Understanding WAR
In the last two weeks, we have discussed the best way to measure a player in terms of both offensive (wRAA) and defensive (UZR) production. Those two statistics express player value in terms of runs above average. We can add them together to find their overall Wins Above Replacement, or WAR. WAR is the most all-inclusive statistic to evaluate a player. The idea of wins above replacement is if the player got injured and the team had to replace him with an average player, how much value would the team be losing. Calculating WAR is simple: just add the wRAA and UZR values together, add in a positional adjustment, convert the numbers from league average to replacement level, and convert the run value to wins (10 runs = 1 win). For pitchers, WAR is based solely on FIP. FIP is turned into runs from based on the number of innings pitched, then to replacement level, then from runs to wins.
A full-time position player is usually worth around 2 WAR. Bench players are usually less than 1 WAR. Average starting pitchers are worth around 2 WAR and relief pitchers rarely reach 1 WAR. Now for some rankings.
From 1871 until today, the career leaders in WAR among hitters are Babe Ruth (178), Barry Bonds (168), Ty Cobb (164), Willie Mays (163) and Hank Aaron (150). Taking just this season, the best players in baseball have been Jose Bautista (4.5), Jose Reyes, Adrian Gonzalez, Curtis Granderson and Matt Kemp. On the other side of the spectrum, the worst players in baseball have been Raul Ibanez (-1.2), Juan Pierre, Chone Figgins, Adam Dunn and Aubrey Huff.
Looking at the pitchers now, the career leaders are Roger Clemens (145.5), Greg Maddux (120.6), Randy Johnson (114.8), Nolan Ryan (92.3) and Pedro Martinez (89.4). In 2011, the most valuable pitchers have been Roy Halladay (4.0), Jered Weaver, Cole Hamels, Dan Haren and CC Sabathia. The least valuable pitchers have been Brett Myers (-0.3), Wade Davis, Chris Volstad, Bronson Arroyo and Colby Lewis.
None of these stats should surprise you. These players are valued as high or low as they are for good reason and this statistic just quantifies what should be noticed by the naked eye. We use all of these statistics to help further analyze a player and predict his performance for the remainder of the season. It’s the best way we have of determining whether a hot streak is merely that or if we have a Jose Bautista in the making. For a few weeks, we discussed how Indians pitcher Josh Tomlin was, according to advanced statistics, greatly overachieving so far in 2011 with his 2.74 ERA, 4.43 FIP, .197 BABIP and 85.1% LOB. Today, Tomlin’s LOB% and BABIP have evened out to a more reasonable 70.2% and .259 respectively, and his FIP has dropped only a little to 3.89. As a result, that nice sub-3.00 ERA is now up to a more reasonable 3.93. I was able to predict that last month from using only the power of these statistics. This will be the last post in the sabermetric series. I hope you enjoyed them.
