Thursday, January 20, 2011

Fielding Independent Pitching (FIP) Explained

After what was a crushing day of writing for me yesterday, what with 9 AL Central players agreeing to new contracts before heading to arbitration, a shocking retirement announcement from the Royals, and a baffling signing-followed-by-designation from the Tigers - today is about as slow as it gets.

With that in mind, what better thing to do than kick off a series I have been meaning to do for some time. Today I'll begin posting a series of articles explaining all of the statistics you see in my articles. Some people might be familiar with them, but I'd like to give anyone that isn't an easy an intuitive guide that explains what the different metrics (stats) are. What they mean, how they work, and why they might be better than some of the more traditional reference points you're used to.

I'll start today with Fielding Independent Pitching, or FIP, which is how you'll see it refered to in my writing.

What is FIP?

FIP (or xFIP) is a statistic that attempts to marginalize to the greatest degree possible, the effect of a teams fielders, home ballpark, and the effect of luck from a pitchers performance. In doing so, the idea is to give a clearer view of precisely how well a pitcher performed, and how they should perform over the long-run. It's like ERA, except it tells us how a PITCHER performed, whereas ERA tells us more about how the pitcher, in combination with his defense, ballpark, and simple luck performed.

OK, give me more.

FIP as we know it today is a somewhat dated concept that was researched in detail just before the turn of the millenium (yeah, I'll take any chance I can get to say that) by Voros McCracken. When he first published his research on the topic it was known as DIPS - or defense independent picthing statistics.

The idea has evolved over the past decade, but the basic idea is - as stated above - to focus solely on the pitchers performance.

Intuitively anyone can understand that no two defenses (or defenders) are created the same. For example, Royals pitchers over the past three seasons have pitched with Yuniesky Betancourt at shortstop, and Betancourt is a horrible defender. By comparison, the Twins have generally run out strong defensive shortstops - be it Nick Punto, Jason Bartlett (errors and all), or J.J. Hardy.

It doesn't take a big leap of logic to understand that if you have bad defenders behind you, you're going to give up more runs. As a result, ERA can end up punishing (or rewarding) pitchers for the defense they have playing behind them.

Additionally, anyone can understand that a pitcher who plays most of his games at US Cellular Field (home of the White Sox) is going to be at a disadvantage because it's a hitters paradise with very little foul ground and winds that send far more balls out of the park that elsewhere. Further, pitchers who spend their year calling Target Field home are at an ERA advantage. A long fly out at Target Field for example, is almost certainly a home run at The Cell. Again, FIP tries to marginalize these effects.

So how does it work?

FIP works by counting only the things directly* under a pitchers control. Those things are strikeouts, walks, hit-by-pitch, and homeruns. In other words, things fielders can't control. A fielder can't prevent a walk, nor can a strikeout skip over/under/around his glove for a single and he can't prevent a homerun from landing in the 6th row.

By assigning specific values to each occurence (values derived from the compilation of data over the past hundred years), we get a far clearer idea of how a pitcher actually performed. We then take those numbers, make some nifty and spit the whole thing out in an easily understandable format just like ERA.

I'm almost as big of a nerd as you are Corey, show me the math!

The simplest version of FIP can be done with a pretty easy formula created by Tom Tango. It is:
((13*HR)+(3*BB)-(2*K))/IP = FIP

We can go further by adding in HBP at the same rate as walks, and subtracting intentional walks at the same rate. If you want to do that, simply remove the (3*BB) portion of the equation and substitute (BB+HBP-IBB*3).

Further adjustments can be made to adjust for HR rate, which research shows is a fairly static rate, but like fielding, can be effected by a pitchers home ballpark. IE: more fly balls will turn into home runs at Coors Field than Petco Park. This metric is similar to FIP and is called xFIP.

* Directly is a bit of a loose interpretation. FIP can't eliminate all luck/randomness. Umpires still call balls and strikes, and as a result, strikeouts and walks are to some extent still beyond a pitchers control. Additionally, when I say a fielder can't prevent a homerun, well, we all know it does happen on a rare occassion

No comments:

Post a Comment