A Thought About WAR

This is not related to the debt ceiling deal, but since it's the Trading Deadline today (five+ hours ago) I thought I'd mention this little thought I have about baseball statistics. The formula for Wins Above Replacement for position players is that you take the runs above average that this player contributes from various components of the game, like hitting, fielding, and baserunning. To that you add a number of runs that represents, more or less, how grateful the team is that they had this player playing for them instead of the random AAAA-level player they would have to call up instead, and then, and this is my point, an adjustment for the player's position. As I understand it the point of the positional replacement is to make it so that an everyday starter who is league-average at his position will have the same WAR regardless of which position they play. There are several points to this proposal, one of which is that it makes it easier to compare different players across positions fairly, which I think are important. But I want to point out one little flaw in that analysis:

For a team as a whole, both the "runs from replacement" and the positional-adjustment runs are fixed over the whole year. That is to say, in any given league in any given year you get x runs per game for not being a replacement player, and y runs per game for playing your position. But the team has someone who counts as not being a replacement player playing every position every day. For instance: in 1986 the New York Mets won 108 games, while the Pittsburgh Pirates won just 64. The Mets, as a team, had 39 runs from positional adjustment, and 196 runs from replacement level. The Pirates had 39 runs from positional adjustment, and 191 runs from replacement level. The difference in replacement level runs is probably due to the slight difference in team plate appearances ('cause the Mets were a better offensive team, I assume); the Mets had 3% more plate appearances than the Pirates, and about 3% more runs from replacement level. So, the same.

In other words, a team does not need to worry about how many runs it gets from these two categories overall. It just needs to think about how many runs above average it gets from hitting, fielding, and baserunning. So when someone says something like, "Michael Bourn (CF) has more fWAR than Prince Fielder (1B), wouldn't people go crazy if the Braves acquired Prince?", that's not quite the right way to look at things. The Braves already have somebody playing center field. They already have someone playing first base. That doesn't change. The question is whether Bourn is an upgrade to the team compared to their very specific replacement center fielder (he is), and whether Prince Fielder would be an upgrade to the team compared to their very specific current first baseman (he would be). I like the positional adjustment for comparing different players to one another for many purposes, but I think it can kind of obscure things when it comes to things like trades.