The "As He Goes, The Team Goes" Fallacy

Steve Lombardi was one of a handful of bloggers who answered questions from the NY Times Bats Blog about the upcoming season, and the following statement caught my eye:

Considering the ultimate price tag ($8 million for one year), it may have been the prudent move to bring Damon back for a year. But, then again, there was that “budget” concern which we heard so often. It may just be conventional wisdom on my part, but it seemed like Damon was in the middle of things happening last season more times than not. For the record, his OPS in wins was over .900 and he batted .245 in games that the Yankees lost.

I advocated that the Yankees bring Damon back, so I understand where Steve is coming from, even though he dismisses the budget concerns. What interested me was the idea that Damon was in the middle of things last season, with significantly better numbers in wins than losses. I believe Steve falls for a common fallacy here, as teams obviously hit significantly better in wins than they do in losses. American League teams hit .305/.376/.506 in wins compared to .226/.291/.346 in defeat (of course, these numbers are meaningless as well, as they simply reflect that poor teams do not hit well, an obvious conclusion). Most players will have vastly superior numbers in wins than losses, because their superior hitting in those games will lead to victories. This presents a circular train of logic whereby the player’s performance is attached to a particular result by nature, yet people attempt to read meaning into the data. Looking at a player’s numbers in team wins and losses and concluding that he was a more significant part of those results than other players without looking at the numbers for those teammates is silly, as you are lacking context. Take a look at the numbers for the Yankees starting lineup last season:

[image title="Picture 1" size="full" id="16346" align="center" linkto="full" ]

Outside of Jeter, who was fantastic in both wins and losses, every other regular shows a marked split between performance in wins and performance in losses. Splitting by wins tends to capture good performances for most players, as good performances from multiple players will often lead to victories. It is a split that tends to tell us little, if anything at all. Johnny Damon may have been “in the middle of things” last season, and there are a number of ways that one could attempt to support that notion. The win-loss split is not one of them.

Follow Me On Twitter