Is Ubaldo Jimenez Inconsistent?

Over the last week, Ubaldo Jimenez has become a popular conversation piece, as he has been linked to the Yankees and Reds (among other teams) as a potential acquisition before the trade deadline. While Jimenez is generally seen as one of the game’s better young starting pitchers, he’s also developed a bit of a reputation for being inconsistent.

As Buster Olney noted on Twitter yesterday, some people in the game see him as similar to A.J. Burnett, and Olney used the phrase “enigmatic inconsistency” to explain why the Rockies might be willing to move him. After I noted on Twitter that Jimenez’s xFIPs have been insanely consistent over the last three years, Buster and I got into a short discussion about the topic. Obviously, the point he’s making about Jimenez’s consistency from start to start isn’t going to be measured using yearly xFIP, though, so today, I thought I’d try to look at the question differently.

The assertion is that Jimenez’s results fluctuate wildly from start to start, or at least more so than other pitchers of similar quality. While xFIP might be a better way to determine Jimenez’s underlying true talent level, that isn’t the question being asked, so instead, we’ll look at Game Score, which measures the results of a pitcher’s performance each time he takes the mound. It’s not perfect, but it’s a pretty decent proxy for results, and should help us measure what is being asserted about Jimenez’s consistency.

Because Buster specifically asked about Ubaldo’s consistency since the start of last season, that’s the sample we’ll stick with. In 2010, Jimenez had an average Game Score of 60.64 with a standard deviation (a mathematical tool used to measure variance) of 16.43. In 2011, those numbers have both gotten worse, as his average has fallen to 52.11 and his standard deviation has gone up to 16.91 – if you take the standard deviation of all of his games since 2010 started, it’s 16.88. Still, we need to know how those numbers compare to other good pitchers to see whether Jimenez’s results actually are more inconsistent than his peers.

For comparison, I took the three pitchers on either side of him in WAR since the start of 2010 – a pretty impressive group consisting of Felix Hernandez, Clayton Kershaw, Dan Haren, Tim Lincecum, C.J. Wilson, and Cole Hamels. How did those guys compare in terms of variation in game score over the last year and a half? Check out the results below.

Hernandez: 17.13
Kershaw: 16.30
Haren: 22.63 (!)
Lincecum: 17.75
Wilson: 14.40
Hamels: 16.37

Of the pitchers who WAR judges to be similarly valuable since the start of 2010, three have been more consistent than Jimenez and three have been less consistent. In reality, though, there is one guy who stands out at each end of the spectrum (Wilson on one side, Haren on the other), with the other five all being pretty similar. It’d be hard to argue that the differences in standard deviation of Game Score for Felix, Kershaw, Lincecum, Hamels, or Jimenez were all that significant.

In fact, most pitchers settle into that 16-18 range. By deviation in Game Score, Jimenez appears to be perfectly normal. The narrative just doesn’t hold up to scrutiny.

On a year to year basis, Jimenez has been pretty darn consistent. From game to game within those years, he’s been about as consistent as his peers. His velocity has certainly taken a tumble this year, but the idea that it’s made him a demonstrably worse pitcher isn’t really accurate.

It seems like most of the perception of Jimenez’s inconsistency comes from his crazy low run suppression in the first half of last season, when he had a 1.15 ERA through his first 14 starts. Naturally, he couldn’t keep that up, but we shouldn’t confuse regression to the mean with an inherent lack of consistency on Jimenez’s part. ERA just wasn’t doing a very good job of measuring Jimenez’s true talent level, and given a larger sample size, his good fortune ran out. It happens, but it doesn’t make Jimenez inconsistent, and it certainly doesn’t make him A.J. Burnett.