Like most deeply patriotic American progressives, I love baseball above nearly all other things—mom and apple pie excepted, of course. I grew up a Detroit Tigers fan and was profoundly wounded by the experience. They dropped below .500 just as I got old enough to understand what was going on…and stayed there for almost two decades. Under those circumstances, my brothers and I had to love baseball qua baseball—since there was little point in devoting our emotions in the likes of Skeeter Barnes, etc. Without much incentive to follow Wins or Losses very closely, we dove into the numbers on the backs of baseball cards in search of less depressing stats (Me: “Hey! Tim Belcher lost his first seven games for us in 1994!” Brother: “Keep looking.”).
Fast forward a few decades and the Tigers are respectable—maybe even good (maybe). What’s more, my brother Tyler is a sports statistician who moonlights as an economist at MIT. He’s been analyzing the debate over this year’s AL MVP—should it be Detroit’s Miguel Cabrera or the Angels’ Mike Trout? He’s considerably more accomplished than I am in many ways, but he really takes the cake when it comes to statistical analysis.
But I am the family political theorist, and reading his analysis got me thinking about the limits of quantitative analysis in sport AND politics.
I’ve been suspicious of WAR [Wins Above Replacement] and other complex wins measures for awhile, and since I’m a Tigers fan, I figured this is as good a time as any to try an alternative. I’ll stick to the offensive side due to data constraints for now. I just want a measure of how many runs a player accounts for over the course of the season with his bat.* There will be none of this “replacement player” business, which introduces another parameter to be estimated (read: more obfuscation and error) with no clear improvement in determining the MVP.
After developing an alternative standard for evaluating the two players, he continues (in a follow-up post):
My hope is that True Runs yields a simple and transparent starting point for the debate. By keeping it simple, I miss some important aspects of baserunning and I don’t adjust for park differences,…defensive contributions, or opposition quality. Most likely, Trout wins out on these categories (especially defense and baserunning — Fangraphs breaks down every side of the debate, though I can’t find any explanation of their stats).
However, I also ignore things that matter to MVP voters — playoff qualification, rare achievements (Triple Crown), and clutch performance (RBI)…players likely face harder pitchers/better effort in clutch situations, and Cabrera probably had more opportunities because of his place in the order. He did not back down from the challenge — his RBI total reflects that. These statistics may not measure true skill (whatever that means), but I think they are going to push Cabrera over the top with the voters.
This final point is where
political philosophers like me come in. Statistics can inform debates over which player is the “most valuable,” but only insofar as we’ve come up with a specific framework of what constitutes value. Is the most valuable player the most “complete?” Is he the one with the best developed offensive skills? Is he the one who best fills a specific niche on his team? Maybe. Maybe. Maybe.
Statistics can’t answer these questions without making reference to an implicit definition of baseball value. Models like Tyler’s depend upon making various assumptions: baserunning and defense matter, but perhaps not as much as raw offensive power or positional flexibility or hitting under pressure. These particular choices may or may not be defensible, but some such judgments are unavoidable. That’s because baseball, like politics, and like life more generally, is amenable to being framed in terms of a nearly infinite constellation of goods. Questions of value require us to take a stand about what matters. But since every statistical model makes these judgments, the real debates happen at that level—on philosophy’s turf.
I find Tyler’s arguments compelling—not because they’re statistically extensive, but because the claims underlying his model are compelling.
Maybe it’s worth reiterating that this is just as true for politics as it is for baseball. If you watched this campaign’s first debate, you know that data are too malleable to provide knockout arguments. Strong, convincing rhetoric requires moral language (h/t E.J. Dionne on this point). Otherwise we’re just trading “facts” and “figures” from studies that come to different conclusions because of the different emphases built into their models.
But hey, it’s playoff season, and it’s occasionally best to put politics aside (even during an election year). Go Tigers!