)
Even though we know we shouldn't completely trust a statistic like RBIs in evaluating a player's contribution -- it's heavily dependent on opportunities and where you hit in the lineup -- it's still a key influence in MVP voting. Last year, eight NL players drove in 100 runs; six of them finished in the top 11 of the MVP voting. In the AL, nine players drove in 100 runs and five finished in the top 11 of the MVP voting. (Don't get me wrong: I'm not comparing Cabrera, who had an outstanding season, to Dawson.)
I'm not even suggesting RBIs is a horrible statistic and should be ignored; I still enjoy knowing how many runs a player has driven in. But the context needs to be understood. That's why I often cite statistics like WAR or Defensive Runs Saved or other advanced metrics; these are all attempts to better understand the game and understand why teams win or understand more accurately the value of a player beyond batting average, home runs and RBIs.
Yet there is still a lot of push back against stats like WAR. The majority of fans still don't trust it or believe in it. A few weeks ago, SportsNation asked on ESPN.com, "Do you consider WAR to be a useful statistic in evaluating baseball players?" The "no" votes outweighed the "yes" votes 63 to 37 percent. It was a landslide rebellion against new metrics. It was Andre Dawson winning the MVP Award.
As Sam Miller wrote in his excellent essay on WAR in ESPN The Magazine, "I'm not a mathematician and I'm not a scientist. I'm a guy who tries to understand baseball with common sense. In this era, that means embracing advanced metrics that I don't really understand. That should make me a little uncomfortable, and it does."
I'm not telling you that you have to pay attention to all these new stats. There are many ways to enjoy baseball. I try to embrace all of them -- a hot dog at the park, the confounding flutter of R.A. Dickey's knuckleball, a rally in the bottom of the ninth by your favorite team, the hope of a rookie, Pablo Sandoval's girth, arguing whether a bunt should have been attempted, a game on TV on a lazy Sunday afternoon.
But the stats are here. The people who run teams are using them. Although as James wrote, you still sometimes wonder if they're using the right ones. (Did the Angels give Josh Hamilton $125 million for his second-in-the-AL 128 RBIs or for his 25th-in-the-AL 6.9 WAR over the past two seasons?)
I get responses from readers all the time on Twitter about WAR; it makes them angry. That's the part I don't understand. It's just another screwdriver in the toolbox, albeit a very high tech one.
Think about it this way:
Player A goes 150 for 500.
Player B goes 137 for 500.
Can you tell the difference between those two hitters, all other things being equal? You couldn't. Player A got an extra hit every two weeks. And yet Player A is a .300 hitter and Player B is a .274 hitter. One hit every two weeks.
We know Player A is the better hitter in this example because of statistics.
We all use them. All the time. They are a part of baseball, whether you're a mathematician or not.