Matt Cain ignores xFIP, again and again

Matt Cain is pretty darn good, I mean, right?

Bill at Crashburn Alley doesn't think so. He recently said Cain "... may be one of the most overrated pitchers in baseball (though you can’t fault him for pitching to the strengths of his home ballpark)..." Is this true? It wasn't just Crashburn Alley that brought about this post; Cain's worth is bandied about constantly, at least in my experience.

If we're to simply look at his FIP and xFIP, a lot of people would say Cain has been lucky. Really lucky. Boy, he's been lucky with that home rune rate, it must be because he pitches at AT&T. Boy, he's been lucky with his batting average on balls in play (BABiP). It must be because of spacious AT&T.

Cain made his debut in 2005, but his first full season didn't come until 2006, when he'd throw 190.2 innings. He'd arrived. All said and done, he posted a 3.5 WAR (Wins Above Replacement) and finished with a 4.15 ERA and 13 wins. His FIP (fielding independent pitching) was 3.96 but his xFIP (adjusted FIP that pushes the home run per fly ball (HR/FB) rate to league average) was just 4.54. Perhaps it was just beginners luck.


In 2007 he was unlucky in that the Giants' offense made the Bad News Bears look like the '27 Yankees. He would go 7-16 with a 3.65 ERA, but his FIP was 3.78 and his xFIP way higher at 4.61. ERA-wise, lucky again.

In 2008, Cain continued to pitch quality innings despite the complete and utter lack of run support from his team. He would finish up with another 200+ innings and a 3.76 ERA, which would go along with 8 of those all-telling wins (8-14).

Cain just continued that wonderful magic in 2009, besting his xFIP (4.22) and FIP (3.89) with a career best 2.89 ERA. And this time it came with some wins as he went 14-8 while the Giants returned to respectability with an 88-74 record. This string of luck had to end in 2010, didn't it?

It did not. Cain would again post an extremely quality ERA with a 3.14 in 2010 while his FIP of 3.65 was good and his xFIP of 4.19 pretty unimpressive.

FIP is no friend to Cain, and xFIP is his mortal enemy. Fangraphs bases it's pitching WAR off of FIP, and Cain has been penalized in their rating each season because of it. While the vast maority of pitchers don't control how many and how few of the balls that are hit into the air off them go over the fence, all pitchers are not created equal. It is his ability control where those balls end up (either in the seats or a glove) that Cain is exceptional.

Cain is an extreme fly ball pitcher, and usually that's not a good thing. But in his case, it is.

In 2006 the league average fly ball percentage (FB%) was 28.20, 36.03 if you include infield pop flies. Cain stood at 31.9 and 45.5, well above average. The average pitcher gave up a home run on 13.77% of fly balls, 10.78% if you include infield pop flies. Cain finished at 10.34% and 7.26%.

In 2007 he league average FB% was 27.97, 35.77 if you include infield pop flies. Cain stood at 34.2 and 42.7. The average pitcher gave up a home run on 12.79% of fly balls, 10% if you include infield pop flies. Cain finished at 7% and 5.6%.

MLB 2008 fly ball averages: 28.01 / 35.73
MLB 2008 average HR/FB rates: 12.72 / 9.97

Cain 2008 fly balls: 34.7 / 44.2
Cain 2008 HR/FB rates: 8.41 / 6.6

Is anyone seeing a trend here? No?

MLB 2009 fly ball averages: 28.51 / 36.14
MLB 2009 average HR/FB rates: 13.06 / 10.30

Cain 2009 fly balls: 33 / 41.3
Cain 2009 HR/FB rates: 10.43 / 8.33

The march goes on...

MLB 2010 fly ball averages: 28.04 / 35.53
MLB 2010 average HR/FB rates: 14.33 / 11.30

Cain 2010 fly balls: 34.1 / 45.5
Cain 2010 HR/FB rates: 9.82 / 7.36

Alas, it must be because his home yard is so incredible at converting fly balls into outs, rather than four-baggers. But in 2010, his HR/FB on the road was 7.5% to 7.1% at home. Wash. In 2009, his away HR/FB was 7.8% to a higher 8.8% at home. Funny, that. In 2008, it was 7.1 away and 6.7 in San Francisco. Wash. How about 2007? Away yielded a rate of 4.5% HR/FB, while he finished with a sizably larger 6.3% at home. Peculiar. And in 2006 we finally find a major discrepancy, as his rate of 11% on the road versus just 4.2% at home was significant. As for his career, Cain has produced a HR/FB of 7.4% while away and 6.7% in his home yard, a delta of roughly 10%. But in two of the past five seasons, his rate at home has been higher while in two others, the difference was negligible.

What if Cain were to have the same HR/FB percentage at home as he has on the road? What would that do? Knowing that will help some to strip all that benefit from his home park, won’t it? Batters have hit 1,415 fly balls off of Cain throughout his career. We’ve established His HR/FB percentage is 6.7% at home and 7.4% away – it comes to 7.0% cumulative. If we multiply 7.0% by 1,415, we get 99.05, which is just right because Cain his given up 99 home runs in his career. To neutralize all that good fortune AT&T has afforded him, let’s multiply 1,415 by 7.4%, which will assume he carries his HR/FB percentage wherever he goes, which is a better representation of his true skill. What do we get? We get 104.71 home runs, or a difference of 5.66 home runs. Let’s call it 6. That’s approximately one per season. Now I’m not expert, but that doesn’t seem like an awful lot.

But let’s find out. What’s the worst case scenario, a grand slam? That’s four runs (and an out he didn’t get). Now this isn’t perfect, but let’s add 4 runs to each of Cain’s six seasons, i.e. one grand slam per season, and see what that does to his as of right now overrated career. You’ll have to trust me on this, but his career ERA is 3.45 now, and with the 4 runs per year it ascends to 3.65. Again, I’m no expert, but that doesn’t seem like an awful lot. It seems to me AT&T has reduced Cain’s ERA by up to (but probably a fair amount less than) 20 points in ERA.

Season in and season out, Cain gives up a lesser percentage of home runs on fly balls than the average pitcher, both at home and away. Why? I can't say with confidence, but Dave Pinto at Baseball Musings has come up with a theory that's as good as any I've read: he surmised that Cain's fastball is dropping less than the hitter expects, resulting in the ball being struck somewhere below the center of the ball, and either going straight into the air (as his 12.9% career infield fly ball percentage (IFFB%) might suggest) or at least somewhere inside the park*.

*Quick aside that I had not considered at any point until NOW: wouldn't Oakland be ideal for him, given the enormous foul ground territory? *Checks*  In a small sample, granted, he's pitched to a 1.16 ERA in 3 starts, his best ERA at any park where he's had two or more starts. This could be coincidence, but I'm intrigued.

And it was those heat charts and Dave's title: Rising Cain, that got me to thinking. There was once a pitcher who was renowned for throwing a rising fastball. We all know that physics has proven no such thing is possible with an overhand throwing motion, and Sandy Koufax had one if one ever existed, but it at least appeared to do so. He also threw a devestating, 12 to 6 curveball, but I'll have to leave that for another time.

I won't pretend Cain is a peer to Koufax, because he's not. Koufax was a devastating strikeout pitcher. But, with that he also possessed an ability to keep down his home run rate despite being an extreme fly ball pitcher, not unlike Cain.

Koufax was a bonus baby back in the 50's, and because of this the Dodgers had to keep him on the major league roster to protect their investment. Unfortunately, Koufax had no idea where the ball was going and struggled through 1960. But in 1961, something clicked as he improved his control by walking just 3.4 batters per nine innings when he'd regularly walked anywhere from 4.4 to 6.0 prior to that season, while also striking out 9.5 batters per nine along the way to a league-leading 269. From 1961 until 1966, Koufax would unleash one of the most dominate pitching stretches in the history of baseball, a furious and complete assault on Major League hitters.

If '61 was an indication, '62 was something of a revelation. It was with this new found control that Koufax was also able to induce harmless fly balls. If we include infield pop flies with the fly ball data, Koufax allowed 6.7% of his fly balls to go for home runs, whereas the league average was 9.88%. He did better in 1963 (5.34% versus 9.25% ), and even better than that in '64 (5.14% versus 9.70%). He again beat the average in 1965 while yielding 8.0% to the league average of 9.6% and one last time in 1966 before going off into the sunset (5.16% versus 9.61%).

There was no such thing as xFIP when Koufax pitched. But had there been, would we have called overrated because, separate from his batter-whiffing exploits, he could really keep the ball in the park despite being a fly ball pitcher? After all, he too pitched in an excellent park for pitchers, Dodgers Stadium. I sincerely hope we would not have.

Let’s give credit where credit is due. Luck is often the culprit, but I see no reason to assign it any (or much) credit here. When in 2006 he did this, we might have hypothesized it was lucky. We might have said the same in 2007, or even 2008. But when we have 1,100 innings spanning 6 six seasons to go on, I think we ought to be careful. I love FIP and xFIP as much as the next guy, but this seems a clear case where we’ve abused their worth.

If we strip Cain the right of claiming he possesses a unique skill to produce harmless fly balls, we’ve done him a great disservice, because believe it or not, this is more or less the only thing that separates him from the pack.  But if you must, you better also prepare to take a shell from a turtle, and that would be evil.

All of my data is compliments of Baseball Prospectus, and I’d like to extend a hat tip to Otis Anderson of Bay City Ball for sending me there. Once I'd pulled the hard data, it only took staring at spreadsheets until my eyes were blurry to get the information I wanted.