Last week, in Part Two of this series, we looked at BABIP and how it can evaluate whether a batter or pitcher “under or overachieved” statistically during the previous season, when compared to career BABIP averages.
This week, I want to try and make sense of a different baseball statistic called FIP. What is FIP, exactly? Well, for starters, Mack listed the statistic in a recent article, so I wanted to help explain it in further detail in case anyone had any questions. More specifically, FIP stands for Fielding Independent of Pitching and is similar to another pitching statistic called ERA (earned run average), which is also used to evaluate a pitcher’s performance.
I suppose that grouping ERA and FIP together is a bit too simplistic. ERA, as most fans know, is the average number of earned runs a pitcher gives per nine innings pitched. ERA is more reflective of a team’s performance, while FIP is more indicative of the individual performance. FIP was primarily developed by an extremely smart fellow named Voros McCracken, approximately ten years ago and it was viewed as pretty controversial in “Sabermetric circles”.
Why you ask? Well, according to Voros McCracken (pretty cool first name by the way), a pitcher has very little to do with what occurs after they release the baseball towards home plate. Pitchers only really have control over home runs, walks, and strikeouts (things that are directly dependent on the pitcher). Everything else that occurs, from hits to errors to runs that score on anything other than a home run, is essentially a product of chance.
Taking this a step further, there is very little deviation between different pitchers’ abilities to prevent hits on balls that are put into play.
Why is the statistic important and why should you care as a fan? According to a quote from the Hardball Times, “FIP helps you understand how well a pitcher pitched, regardless of how well his fielders fielded” on that day. A bit more subjective way to describe the statistic is to say it measures how well a pitcher “should have done” compared with how they really did.
So, if a pitcher has a lousy defense behind him, they will presumably perform worse then a pitcher who benefits from a more efficient defense. Does that mean the first pitcher is worse then the second one? Not necessarily. But, it may mean that the first pitcher is undervalued since their more common statistics (such as wins and losses, or ERA) may be skewed. Hidden value, or undervalued players are what keep Paul DePodesta up at night!
So, how do we figure this statistic out? I am glad you asked (but you may not be in another paragraph or two).
FIP = 13*HR + 3*BB - 2*K divided by IP
HR’s mean home runs allowed by the pitcher, BB’s mean walks allowed by the pitcher, K’s mean strikeouts by the pitcher and IP’s are total innings pitched, expressed in one third increments. Some statisticians will also add an additional constant developed to account for different ball park factors, but that is a different article. Just don’t be alarmed if you see some sites add an additional number at the end of the computation. For our purposes, we will focus on the basic formula, with the understanding that a pitcher will typically do better in a pitcher’s park versus a hitter’s park (I know, “Duh”).
The numbers listed are formulaic constants developed by another creator of FIP, named Tom Tango. He developed a matrix with run values for each possible play outcome. The constants 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. Clear as mud?
The cool thing about constants is that you don’t worry about them, you just plug in the values that are represented by the coefficients (letters) and do the math! Remember the proper “order of operations” (parenthesis, exponents, multiplication, division, addition and subtraction) or your stats will be off. OK, this is getting a touch off track, so let’s get back to baseball.
FIP is a nice statistic to try and measure not only a current pitcher’s performance, but it can also be used as a predictive tool, for future performance (which you know is being used by Sandy and Co). This is especially the case when a pitcher’s FIP is drastically lower then ERA.
For an example, let’s look at our very own Jon Niese from 2011 (thanks for the idea, Mack). Jon started 26 ball games (before getting injured) and threw a total of 157.3 innings while doing so, which was good for a record of 11 wins and 11 losses. He also struck out 138 batters and only walked 44, while generating an ERA of 4.40 and a WHIP of 1.41 (which is walks plus hits divided by innings pitched, i.e. base runners allowed per inning), which is the direct result of a whopping 178 hits allowed (primarily out of his control). Wow is this Oliver Perez reincarnated?
Your first reaction might by that he had an average, or below average season. That may be, but scouts and opposing general managers would disagree with you and they would gladly take him in trade. Using our new, favorite statistic (FIP) shows that Niese had a FIP of 3.36 (using the above listed formula), which is a whole run lower then his ERA!
If you look at pitchers with at least 100 innings pitched in 2011 (which focuses primarily on starting pitchers), Jon Niese was 30th in all of MLB, ahead of pitchers like Tim Hudson and James Shields, and just behind David Price and CJ Wilson. That is hardly average!
What the hell is going on? Could this be due to the Mets’ having a poor defensive team? Not just errors, but poor defenders that didn’t even get to balls they should have? Perhaps, but more importantly, it shows that we have a nice, young lefty for our rotation now and in the future. Consider another lefty, named Cole Hamels (who I think is a nice comparison). With a few extra seasons under his belt, he produced an FIP in 2011 of 3.05, which I think is within reach for Jon in 2012 and beyond. Would you take Hamels in our rotation right now?