Yesterday’s debate over my Andy Pettitte piece was for the most part informative and conversational. Sure, there are some people that think the W-L record is an important way to figure out “who’s good at bearing down”, but for the most part we were able to have a good discussion about what stats were the most appropriate for evaluating who should be in the All-Star Game. Prompted by one commenter who questioned my inclusion of James Shields, given his 4.55 ERA and his 1.41 WHIP, I promised a more thorough examination of the pitchers that I considered to be the best. Upon further review, Shields missed the cut. To the list we go!
1. Francisco Liriano: 2.90 ERA, 2.11 FIP, 2.95 xFIP. Has struck out 9.71 K/9 batters per nine innings, and has a 4.14 K/BB ratio. He’s been dominant, and his peripherals suggest his start is as good as it seems at first glance. You could make the case that Cliff Lee has been as good as Liriano, but I’ll take the flashier K rate any day. Liriano gets my #1 slot.
2. Cliff Lee. 2.88 ERA, 1.93 FIP, 2.95 xFIP. Cliff Lee has struck out 7.86 batters per nine, and has an incredible 15.00 K/BB ratio in 68 innings. You read that correctly. In 68 innings, Cliff Lee has struck out 60 batters and walked 4, and his 0.52 BB/9 rate is the best among American League starters. It’s true that he missed most of April, but he’s made 9 starts already and has pitched almost as many innings as Phil Hughes. It’s been a masterful start for Lee.
3. Jered Weaver. 3.29 ERA, 3.20 FIP, 3.48 xFIP. Weaver has quietly put together an excellent 87 innings, striking out 9.86 batters per nine innings and walking only 2.36 batters per nine innings. His 4.17 K/BB ratio is fourth best, behind Cliff Lee, Zack Greinke and James Shields. In the comments yesterday, bexy noted that Weaver will be the hometown kid at the All-Star Game in Anaheim and should be considered for the start. While Liriano and Lee have been better than Weaver, I’d agree with that sentiment.
4. Phil Hughes. 3.11 ERA, 2.90 FIP, 3.70 xFIP. The xFIP is higher because of an abnormally low HR rate. That said, Hughes has still been dominant with an 8.84 K/9 rate, a 2.66 BB/9 rate and a 3.36 K/BB ratio. Hughes has been the best pitcher on the Yankees. I wish it was 2011 already and he had no innings cap.
5. Jon Lester. My preseason AL Cy Young pick has pitched very well after a rough start. He has a 3.18 ERA and his FIP is 3.09, fourth-best in the American League. Over 85 innings he has posted a K/9 of 9.42 and a BB/9 of 3.81, which leaves him with a K/BB ratio of 2.47.
6. Ricky Romero. 3.29 ERA, 3.17 FIP, 3.23 FIP. Gotta love it when those three line up neatly, it makes my job easier. Over 90 innings, Romero has a K/9 of 9.07 and a K/BB of 3.39, which leaves him with a K/BB ratio of 2.68. His BABIP, strand rates and other peripherals all seem largely in line with past norms, which can only lead to the conclusion that Romero’s great start is no illusion. Give the man his dues.
7. Justin Verlander. In 13 starts and 86 innings, Verlander has a 3.56 ERA, a 3.28 FIP and a 4.15 xFIP. The xFIP is higher because of an artificially low HR/FB% of 4.9%. He’s striking out batters at a rate of 7.74 per nine and walking batters at a 3.24 BB/9 clip, which leaves him with a K/BB ratio of 2.39.8.
8. Felix Hernandez. 3.61 ERA, 3.67 FIP, 3.61 xFIP. He’s striking out 8.38 batters per nine innings and walking 3.14 per nine, slightly up from his 2009 mark of 2.68. It’s another solid start for King Felix, and he’d be my final selection for the All-Star Game.
The Just-Missed List
9. Shaun Marcum. Another underrated Blue Jay, Marcum has a 6.53 K/9 and a fifth-best BB/9 of 1.75. His 3.72 K/BB ratio is 7th best, right behind Carl Pavano (gag). In 14 starts, Marcum has logged 92 innings and has put together a 3.31 ERA, a 3.48 FIP and an xFIP of 4.06. His performance is impressive for someone coming off Tommy John surgery, in particular his ability to control the walks. He’s my “last out” All-Star.
10. John Danks. Danks has a 3.27 ERA and an eight-best 3.31 FIP. Like Verlander, Danks’ xFIP (4.17) is higher, thanks to a low HR/FB ratio is 4.6%. Despite that, Danks has K/9 and BB/9 rates of 7.13 and3.16, respectively, which gives him K/BB ratio of 2.26.
11. Andy Pettitte. We went through this yesterday, but to summarize: 2.46 ERA, 3.75 FIP, 4.09 xFIP. High strand rates and a low BABIP suggest some luck has come into play, and Pettitte has a K rate of 6.16 per nine, a BB rate of 2.46 per nine, and a K/BB ratio of 2.5, good for 24th best in the American League. I’m ranking him higher than Buchholz, despite having a higher FIP, because of his ability to limit the walks.
12. Clay Buchholz. This piece on FanGraphs does a good job of analyzing Buchholz’s season thus far. Buchholz has an ERA of 2.67, a FIP of 3.42 and an xFIP of 4.20. His K rate is down from past seasons at 6.19, and his BB/9 is 3.63, leaving him with a somewhat mediocre K/BB ratio of 1.71.
13. James Shields. Over 85 innings, Shields has posted a 4.55 ERA, a 3.89 FIP and a 3.39 xFIP. It’s worth noting that his total number of earned runs allowed is 43, and that 17 of them came in two disastrous starts against the White Sox and Marlins. This isn’t to say that those two outings shouldn’t count against him. They should. It merely contextualizes his year slightly. Shields has a 8.68 K/9 and a 2.01 BB/9, and his 4.32 K/BB ratio is third-best in the American League. Despite that, I can’t get past his ugly ERA. I overrated Shields the first time around, and he’s fallen down my list.
14. Doug Fister. Fister has the second-best ERA in the American League with 2.45. His FIP stands at 3.66 and his xFIP is 4.23. Fister also has the fifth worst K/9 rate (4.13) amongst qualified AL starters, but his BB/9 (1.29) is second-best, behind Cliff Lee. He’s an anomaly, and while his BABIP (.240) suggests that some correction is coming, he’s managed to get results thus far in 2010. To be honest, I can’t figure out where to put him on the 9-15 list because his performance is so odd. Do you really want someone with a K/9 lower than his xFIP on the All-Star team? Maybe he should earn a spot because of that. I can’t decide.
15. David Price. Price is the cause celebre of ERA and W-L enthusiasts, thanks to his 2.31 ERA and his 10-2 record. Far be it from me to rain on his parade, but I’m not overly enthused about everything else. His FIP is 3.87 and his strikeout and walk rates (6.72 and 3.68, respectively) leave him with a K/BB of 1.83. This is good, not great.
Honorable Mentions: Zack Greinke, Jeff Niemann, Jason Vargas, Carl Pavano and Colby Lewis.
At the end of the day, I understand the argument posed by Moshe in the comments of my post about Pettitte. ERA shouldn’t be ignored, as it stands as an historical record of what actually occurred. Since the All-Star Game is ostensibly about celebrating and showcasing the best players of the first half of the year, it would be unwise to simply look at projection stats like xFIP and select the players that we expect should have done the best. I buy this, and it’s why I chose Price over Greinke for the last spot despite the fact that they have identical FIPs.
At the same time, we should not conflate stats like FIP with projection. Like ERA and WHIP, FIP shows how well a pitcher has done at controlling things like strikeouts, walks and home runs. It certainly has weaknesses. xFIP was created, among other reasons, when later research demonstrated that pitcher’s don’t have the ability to control their HR/FB rate. Regardless, FIP shows is something, and it shows us something real. Despite what some of the commenters yesterday suggested, ERA is no more “real” than FIP. It has simply been around longer and is ingrained in our consciousness. But simply because we’re used to it doesn’t make it better, and it doesn’t mean our analysis should stop there.