Saber Smart: Fantasy Baseball Stats and Terms Glossary (Part 2)

In this series of articles, we'll take a look at some sabermetric stats which you'll hear tossed around many RotoBaller articles and many other sites as well. The goal of these glossary pieces is to help you understand these statistics and how to use them in your fantasy baseball analysis.

In Part 1 of this series, we looked at BABIP, LD%, and HR/FB %.  In part 2, we'll look at FIP and xFIP. We'll continue to examine sabermetric stats that can help you in fantasy baseball going forward.

Pitching Sabermetric Statistics

FIP

FIP stands for Fielding Independent Pitching and it's one of the more useful pitching metrics out there for predicting future pitching performances. Basically, what FIP attempts to do is isolate how well a pitcher pitched, independent of things outside of his control such as the team defense behind him or his BABIP-allowed numbers.

The actual formula used to calculate FIP:

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

You can see that the only factors used to judge a pitcher through this metric are his home run rate, his walk rate, his hit-by-pitch numbers, his strikeout rate and the number of innings he's pitched. That's it. The 'constant' you see in the calculation is just used to place FIP on the same scale as ERA to make comparisons easier.

Why do this? Simply put, research (started by the legendary Sabermetrician Voros McCracken) has strongly suggested that a pitcher, on average, has very little control over his BABIP-regulated results, which drive many popular metrics that assess pitcher performance based on actual runs allowed, like ERA. And indeed regression analysis has shown that FIP has far and away more predictive value than ERA alone. This makes FIP a useful stat for when you want to try and tease some of the random variation out of your pitcher's numbers and determine what they might do going forward. A large deviation between FIP and ERA strongly suggests that some regression is due going forward.

xFIP

In the same way that research has suggested that pitchers have little control over their BABIP-allowed, it has also revealed that, on average, the home run rates for pitchers will regress to the league average over time (roughly 9-10%). What xFIP does is calculate FIP, but instead of using the actual HR rate a pitcher has allowed, it instead plugs the league average HR rate into the equation.

The actual formula used to calculate xFIP:

xFIP = ((13*(Flyballs * League-average HR/FB rate))+(3*(BB+HBP))-(2*K))/IP + constant

Again, that 'constant' you see in the equation is just used to bring xFIP onto the same scale as ERA for the sake of comparison.

Statistical analysis has shown that of ERA, FIP and xFIP, it is xFIP that has the most predictive value. Therefore, when attempting to assess what a pitcher is likely to do going forward, it is a very useful metric to look at (in conjunction of course with other important stats).