In a continuation of the free-agency bonanza for players in general and relievers in particular, Matt Guerrier is reportedly going to be signing a three-year, $12 million dollar contract with the Los Angeles Dodgers. Three years for a middle reliever in his thirties seems to be a bit much, but in the case of Guerrier, what is particularly interesting is the great difference between his defense-independent pitching statistics and his ERA/RA.
It is still hard to tell how the market for a marginal win is going to settle. I’ve been assuming five million dollars this offseason, which seems a bit high given the past seasons, but generally is “working” so far. I’ll also assume that amount will increase by ten percent each season. I usually use a generic 0.5 win-per-season decline curve for players, but for relievers I use 0.3 per season. Given these components, $12 million dollars over three years indicates that the Dodgers are paying for Guerrier to be about a one-win “true talent” pitcher in 2011.
Looking at Guerrier’s player page, the initial response is incredulity (assuming a lack of familiarity with Dodgers General Manager Ned Colletti’s body of work). Sure, Guerrier’s ERA the last couple of seasons looks good, and he doesn’t walk many batters. However, he also doesn’t get many strikeouts, his groundball rate isn’t anything special, and he seems to have been “lucky” on balls in play the last two years. In fact, the only time he has gotten close to performing as the 1 WAR pitcher the Dodgers are paying him to be was back in 2007 when he was 0.9 WAR. Over the last three seasons, he has accumulated a grand total of 0.2 WAR. He did have a bit of bad luck with home run/fly ball rates in 2008, but overall his xFIP the last few years doesn’t indicate he’s anywhere near being a 1 WAR pitcher. Moreover, in going from Minnesota to Los Angeles, he’s moving to a park much more friendly to home runs.
That might be that, but there’s another issue lurking with Guerrier: his career ERA (3.38) is significantly better than his career FIP (4.44) and xFIP (4.41). This isn’t the place to lay out the place of FIP (and other defense-/”luck”- independent pitching stats) relative to ERA. I generally agree with Dave Cameron’s two posts on the issue. But while FIP does a lot of work, it has (acknowledged) limitations. Some pitchers do have the skill of “outpitching” their FIP (something I have been guilty of downplaying in the past). The problem is determining which pitchers those are, since it takes a really large sample relative to other skills. This is important in Guerrier’s case because while he’s been barely above replacement the last few years according to FanGraphs WAR, according to Baseball-Reference’s RA-based implementation of WAR (I’ll leave aside the defense issue for the sake of space), Guerrier has actually been quite valuable for a reliever lately, sporting a 2.3 WAR in 2009 and 1.4 in 2010. Of course, he also put up a -0.6 in 2008, perhaps illustrating the point that FIP is more “consistent” from season-to-season because it reflects true talent better than RA/ERA.
Without getting bogged down in a debate about whether FIP is appropriate for WAR because it reflects true talent rather than actual value, in this case we definitely want to know what Guerrier’s true talent is because that is what he is being signed for. It is likely that he has the skill of outperforming his FIP? Over seven seasons he has pitched 472 innings. In one good discussion about the relative merits of FIP, it was suggested that while FIP is better over a one-year sample, that (E)RA is better after three seasons. That might seem to favor Guerrier until you remember that he’s a reliever, and that 472 innings is just over two seasons of what a starter would produce. The reliever small sample/variance issue strikes again!
I don’t have a conclusive answer on the general issues mentioned here or in Guerrier’s case in particular. I can understand why some might be tempted to think that Guerrier is better than his FIP indicates. However, given the reality of the sample size generally needed to indicate such a skill along with Guerrier’s peripherals and age, I think the Dodgers overpaid, with the third year being a particularly bad idea. Sure, other relievers are getting three years, but if Dave Dombrowski wants to jump off of a cliff, does that mean Ned Colletti has to, as well?