[nat]

Since it keeps coming up (especially in my posts), maybe a brief explanation of Wins Above Replacement (WAR) is in order.

The idea is to quantify how much value a player produced, in a way that allows you to compare players across teams and across eras. If the player hadn't been playing, there would have been an open roster spot on his team, which would probably have been filled by some guy from AAA. The performance of that AAA guy is the "replacement" from the stat's name. So WAR tries to calculate how many wins a player would generate for a random team (the randomness is necessary to allow cross-team comparisons), beyond what would be produced by the kind of AAA player that every organization has hanging around.

They do this by finding the "run expectancy" of every event that the player takes part in. Because baseball keeps very good records we know, for example, how many runs, on average, are scored after a player hits a single (or a double, or steals a base, or gets caught stealing, or strikes out, or etc.) That number is the run expectancy for the event. The last time I saw a table setting these out (which was a few years ago, so the numbers in this post are a bit out of date) the run expectancy of a single was about 0.3. The run expectancy of a home run is 1.4. (It's greater than one because there are often players on base when a home run is hit.) We do this with defensive plays too (although it's a bit more complicated with defense). Adding all of those up gives us how many runs the player would have been expected to produce, had he been playing for a random team. We then subtract the number of runs our replacement player (the guy from AAA) would have been expected to produce. That leaves us with the player's net contribution to scoring and preventing runs. Then we divide those numbers by the number of runs scored (or prevented) that it takes, on average, to win a ball game. And the resulting number is the player's WAR.

Edit: Here's a rough guide for what's a good/bad WAR total. Major league average players produce about 2 WAR in a full season. Bench players might get 0.5 to 1. The league MVP usually has around 8 (although there's lots of variation on this). Mike Trout has been averaging about 9 per year. The best season from a position player was Ruth's 1923, which was worth 14. The highest season WAR total ever was Tim Keefe's 1883, which was worth 20 WAR, because he pitched more than 600 innings that year. The best post-1920 pitching season was Dwight Gooden's 1985, worth 13 WAR. It usually takes about 60 WAR to make you a serious hall of fame candidate, although plenty of guys have gotten in with less than that, and a few with higher totals have been left out. Babe Ruth has the all-time career record, with 183 WAR. Cy Young, Walter Johnson, and Barry Bonds are next, with figures in the 160 range.