Thursday, June 14, 2012

Cliff Lee and the Problem of Wins

With dominant performances last night by Matt Cain (perfect game and whatnot) and R.A Dickey (America's favorite knuckleballer/adventurist) it seems a little pointless to discuss other pitchers. That being said, it seems as though everyone in the baseball blog-sphere has written about the perfect game. Again and again and again (my personal favorite) and again. Instead of not doing his awesome performance justice, I will instead focus my post on Cliff Lee in the lens of Wins as a stat.

If you are a fan of this blog, you know that we dislike the use of Wins as a counting stat. Sure, getting 300 wins is impressive. 20-win seasons are rare because it is tough to do. But wins are simply too era-dependent to tell a reliable story for a given pitcher. In other words, Cy Young could win 511 games because he played in an era that did not utilize relief pitchers, not to mention starting pitchers threw more often, thereby giving him more opportunities to win.

Wins are also drastically team-dependent. Average pitchers can win a lot more games in a given season if they play on teams with high-scoring offenses. The recent example of this is Ivan Nova, who went 16-4 last year with the Yankees. Nova had an ERA of 3.70, 4.01 FIP and a FIP- of 96. Translation: Nova was aggressively league average in peripherals. It is hard to model what his win-loss record should have been but considering the average ERA and FIP numbers for pitchers with 100+ innings pitched were 3.95 and 3.93 respectively. (Extreme side-note: using a Pythagorean Win Theorem here is not correct. If you, like me at first, attempted to use that to find an expected/corrected win-loss record, please consult me on why that is a wrong application of the PWL tool.) Comparable data shows Ervin Santana with a 4.0 FIP and an 11-12 record. Zach Britton had a FIP of 4.0 also and an 11-11 record. A FIP of 4.1 got Carl Pavano a record of 9-13. All that to say this: win-loss records can tell us something but more often than not, it tells us who plays on a better team.

And then we have poor, poor 2o12 Cliff Lee. Coming into his start on Saturday, he will have a 0-3 record in 10 games pitched. In 70.2 IP, he has a K/9 and BB/9 of 9.42 and 1.66. His FIP is 2.99, while his xFIP is 2.67. These are not stats of someone you would expect to be win-less through   mid-June. A studious review of his game logs show the problem: the Phillies have an anemic offense that has not provided much run support. Notable games include: throwing a 10-inning masterpiece against the Giants (guess who threw a 9-inning two-hitter against him. Anyone? Anyone?) but not getting the decision as the Phillies lost 1-0 in 11 innings. Or, the May 15th game against Houston where he went 8 innings, struck out 10 and gave up one run, yet the Astros scored two runs after Lee left the game to tie the game up and force extra innings. In fact, 6 of his 10 starts have seen Cliff Lee leave the game with the game tied or the Phillies ahead. He lost to the Dodgers 2-1. He has not been chased out of the game before the 6th inning. Put all these facts together and you get: Cliff Lee getting extremely unlucky.

Turning to Run Support numbers, where does Lee  find himself? He has gotten 20 runs in support so far this season. Compared to starting pitchers with enough qualified innings, this places him 3rd to worst, behind Edinson Volquez and Anibal Sanchez. To add insult to injury, those pitchers have won 2 and 3 games respectively while carrying FIPs of 4.31 and 3.21. And yet, friends still try to tell me that Wins are not based somewhat on luck. In case you care, Ricky Romero has gotten 69 runs of support, has a FIP of 4.79 and is 7-1. Yep, wins definitely paint a complete picture of individual talent.

I am very glad that baseball is largely removed from the archaic idea that wins=talent. I am happy to let other commentators/journalists of other sports get into logical fallacies in that regard (hey, it makes me look smarter!) but Cliff Lee stands as a shining, if not unfortunate, example of why Wins need to be viewed as only a part (a very, very, very small part) of the evaluation of pitchers.

1 comment:

  1. [...] to score and their pitchers have been good but not great. Cliff Lee has become a poster-boy for why ‘Wins’ is a lousy category. Sure, it does not help that the Nationals are going bananas (Gwen Stefani reference, ftw!), but [...]

    ReplyDelete