All About xFIP

This morning, the Phillies’ official twitter account somewhat bizarrely sent out the briefest of introductions to xFIP.

Considering the organization’s, shall we say, hesitance to embrace sabermetric analysis in the pre-Klentak era, this little pebble thrown into the gaping chasm of the interwebs — even if done so with tongue planted firmly in cheek — came as a bit of a surprise. So, instead of mocking the team for doing what people have been criticizing it for not doing…

and at the risk of explaining something already known to an audience that actively seeks out this site for its sabermetric bent, indulge me in a (very) broad overview of xFIP.

xFIP is Expected Fielding Independent Pitching. It’s an advanced sabermetric statistic that attempts to show what a pitcher’s run prevention capability would be assuming average outcomes on balls in play and a home run per fly ball ratio based on the league average. If you’re familiar with ERA — so, basically everyone — you can think of that statistic as a representation of a pitcher’s run prevention capability on the surface. But really, a pitcher can’t control what happens once the ball leaves his hand. He can’t control whether his teammates catch the ball or not, or whether the fences are deep or shallow relative to “league average,” or plain old dumb luck. So while ERA is the starched, white cotton shirt of pitcher performance measurements, it doesn’t really capture what the pitcher himself actually does to control outcomes.

If you read that as a criticism of ERA, well, you’d be right to do so. I don’t believe that strongly in the power of ERA to tell us about a pitcher’s skills. Having said that, xFIP is hardly chrysopoeia. xFIP is the child of FIP, which also attempts to estimate the pitcher’s ability to prevent runs while controlling for things the pitcher can’t. FIP uses strikeouts, walks, HBPs, and home runs allowed, then assumes average outcomes on balls put in play. The leap that xFIP takes is in assuming league average home run to fly ball ratio, and therein lies the best criticism of the measurement. Some guys — James Shields, for instance — just give up tons of homers because they were born to give up homers (or, you know, because they’re “flyball pitchers” or they “pitch to contact”). If you assume James Shields will have a league average ratio of home runs relative to fly balls allowed, you’re taking away the one thing that makes James Shields, well…James Shields.

I think it’s great that the Phillies are at least trying to show they’re hip to the newfangled gobbledygook. If this helps get more fans interested in analyzing baseball from a different perspective,  that’s a fantastic outcome. Still, most consumers of baseball seem to be comfortable with the old stuff, and that too is fine. It took a decade just to get OBP and SLG into the layman’s lexicon. Considering that, I find it hard to believe that we’ll see xFIP on broadcasts in the place of ERA any time soon.

Leave a Reply

*

8 comments

  1. Romus

    July 19, 2016 02:09 PM

    Why the sudden metric ‘burp’ by the Phillies?
    Assuming the boards may have lit up after Tom McCarthy decided to go into his FIP explanation on last night’s broadcast with co-host Ben Davis.

  2. Bob

    July 19, 2016 02:45 PM

    Adam,

    Are the expected homeruns calculated at league average HR/FB rate or the HR rate of that specific pitcher? For example, James Shields has an average HR/FB rate of 11.9%. This year he is giving up HRs at a 15.3% clip. League average is 12.8%. Are they using the 11.9% or the 12.8%?

  3. Pete

    July 19, 2016 11:06 PM

    “I find it hard to believe that we’ll see xFIP on broadcasts in the place of ERA any time soon.”

    I wouldn’t mind seeing both, but why must xFIP replace ERA? ERA lets the viewer know what actually happened. Do we really want to lose that completely?

    In my mind ERA and xFIP (or FIP) are complimentary. One without the other gives you an incomplete picture.

    • Pete

      July 19, 2016 11:11 PM

      *complementary

    • Romus

      July 20, 2016 07:45 AM

      Pete….agree…..the first generation analytics…..ERA, BA, Fld Pct et al….are the foundations….when you build a house….you want it on a good foundation. Now the SABR is everything from the ground up…maybe even the solar panels on the roof..

  4. Frank S.

    July 20, 2016 11:05 AM

    Or, to put it more simply and generically, one of Bill James’ Ten Commandments:

    “Much of what is perceived to be pitching is, in fact, defense.”

    • Steve

      July 22, 2016 07:45 AM

      In 1 year sample sizes i would mostly agree. There are some pitchers however, who are able to consistantly pitch to weak contact, causing to a higher percentage of batted balls to lead to outs. This becomes evident after multiple years, and xFIP may not always do them justice.

Next ArticleWho Had The More Encouraging Game: Nola or Velasquez?