A couple days ago, I posted a brief overview of WAR and what it entailed. Today I’m going to be doing the same with a pitching statistic known as FIP. FIP stands for fielding independent pitching and basically it attempts to evaluate a pitcher based on things only in a pitcher’s control – strikeouts, walks (and HBPs) and homers allowed. Defense is taken out of the equation, removing a major problem with ERA. For instance, a ball hit 7 or so steps to Derek Jeter’s left would probably go under his glove as he dove for it, but Adam Everett would make the play routinely. In instance one, the pitcher would get debited and in version two credited, even though he gave up the same batted ball.
So here’s the formula. It’s super long and you’ll never need to compute it, but it’s nice to know that you can: (HR*13+(BB+HBP-IBB)*3-K*2)/IP you then take the resulting number plus some factor (usually 3.20 or so) to get it set to the same scale as ERA so people know what is good and what is bad. Plus that makes it easier to compare to ERA so we can see if a pitcher is “overperforming” or “underperforming.”
FIP does have some shortcomings. For one, it assumes all pitchers will give up a league average in batted balls. Meaning they will give up x amount of grounders, x amount of flyballs and x amount of linedrives. To determine whether or not a player is outperforming his FIP, a good place to start would be to look at the type of batted balls given up. Grounders are good, flyballs are bad (but only in the sense that they go for home runs) and line drives are very bad.
What else is there?
The next step up the complicated ladder is probably tRA. tRA is computed on the fact that between a pitcher and a batter, there are 8 outcomes that, with a large enough sample size, can be attributed to the pitcher. Those are: strikeout, walk, HBP, line drive, ground ball, infield fly, outfield fly, and homer. I am a tiny bit skeptical at the infield fly/outfield fly thing, but I definitely don’t have the math to prove my theory so I’m sticking with it. So, diatribe aside, next what we need to do is assign values to those outcomes. Namely, we need to know how many runs or outs result from each of these outcomes. These differ from year to year and from league to league, but figuring outs is relatively easy: just use play by play data to calculate how many outs they have been worth. Runs are a little more difficult, and I’m not going to get into it, but basically once your run expectancy matrix is computed, you then find the value of all the events listed above using this formula: VALUE = runs scored + (Win expectancy after event – Win expectancy before). Next this is combined with the number of times a pitcher gives up said event, correct a bit using a park factor, and multiply it by total batters faced. tRA can now be determined with this formula: tRA = expected runs/expected outs*27