O’s By the Numbers: Zach Britton and xFIP

This week in the By The Numbers we’ll look at Zach Britton and Expected Fielding Independent Pitching (xFIP). This will be the second entry in our Advanced Metric Primer Series. The goal of the series is to introduce readers to advanced baseball metrics in a fun and Orioles-centric context.  This week it will also provide an interesting look at what has happened to Zach Britton since his first nine starts. The table below shows Britton’s statistics for his first nine starts and his last ten starts.

Looking at these numbers it seems as if the clock hit midnight when Britton left his May 18th start and he turned into a pumpkin. In fact, the difference is so egregious it seems undebatable: Britton was good, something happened and now he isn’t good anymore. However, xFIP tells us a very different story.

Recall from our previous entry on Fielding Independent Pitching (FIP) that “there is little if any difference among major-league pitchers in their ability to prevent hits on balls hit in the field of play.” FIP offers an ERA like metric that only includes home runs (HR), walks (BB), and strikeouts (K). Formally FIP is: ((13*HR)+((3*BB)-(2*K))/IP) + ~3.2. Expected Fielding Independent Pitching (xFIP) is a regressed version of FIP, developed by Dave Studeman from The Hardball Times. It’s the same as FIP, except it substitutes a pitcher’s Home Run to Fly Ball rate (HR/FB) with the league-average rate. Currently the MLB average HR/FB rate is ~9.1%. This substitution is due to a bevy of work that has shown that pitcher homerun rates are extremely variable over time. A pitcher may allow Home Runs on only 5% of their flyballs one year (Clay Bucholz in 2010), yet then turn around and allow 11% the next year (Clay Bucholz in 2011). Due to this variability Home Runs rates are really difficult to predict; xFIP attempts to correct for this. FIP and xFIP are among the best metrics at predicting a pitcher’s future performance because they measure the pitcher’s ability to prevent runs scored, independent of factors he cannot control. In the case of FIP these factors include the ability of the pitcher’s teammates to play defense. In the case of xFIP this also includes the pitcher’s HR/FB rate. In fact, xFIP has the highest correlation with future ERA of all the pitching metrics.

Zach Britton is a poster child for xFIP; there is a pristine symmetry to his sudden decline into pumpkin-hood. In Britton’s first nine starts he posted an xFIP of 3.95. In his following ten starts he has posted an xFIP of 3.97. Not much has changed about Britton except the results on non-Home Run balls in play. He was very lucky in his first nine starts and very unlucky in his following ten.

Overall, I think Britton will be fine. He isn’t the next coming of a 2004-2008 Brandon Webb but he is a very good ground ball pitcher, a potential #2 or #3 rotation pitcher in the AL East. The biggest problem Britton has right now is the atrocious Orioles defense. Heath Bintliff wrote an excellent article at MASN on this last week. I won’t reiterate all of Heath’s analysis but the take away message is that the defense is really bad except for Wieters and almost no one is hitting well enough to justify how poorly they play in the field. For a groundball pitcher like Britton this problem is exacerbated. Britton relies on turning ground balls into outs and the Orioles’ defense is really bad at accomplishing that task. So even while xFIP tells us that Britton hasn’t changed as a pitcher, we shouldn’t expect him to post an ERA that matches his xFIP until the Orioles improve their defense. For more information on FIP please see our advanced metric primer here.