Using advanced stats for Fantasy - xFIP

Homer Bailey's xFIP points to a nice rebound. (USATSI)
Homer Bailey's xFIP points to a nice rebound. (USATSI)

We continue the advanced stats posts with yet another pitching metric. If you missed the first two installments, you can find them here and here. Let's dive right into it.

Stat: xFIP (Expected Fielding Independent Pitching)

Where can I find it: FanGraphs carries xFIP.

What does it do: It's basically a slightly altered version of FIP, which I covered yesterday. The main difference with xFIP is that it estimates home run rates for each pitcher. Basically, it attempts to predict how many home runs a pitcher should have given up based on their batted ball rates.

I still don't understand: The FanGraphs glossary is a good place to start.

How can I use it: xFIP can be implemented the same way as FIP. The main difference here is that xFIP can tell you whether a pitcher has been lucky or unlucky with the long ball. This may not sound all that advantageous, but it can be effective in the right situations.

Example: xFIP is the primary reason I'm not freaking out about Homer Bailey's poor start. Though both his ERA and FIP are awful, Bailey has a solid 2.70 xFIP. This is due to the fact that 60 percent of fly balls hit against Bailey have gone for home runs this year. That's a laughably high number. For reference, the league average HR/FB rate last year was 10.4 percent. Bailey's career HR/FB rate is is 11.2 percent. No matter how you look at it, Bailey has been extremely unlucky with fly balls thus far. That's certain to correct itself, and Bailey should see his numbers drop to normal levels.

What are the problems with the stat: Like FIP, you want to be careful using xFIP in small samples. There's also the issue of players consistently beating their xFIP, like Matt Cain. For years, Cain consistently finished with a much lower ERA than xFIP. The metric kept expecting him to give up more home runs. That didn't happen until last season. Still, Cain has beat the stat enough to make it unreliable in his case. On top of that, some players are just prone to giving up home runs. So, using the league average home run rate for every player doesn't always work out. 

Anything else: xFIP can be useful, but it requires a much more specific example. In most pitching instances, I'll rely on FIP unless a player is exhibiting some really odd home run trends. 

Next post: I'm planning to look at walk rate, strikeout rate and HR/FB rate. All of these things are factored into both FIP and xFIP, but sometimes it's important to break them down even further. If you find it hard to accept, or get into, FIP and xFIP, this is a slightly simplified way to come to similar conclusions.

Show Comments Hide Comments
Our Latest Stories