On DIPS Theory, Defense, and the 2008 Milwaukee Brewers

Right Field BleachersCorrespondent IMarch 3, 2009

In 1999, Voros McCracken developed what he called Defense-Independent Pitching Statistics, or DIPS. DIPS theory is based on the idea that a pitcher has very little control on what happens to a ball once it is put in play—then turning that ball in play into an out becomes the responsibility of the defense.

This new way of thinking has had an extraordinary impact on the way the sabermetric community views pitching. Most notably, DIPS theory has led to the development of FIP, currently the metric used by fangraphs.com, and tRA, used at statcorner.com, which has been quickly catching on and has the potential to become even more valuable as bugs are worked out.

I’d like to turn the attention of this post now to FIP, or Fielding Independent Pitching. I feel like I’ve been referencing this stat quite a bit lately and that now would be a good time to go a little bit more in depth. FIP is the essence of DIPS theory.


FIP looks at three components of pitching statistics: HR rate, K rate, and BB rate. None of these three outcomes include balls in play, and as such they are mostly controlled by the pitcher. The formula for FIP is as follows


FIP = (13*HR+3*BB-2*K)/IP + N


Where N is some factor (usually 3.20 or so) to make sure that the league average FIP is the same as the league average ERA. That is, that’s why the FIP scale and the ERA scale are the same. What’s important to note here are the ratios.


A home run is 4.33 times worse than a walk, and 6.5 strikeouts approximately cancel out the effects of a home run. Let’s take a look at how this correlates to defense.

For now, let’s take a look at the differences between team FIP and team ERA. The graph above shows the run difference between FIP and ERA (earned runs minus expected earned runs from FIP) on the x-axis and team UZR on the y-axis.


There’s a pretty decent correlation here (r = .7). However, the Brewers, in yellow, appear to be somewhat of an outlier. They allowed 76 runs fewer than FIP projected, with only a 14.7 UZR.


A team with 76 runs fewer than FIP projects should have approximately a +45 team UZR. So why the 30 run (or three win) discrepancy? I’m not sure I can answer that question right now, to be honest. Spray charts for Brewers pitchers would help.


The Brewers were much better at fielding on the left side of the field than the right side, so maybe that has something to do with it. Also, it’s possible that Suppan and Bush had abnormally high FIPs because they had career high (in a bad way) HR/9 numbers.


Clearly more observation and analysis is needed here to complete these ideas, but I think it’s at the least an interesting think piece at this point, and we can clearly see that yes, defense does matter.