December 02, 2009
Today's statistic: FIP
What it stands for: Fielding Independent Pitching, devised by Tom Tango, based on research by Voros McCracken.
How they calculate FIP: Tango's simplest version is expressed by the following equation.
Usually a constant is added to scale the FIP to the league average, so the formula winds up looking like:
(13 * HR + 3 * BB - 2 * K)/IP + 3.20
(Not to confuse everyone right off the bat, but BB can also be represented as BB + HBP - IBB, which forgives a pitcher for walks his manager ordered but penalizing him for beanball.)
There is also a stat called xFIP (aka expected FIP) that utilizes many more pitching components in an attempt to normalize for the luck in home runs allowed, but it's outside the scope of this article.
What FIP is good for: FIP is a stat that's supposed to look like ERA, but it sure isn't ERA. The formula is shocking in its simplicity, a result of McCracken's revolutionary finding that pitchers only have control over home runs, walks, and strikeouts and that nearly every other outcome — from hits to errors to runs that score on anything other than a homer — is essentially a product of random chance.
Here's the sentence that shook sabermetrics, all the way back in 2001:
There is little if any difference among major-league pitchers in their ability to prevent hits on balls hit in the field of play.
(There are other things that are a pitcher's fault — hit batters, balks, wild pitches, etc. — and it's possible to create more nuanced formulas that account for exactly how much of each possible outcome is the pitcher's fault; this is the simplest form.) As the Hardball Times says:
"FIP helps you understand how well a pitcher pitched, regardless of how well his fielders fielded."
Put another way, it tells you how well he should have done, not how well he actually did.
How FIP works: You're probably wondering where the coefficients came from — 13 * HR, 3 * BB, and 2 * K. The simple answer is that Tom Tango created a matrix with run values for each play outcome. The coefficients attempt to adjust for how much each home run and walk contribute to the other team's runs scored and how much each strikeout contributes to preventing the other team's runs scored.
Basically, the fundamental difference between old-school stats and new-school stats is that old-school stats measure what happened at the surface level — batting average, earned run average, wins. New-school stats try to measure each player's contribution to those surface stats, while filtering out the contributions of their teammates and the random fluctuations of chance. Because of this, newer stats tend to do a much better job of predicting the future. FIP is a much better predictor of future performance than ERA, wOBA is a much better predictor of future performance than batting average, and so forth.
However, like the Oracle at Delphi, they do a much better job of predicting the future than measuring the present. On a year-to-year basis, due to luck or defense, a player may vastly overperform or underperform his FIP.
Javier Vazquez(notes) is particularly notorious for this. He's had an ERA higher than his FIP for four of the last five years and by a fairly substantial margin. From 2004 to 2008 he had a 4.50 ERA, but only a 4.08 FIP. In 2009, he had the best season of his career and had his first top-four Cy Young finish, surprising many people who just looked at his ERA and won-loss record. But from his FIP, it was clear that he was a better pitcher than many people thought.
When FIP doesn't work: The one thing FIP can't tell you is how many runs the other team scored. So it is a nice tool for player evaluation, but the one thing it doesn't measure is the one thing the team cares about most — whether runs were actually prevented. Fairly or unfairly, Javier Vazquez disappointed a lot of fans in New York and the Bronx. A lot of runs scored on his watch, whether they were his own fault or the fault of something like Derek Jeter's(notes) poor range at shortstop.
Why we care about FIP: It's one of Cy Young winner Zack Greinke's favorite stats! Look, man cannot live by FIP alone. Ultimately, baseball is about runs, runs scored and runs prevented. FIP tells you something about how much a pitcher contributed to his team's cause, but you obviously still need to look at your favorite old reliables: ERA, Runs Scored/Runs Allowed, and so on.
Still, FIP is a great tool to have. It's a version of ERA that tells you how well a pitcher actually pitched, rather than how well his team did. There are always more complicated measures of a pitcher's exact contributions on the mound, but certainly none as simple.
Next week's lesson: wOBA