Saturday, May 14, 2011
Searching the low SIERAs
By Christina Kahrl
James Shields’ big turnaround has been remarkable enough, but various interpretive metrics helped anticipate that better times were coming for the Rays’ workhorse after last year’s ill fortune. Shields’ league-leading 34 homers allowed were part of the problem, the product of a career-high ratio of home runs to fly balls of 10 percent.
James Shields has a 4-1 record and a 2.08 ERA so far this season.
As Jerry Crasnick noted, a big part of Shields’ ill fortunes were the product of defensive execution -- or the lack of it -- in his starts, producing his MLB-high batting average on balls in play (BABIP). At the same time, Shields was delivering an excellent strikeout rate, getting K’s on almost 21 percent of all plate appearances.
Those kinds of factors go into an interpretive metric developed by Matt Swartz and Eric Seidman published at Baseball Prospectus that does just a tiny bit better than xFIP in anticipating where a pitcher’s ERA should be headed. They invented Skill-Interactive ERA, which estimates ERA through walk, strikeout, and groundball rates. (That’s the short explanation, the full discussion ran in five parts last year.)
If, by looking at last year’s SIERAs relative to actual ERA, you had come up with a list of the 10 biggest differences between a SIERA lower than an ERA, Shields would have ranked third among all pitchers with 90 or more innings pitched -- an interpretive suggestion that he was due for better things in 2011. Shields’ 3.57 SIERA in 2010 gave you a better idea of what he was capable of doing going forward than did his 5.18 ERA, providing you with a better yardstick of where his actual skill level was.
With that in mind, who were the other nine pitchers on this list of expected success from looking at these low SIERA underachievers of 2010, and how are they faring this year?
Obviously, that bounce-back trio of Beckett, Lohse and Shields makes for an impressive group of guys who were supposed to do better, and have. You can add Nolasco and Correia to the list of happy comebacks so far, giving us a quintet that makes for a fairly solid argument that SIERA is a reasonable tool for anticipating future performance. Move into the 11th through 20th biggest differences between a pitcher’s SIERA and his ERA from 2010, and you add several more names from this spring’s most improved starting pitchers: Bud Norris, Jason Hammel, Justin Masterson, Aaron Harang and Chris Narveson.
Josh Beckett was among the top SIERA bounceback candidates for 2011.
However, there are also the guys not doing well, but they generally all come with reasonable explanations. Even if you allow for SIERA’s suggestion that he was due to come back some way, Bannister’s SIERA was still pushing 5.00, not exactly someone you race to go out and get. He signed with Yomiuri to play in Japan, but came back stateside after the March earthquake and is currently on the NPB’s restricted list as a result.
Moving down the list, Morrow is coming back from an early-season elbow problem that put him on the DL, while Parra is still on the DL. Like Bannister, Rowland-Smith’s difference between his ERA and his SIERA was big, but still wound up suggesting he wouldn’t be someone you’d want; sending him back to the PCL if you were the Astros seems eminently reasonable. The closest thing to an outright miss so far is Ely; he was crowded out of the Dodgers’ rotation, although he didn’t help himself by getting bombed in his one start, forcing him to make that wrong turn back to Albuquerque.
Where statheads don’t do themselves any favors when talking about the results of these kinds of metrics and evaluating what went wrong for these pitchers in the past is by haphazardly ascribing this sort of turnaround to "luck." The data for all of these pitchers represents symptoms of a failure to execute -- failure from the pitchers themselves, and/or their fielders -- and not just random chance. Beckett and Lohse were dealing with injuries; now they’re healthy and able to deliver something closer to their career numbers.
Shields wasn’t going to be able to just throw his glove back out there this spring and expect that things would just get better because the data said it would -- that his luck would simply turn.
Instead, as Crasnick’s piece reveals, Shields worked with his team to proactively change his mix of pitches, using his curve more often as well as earlier in his batter/pitcher confrontations. By working actively to change their fortunes, these are pitchers creating a good measure of their own success.