FIP and xFIP Definitions
In this article, we cover two of the most important modern pitching statistics: FIP and xFIP.
FIP stands for Fielding Independent Pitching, and xFIP stands for Expected Fielding Independent Pitching.
For decades, ERA was the standard measure of pitcher performance, but it comes with some fundamental flaws that we discuss below. While FIP and xFIP are not perfect in terms of measuring the performance of a pitcher, most would agree they do a better job representing the pitcher’s true ability than ERA.
The Concept of FIP and xFIP
The need for something like FIP (and xFIP) arose from the inherent flaws in ERA. As you likely know, ERA stands for Earned Run Average, and it is just that. The number you see listed as a pitcher’s ERA when watching a game on TV is the number of earned runs that pitcher has allowed, on average, per nine innings over the current season. Obviously, giving up fewer runs is better than more, so a low ERA is correlated with quality pitching.
That’s fine, but there is a glaring problem that needs to be addressed: what about the other 8 players on the field? The goal of the entire defense is to prevent runs from scoring, and while the pitcher plays an important role in that task, he is not alone. A pitcher who is supported by a quality defense—strong-armed catcher, fast outfielders, etc.—will have an easier time suppressing runs than a pitcher playing with poor defenders.
In other words, ERA tends to assign the credit for good defense, or the blame for bad defense, to the pitcher. This is where FIP comes in. As the name would suggest, FIP aims to remove the defensive component from the equation, focusing instead on the things the pitcher can directly control. By taking the impact of the defense out of the picture, the resulting number should be a more accurate reflection of a pitcher’s ability.
To create a statistic that isolates pitcher performance from the performance of the defense, it’s important to understand what kinds of things are within a pitcher’s control. Thanks to a variety of research done over the years, it is generally accepted that a pitcher has very little control over what happens to the ball once it is put in play. With some limited exceptions, like pitchers who allow a high rate of fly balls, most pitchers will fall within a relatively narrow BIBIP range over a large enough sample (BABIP: Batting Average on Balls in Play).
If balls in play are not a great reflection of a pitcher’s ability, we are left with three plate appearance outcomes that can be used to evaluate pitchers. Those are home runs, walks (and HBPs), and strikeouts. The calculation for FIP integrates these three elements to provide a numerical representations of a pitcher’s performance.
With xFIP, the only difference is rather than using the actual number of home runs allowed by a pitcher, the calculation will opt for the expected number of home runs based on the number of fly balls allowed. The rate of home runs per fly ball tends to be pretty consistent across pitchers given a large enough sample. So, a pitcher allowing a high rate of home runs per fly ball may just be getting unlucky. By using xFIP to normalize the home run rate, any good or bad luck related to home runs is taken out of the equation.
How are FIP and xFIP Calculated?
The formula used to calculate FIP looks a little intimidating at first, but it only requires the use of some commonly-available statistics to make it work. It goes as follows:
FIP = ((13*HR)+(3*(BB+HBP))-(2*K))/IP+ constant
That might look like a jumble of letters and symbols, so let’s walk through it step by step:
- The number of home runs a pitcher has allowed is multiplied by 13
- The combined total of walks and hit-by-pitches allowed by a pitcher is multiplied by 3, and that number is added to the number from step one
- Next, the total number of strikeouts is multiplied by two, and subtracted from the number reached after steps one and two
- The total remaining is divided by innings pitched
- Finally, the FIP constant is added to the number left from the calculations to arrive at the final FIP for the pitcher in question
What is the FIP constant? It is a number which changes from season to season based on the run environment and is used so FIP is on a similar scale as ERA. Once the constant is used, the resulting FIP will match up with ERA such that a number which is considered to be a good ERA will also be a good FIP. An ERA around 3.00 is generally considered very good and, with the addition of the FIP constant, a FIP of 3.00 is also very good. As an example, here are the FIP constants for the last 10 seasons:
To calculate xFIP, things are going to look largely the same within the formula. Only, instead of using the actual number of home runs allowed, we are going to use the number of fly balls allowed multiplied by the league-wide home run rate.
FIP = 13*(Fly Balls*lgHR/FB%) + 3*(BB+HBP) – (2*K)/IP+ constant
The league-wide home run per fly ball rate will naturally vary from season to season, but it tends to hover just under 10%.
Why are FIP and xFIP Important?
FIP and xFIP are important because evaluating a pitcher shouldn’t involve the evaluation of his supporting defense. Using ERA can make a pitcher on a team with a good defense look better than he really is, while a pitcher with a poor defense is punished. If you are trying to predict future performance of a pitcher, basing your evaluation in part on what a defense has done will give you misleading results. To the greatest extent possible, the impact of the defense should be removed in order to accurately determine a pitcher’s individual performance.
It should be noted that FIP and xFIP are also important for the role they can play in taking luck out of the equation. Much like defense, luck is not something that should come into play when evaluating a pitcher. For instance, if a pitcher gives up a few badly-timed bloop singles, he may give up a few runs despite actually pitching well. On the other hand, a pitcher who has a high walk rate and low strikeout rate, yet isn’t giving up many runs, may simply be getting lucky. Over time, such a performance would be expected to lead to a high ERA. FIP gets ahead of the game by reflecting the quality of the pitcher’s performance before the luck runs out.
What are Good FIP and xFIP Numbers?
For some perspective, let’s view the top-10 FIP seasons from 1900 – 2018.
This table makes a couple things perfectly clear. First, it is obvious to see the importance of home runs in the FIP calculation, as nine of the ten entries are from more than a century ago, when home runs were far less common. Also, this table should highlight the brilliance of Pedro Martinez’s 1999 season.
For an updated look at FIP numbers, the chart below includes the top-five seasons from 2000 – 2018.
What are the Problems with FIP and xFIP?
With FIP, one of the drawbacks is the lack of park factors. This is particularly important given the way home runs are used as a major point of focus within FIP. If a pitcher plays his home games in a homer-friendly park, it’s going to be harder to maintain a low FIP.
Another drawback of FIP is its limited usefulness over short periods of time. If you are looking at how a pitcher has performed over the last two or three starts, for instance, FIP is probably not where you should turn. It’s far more useful to look at FIP from a wider viewpoint, often a full season. However, xFIP may provide you with a better perspective, as it takes out the random statistical noise created by a couple of extra fly balls turning into home runs.