Who Is the Real Travis Wood?
Travis Wood was one of the few bright spots for the 2013 Chicago Cubs. Wood was the Cubs’ lone All Star in 2013, when he posted a 3.11 ERA in 200 innings over 32 starts, good for 4.4 rWAR, Baseball-Reference’s WAR calculation. Wood was also the Cubs’ most consistently good pitcher by a significant margin, particularly after the trades of Scot Feldman and Matt Garza. This led to some talking about Wood as a potential rotation cornerstone, maybe a solid 2 behind whomever the Cubs find to be their ace in 2015 or 2016. It was hard to find a Cubs fan who didn’t think Wood was at least a very good 3 in most rotations.
Unfortunately, 2014 has not been as kind to Wood. In 116 innings over 20 starts he pitched to a 5.12 ERA, a below replacement level performance (-0.4 rWAR on the mound). So which is the real Travis Wood? The one who looked like a potential near piece to build a rotation around in 2013, or the one who has provided more value in the batter’s box than on the mound and looks like he perhaps should face competition heading into 2015 to earn a spot in the starting rotation?
The answer is neither. A pitcher’s ERA correlates highly with 6 peripheral statistics: strikeout rate, walk rate, ground ball rate, home runs per fly ball, bating average on balls in play (BABIP), and left on base rate (LOB%). A pitcher with high strikeout, ground ball and left on base rates, along with low walk and HR/FB rates and a low BABIP will have a very, very good ERA. But the ability of these statistics to provide information regarding what to expect from a pitcher going forward varies greatly.
Strikeout. walk, and ground ball rates are the most predictive of these peripherals. Unless a pitcher’s stuff improves or declines, or if a pitcher meaningfully changes the way he pitches (which few do successfully), the variation in these peripherals tends to stay fairly small on a season to season basis. The predictability of HR/FB seems to depend on the pitcher. Some pitchers have an ability to control whether their fly balls leave the park, although a majority do not. BABIP and LOB% tend to not be predictive season to season, although some pitchers do have lower natural BABIPs than others.
The sabermetrics community developed two statistics to use the more predictable peripherals to determine, given an average BABIP and left on base rate and a neural ballpark environment and defense, what a pitcher’s expected ERA is. The first, FIP (Fielding Independent Pitching) looks at K/9, BB/9, and HR/FB. FanGraphs later developed xFIP, on the basis that most pitchers exert greater control on whether ball are hit on the ground or in the air than they do on whether the balls that are hit in the air clear the outfield fence or not. The question of whether FIP or xFIP is more predictive for any particular pitcher rests on the question of if the pitcher has that ability to induce fly balls that do not become home runs at a reduced rate, or if he is allows balls to fly out of the ballpark with abandon.
In 2013, Wood’s FIP was 3.89, and his xFIP was 4.50. So Wood outperformed his FIP by more than 3/4 of a run, and outperformed his xFIP by over 1.25 runs. As Wood’s BB/9 is a a full walk higher in 2014 than it was in 2013, those numbers have gone up in 2014, but not drastically so: his FIP is 4.29 and his xFIP is 4.69.
So what’s causing the 2 run difference between Wood’s 2013 ERA and his 2014 ERA? BABIP and LOB%. In 2013, Wood posted well batter than average in both of those peripheral statistics, putting up a .248 BABIP (league average hovers around .300) and stranding 77.4% of base runners (league average hovers around 70%, with a few percentage points making a big difference in ERA).
To this point in 2014 Wood has a .315 BABIP and stranded just 66.2% of men who get on base. With 7% more of the runners he faces getting on base, and 11% more of the runners who get on base reaching home plate, Wood’s ERA has ballooned.
It’s not all bad news, though. Wood continues to show an ability, like many left handed pitchers who induce more fly balls than average, to limit the number of home runs he gives up on fly balls, consistently keeping that number between 6.3% and 7.4% (league average is around 10%, with, again, small changes resulting in big ERA differences). This means that Wood’s FIP, which for his career is about 1/3 of a run lower than his xFIP, is a better indication Wood’s true talent.
In other words, Wood’s expected ERA is between the high 3s and low 4s, not the low 3s ERA he posted in 2013 or the low 5s ERA he currently has.. That turns Wood into a solid, innings eating, 3/4 type in a starting rotation… just like the Cubs thought they were getting when they signed Edwin Jackson prior to 2013. But he’s not a 2, and no one should expect him to be.
So what does this mean for the long term with Wood? Well, if he is willing to sign a long term extension well below market value for a pitcher like him (say $10 million per year in the free agent seasons he would give up), it would like still be worth it for the Cubs to do so. But if he’s looking for a big payday, the Cubs should go year to year with him through arbitration and let him leave via free agency or trade him if they have better options.