Explaining ERA-, FIP- And xFIP-

In the sabermetrics world, we hear so much about valuing pitcher using estimators like FIP, xFIP, SIERA, etc. However, these stats still lack context and can be hard to judge sometimes. That is where ERA-, FIP- and xFIP- come in.

Essentially what these stats do is take ERA, FIP and xFIP and give them context (for an explanation of FIP and xFIP, check here). They take into account both park and league factors and adjust a pitcher’s ERA, FIP or xFIP based on them. This is then scaled with 100 being league average.

Like ERA, the lower the number in ERA-, FIP- and xFIP-, the better. Each one point change in these stats represents a one percent difference compared to league average. For example, a pitcher with an 80 FIP- has a park and league-adjusted FIP that is 20 percent better than league average, and a pitcher with a 115 FIP- has a park and league-adjusted FIP that is 15 percent worse than league average.

The formula for this ERA- is as follows. For FIP- and xFIP-, you would simply replace ERA in the equation with FIP or xFIP.

ERA- = 100*((ERA + (ERA – ERA*(PF/100)) )/ AL or NL ERA)

*PF=park factor

The benefits of these stats are simple because they take park and league advantages or disadvantages and attempt to take these effects out of the equation. This makes it easier to compare pitchers from both the same year and differing years by taking into account what environment the pitcher was pitching in.

For example, a 3.50 ERA in 1998 and a 3.50 ERA in 2012 were much different because offense was more prevalent in 1998. Since ERA- is scaled to league average, however, it is easy to look at ERA- in these cases to see what a 3.50 ERA meant in the context of the environments of ’98 and ’12. Likewise, a 3.50 ERA at Coors Field last year was impressive, but at Petco Park it was not, and ERA- can illustrate that for us.

There are some downfall to these stats, however. They only go as far a their base stat, and ERA, FIP and xFIP are hardly perfect. Also, the park factor used to make the adjustments is estimated to be the same for every player who played in that park even though a park might affect different pitchers differently.

For example, a fly-ball pitcher in Tropicana Field, which is notorious for holding in home runs, is going to be affected more by the park than a ground-ball pitcher who would be equally effective in a park that was more home-run friendly. Both of these pitchers receive the same adjustments via Tropicana Field’s park factor even though the fly-ball pitcher clearly has a greater benefit from pitching there than the ground-ball pitcher.

When evaluating a pitcher, ERA-, FIP- and xFIP- are probably the best single stats to use. While they certainly have their limitations, the fact that they give context is key, and they can still do that while taking into account many other factors that FIP and xFIP adjust for on their own.

Want an explanation of another stat? Check out our Saber Glossary.

Load Comments