This feels like the sort of post I should not be writing at the beginning of June. Cliff Lee was roughed up for six earned runs over five and one third innings last night, including two home runs surrendered to Nationals second baseman Danny Espinosa. It was Lee’s twelfth start in a season that has been punctuated both by brilliance, such as his 16 strikeout performance against the Braves earlier this month, and by some rough going. Currently, through 80 innings pitched, Lee’s ERA stands at 3.94. Of course, ERA never tells the whole story (or even, sometimes, any significant part of it), so we look to the ERA retrodictors for a different take on Lee’s inputs — things that Lee controls, such as strikeouts, walks, and groundball rate. By their reckoning, Cliff Lee is pitching extremely well right now; his SIERA is 2.81, and his xFIP is even better, at 2.63. It’s tempting, then, to draw parallels to Cole Hamels’ infamous 2009 season, when, through his first 81 innings, he had posted a 4.44 ERA, thanks mostly to an inflated .366 BABIP. Lee, similarly, sports a .344 BABIP, and seems to have been on the wrong end of some fortunately-placed groundballs.
Still, from the viewer perspective, it can sometimes seem as if something is a bit off with Clifton. For one thing, his control has deteriorated a bit. His 5.6% walk rate so far this season is not out of line with his career figure, but it is his highest since 2007, and in the intervening period he’s developed a reputation for being extremely stingy with the free passes. Nobody could have expected him to repeat last year’s walk total of twelve (twelve!), but it was still surprising to seem him surpass that mark on May 16th of this season. It’s difficult to pinpoint the culprit for this. He’s throwing about the same percentage of strikes that he did in 2010, and opponents are swinging less, but not significantly so. He is throwing substantially less first pitch strikes than in 2010, which, one could speculate, could cause him to be more careful with hitters as he falls behind in the count. On hitter’s counts, he does throw his changeup much more than usual, but he doesn’t have a problem locating it; it is a strike 71% of the time, and it generates the most whiffs of any of his pitches save the curveball. And even considering the uptick in walk rate, he’s still well below the league average (8.6%), and it certainly hasn’t perturbed his SIERA or xFIP.
Lee’s homerun rate has risen, but only by one half of one percent, currently sitting at 2.5% (the league average is 2.7%). He is actually allowing less flyballs than either of his previous two seasons, but his homerun/flyball percentage has increased to 10%, not having risen above 6.5% for any of the seasons from 2008-2010. This is not a metric that reflects significant sustainable skill, and there is no reason to assume there is much more to that than some quantum jostlings and park factors. In his time at Citizens Bank Park in 2009 and Rangers Ballpark in Arlington in 2010, his HR/FB percentages were 8.0% and 8.7% respectively.
|Opposing Hitter wOBA|
|Split||Cutter||2 Seam FB||4 Seam FB||Changeup||Curveball|
|Balls in Play||.489||.358||.309||.445||.245|
|All PA ending||.368||.290||.266||.375||.108|
Notably, hitters seem to be mashing Cliff’s cutter. The chart at left shows opposing hitter wOBAs for his different pitch types, for both balls put into play and all plate appearances ending on that pitch (effectively, adding walks and strikeouts to the batted balls). So far, hitters have been most successful against the cutter, managing a .489 wOBA on cutters put in play, and .368 on plate appearances that ended on a cutter. Compare that to last year’s figures, .361 and .271 respectively. There isn’t what you’d call a significant difference between how lefties and righties handle the pitch; on cutters put in play, left-handed batters have posted a wOBA of .515 (20 PA), right-handed batters .477 (45 PA). For all plate appearances ending on a cutter, lefties have a wOBA of .368 (28 PA), righties .369 (68 PA). The cutter-bashing could be concerning, since Lee has ramped up his use of the pitch recently, using it for 22.7% of his pitches so far in 2011, while it constituted just 9.5% of his pitches from 2009-2010. None of this necessarily means anything, though. As you can see, the sample sizes we’re talking about here are not worthy of drawing serious conclusions from, and wOBA is of course subject to Lee’s previously mentioned unfortunate BABIP.
It might seem like lazy saber-bating to just chalk Lee’s outcomes up to luck, but it’s difficult to read the analysis any other way. It is probably the case that our observations are confounded by too many biases. We see Lee striking out hitters at elite rates, and expect elite overall outcomes. We champ at the bit to watch every start by a Phillie ace, and are disappointed when it goes poorly. That is the reality though — every pitcher has bad starts, and bad luck. Hamels managed to have some of the worst luck imaginable for an entire season, and while to some it seemed impossible that fortune could be the only thing at play, the following season validated his defense-and-luck independent indicators. There is nothing in Cliff Lee’s performance right now to suggest that this is anything but a bad run of fortune and perhaps a few poorly chosen cutters. In his last seven games of 2009, Lee had an even worse stretch, running up a 6.13 ERA fueled by a .386 BABIP over about 40 innings. There was nothing wrong with him then, and there isn’t now. After the next several starts, expect Lee to look less like his ERA and more like his SIERA.