Metrics don’t always tell same tale, as Cubs’ Travis Wood shows
BY JOHN GROCHOWSKI For Sun-Times Media June 30, 2014 9:20PM
Is Cubs left-hander Travis Wood having a substantially worse season than he did in 2013? It depends on which metric you want to use. | Sean M. Haffey/AP
Updated: August 2, 2014 6:17AM
The Baseball-Reference.com and FanGraphs.com versions of wins above replacement are different calculations, but they usually are reasonably close on position players. The occasional large discrepancies usually come down to defense.
Pitching is another matter. Baseball-Reference’s rWAR uses runs allowed as its base while adjusting for opposition, ballpark effects and team defense. FanGraphs’ fWAR starts with fielding independent pitching. FIP puts the focus on pure pitcher vs. batter events — strikeouts, walks and home runs — and regards everything else as a ball in play that the defense turns into an out or doesn’t.
FIP is a better predictor of future ERA than current ERA is, so FanGraphs tries not to let any of the day-to-day differences in defense bleed into WAR.
The different approaches can lead to radically different WARs. Take Cubs left-hander Travis Wood, who has made 16 starts so far this season, half as many as he made last season. His ERA is 4.52, compared with 3.11 in 2013. If you include unearned runs, as rWAR does, he’s at 4.80 runs per nine innings, compared with 3.29 last season.
His 7-6 record is better than his 9-12 mark of last season, but that’s largely a product of run support. The Cubs, who were averaging 3.79 runs entering play Monday, have given Wood some of their best work. They’ve backed him with five runs per start after averaging only three runs in his starts last season. Wood has contributed to that. His 3.7 runs created per 27 outs is right on team average. A lineup of nine players producing at his level would average 3.7 runs.
Neither version of WAR pays attention to wins, losses or run support; they’re focused on the quality of work on the mound. The Baseball-Reference formula picks up on the increase in runs allowed and calculates Wood’s WAR at 0.5. He’s on a pace where 32 starts would take him to 1.0, far below the 4.4 he posted last season.
FanGraphs’ approach sees a 4.09 FIP that’s weaker than Wood’s 3.89 of last season, but not dramatically so. His walks per nine innings are up from 2.97 to 3.67 and his home runs allowed per nine innings are up marginally from 0.81 to 0.84, but his strikeouts are up from 6.48 to 7.24.
Most of the reason for his higher ERA is that opponents’ batting average on balls in play has soared from .248 to .298, a little above his career mark of .288. There are strong elements of chance in BABiP, so Wood’s 2013 ERA was partially a matter of good fortune, and he has been a touch unlucky this season.
FanGraphs ignores the BABiP portion of his record and focuses on FIP. It calculates a 1.1 WAR, putting him on a pace for a 2.2 that’s a little below his 2.8 of last season.
The Baseball-Reference approach sees a huge change in performance level. But it’s different when the base is FIP, and FanGraphs’ fWAR sees Wood as substantially the same pitcher he was last season.