skyking162

September’s the most frustrating time of the year. Sure, there are a lot of fun things going on — football begins, pennant races are on fire, and not owning an air conditioner stops being painful — but it’s also MLB Awards Speculation Season, and, to anyone with a brain, the non-stop idiotic banter about why Jose Reyes is the Mets’ MVP will make you wish the internet doesn’t exist.

In past years I’ve usually written an article about how I no longer care who actually wins the awards, because the criteria are so poorly defined. Instead, this year I’m going to be productive and walk through the discussion for the four major awards (MVP and Cy Young in both the AL and NL) — not because I care who comes out on top, but because the methods used to answer questions are often more interesting than the answer.

First up, the AL Cy Young. It’s obviously Johan Santana. Oh, wait, I was going to discuss methods. Fine…

Comparing starters to starters can be relatively straight forward. We don’t have to worry about clutchiness or leverage, and in 2006 I think we can finally ignore Wins. The decision boils down to innings pitched and the quality of those innings pitched — some sort of ERA-style rate stat. Which one to pick can get interesting, however.

ERA has its flaws and there are any number of alternatives that attempt to fix those flaws. If you don’t like that ERA ignores unearned runs (which are still somewhat the pitcher’s responsibility), Run Average (RA) is a good alternative. If you don’t like that the events given up by a pitcher can combine to yield a fortunately low or unfortunately high number of runs, Component ERA (ERC) is a good alternative. If you think that pitchers benefit or suffer from the fielders behind them, you’ll want something more defense-independent (DIPS or FIP). If you think that homeruns allowed per flyball is flukey, then xFIP is the way to go.

Somewhere between ERA and xFIP we crossed the line from “what actually happened” to “the amount of skill the pitcher demonstrated”. That’s the difference between value and ability. Like the discussion of whether to credit Big Papi’s walk-off homers with extra value beyond a normal homerun, we need to decide whether to credit Jeremy Bonderman with an above-average ability to keep flyballs in the park. Well, we’d have to make that decision if we actually cared about who the Cy Young is. Personally, I think leaving each person to create their own definition of Cy Young isn’t a bad thing. Let’s see the leaderboards for runs prevented, using a few different runs-allowed metrics (I used 6.00 as replacement level for all the metrics, but that’s another factor that’s open to interpretation):

ERA Runs Prevented
72 Santana
64 Halladay
54 Verlander
52 Lackey
50 Liriano
49 Wang
49 Sabathia
48 Zito
45 Haren
45 Escobar

Nothing out of line here. Santana’s a win better than Halladay who’s a win better than the rest. Escobar’s avoided everyone’s radar this season.

RA Runs Prevented
69 Santana
60 Halladay
51 Verlander
48 Liriano
45 Wang
43 Zito
41 Haren
40 Lackey
40 Weaver
40 Schilling

Like Liriano, Weaver’s been a very valuable pitcher this year even with only half a year of adding value. If you don’t pitch as much as the rest, you better be lights out, and those two have been. Schilling also sneaks onto the list thanks to giving up zero unearned runs this year.

FIP Runs Prevented
67 Santana
62 Bonderman
55 Halladay
52 Lackey
49 Schilling
48 Sabathia
48 Mussina
46 Haren
45 Liriano
45 Escobar

Wow, Bonderman comes from off the list to grab second place behind Santana. Why? Because his peripherals are fantastic: 8.9 K/G, 2.7 BB/G, and a measly .7 HR/G. While Bonderman’s ERA is over 4.00, he actually leads the league with a 3.01 FIP. Verlander and Wang fall a bit due to unspectacular strikeouts rates, something that may be unfair. Zito’s nowhere to be seen — his FIP is a Washburn-like 4.85. And not to ruin the surprise, but his xFIP is even more awful — 5.34. Zito’s a flyball pitcher giving up fewer homeruns than you’d expect and benefitting from a ballpark and defense that limit hits on balls-in-play. If MLB had sign-and-trade rules like the NBA, Billy Beane would make a killing this off-season. Is that Mike Mussina turning in the just the third underrated Yankees season ever? Yikes.

xFIP Runs Prevented
62 Santana
58 Halladay
53 Bonderman
49 Schilling
47 Liriano
45 Hernandez
44 Haren
42 Mussina
41 Sabathia
39 Millwood

Bonderman drops below Halladay, as xFIP thinks Jeremy’s stinginess with homeruns isn’t all skill. Adding their names to the list are Felix Hernandez and Kevin Millwood, two pitchers who have given up a lot of (undeserved?) homeruns. xFIP really likes Felix’s groundball tendencies. He’s one guy who’s likely to see his ERA drop next season, either because those homeruns are flukey or because he’ll become better at preventing homeruns with a year of MLB experience.

Which list is the best? I think that depends on your question. If you want to know which pitcher actually prevented the most runs this year, use ERA or RA (or a combination). If you want to know which pitcher prevented the most runs assuming a neutral defense, use FIP. If you want to know which pitcher prevented the most runs based on his exhibited skills, you might prefer xFIP. Personally, I’d define the best pitcher as the one I’d most want on my team for the past season if I knew the skillset he was going to bring to the table. As that decision doesn’t take place in a vacuum (if you had the Twins’ infield defense, Chien-Ming Wang becomes less intriguing), there’s not always an exact order. But I’d probably use xFIP runs prevented, and I’d definitely pick Santana first.

However, a simple IP*(repERA-actERA) calculation might not suit your analytic fancy, especially if you think relievers deserve credit for pitching in higher-leverage situations. Dealing with AL relievers will be the topic of the next post.

If you want to play around with pitchers’ runs-allowed rates, head over to the Hardball Times Stats. They have neat ways of sorting players, including differences between the various runs-allowed metrics.

Popularity: 2% [?]

This entry was posted on Friday, September 8th, 2006 at 2:09 pm and is filed under Baseball. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.