Monday, April 14, 2014
Defending FIP
When I first introduced FIP, it was in this article. I presented each event in a "spectrum", from the event where the pitcher has full responsibility (relative to the other players on defense), to where he has no responsibility. You can quibble about some of the ordering of the events. And you can quibble whether I should be saying "95%" or "90%" or whatever. But, it's very clear that other than HBP, I don't give the pitcher 100% on anything. And other than baserunning outs, I don't give the pitcher 0% on anything.
It's summarized here:
The first line is almost all pitching, the last line is all fielding, and the line in between is a mix of the two.
And FIP is essentially concerning itself with a SUBSET of pitching. At no point, ever, have I said or implied that the pitcher is 100% responsible for those events used in the FIP equation, nor have I ever said or implied that all the other events are 0% the responsibility of the pitcher.
It is akin to OBP, which treats the walk and HR identically. While OBP treats them equally, it does NOT mean that they are in fact equals. It is simply that in the subset of offense, where we treat each event as a binary outcome, then the walk and HR have to be treated identically. OBP is about FREQUENCY. But there is ANOTHER subset of offense that is concerned with the number of bases gained. That subset is about MAGNITUDE.
The "problem" with FIP is that this subset of pitching makes up such an enormous component of pitching that people will simply TREAT it as if it is the ONLY component of pitching. We don't have this problem with OBP, first because it's obvious that it can't be true (walk = HR), but also because OBP is NOT this enormous component of hitting.
Now, one thing that I might regret with FIP is relying on IP in the denominator. This was back in 2003 or so, when IP dominated everything. But with the advent of Fangraphs, and with our insistence on using per PA measures, FIP should have "rolled" with the times. And it's something I have touched upon, notably here, when I deconstructed FIP. I could just as easily say "PA/4" as the denominator instead of IP (or something like that). Furthermore, the other thing ?that I might regret is scaling to ERA instead of RA9.
But, FIP was constructed simply as a shortcut to Voros' more elaborate DIPS. And I never thought of constructing FIP as a way to get some wider acceptance. Indeed, I never even had that constant in there, until Studes at Hardball Times wanted to present FIP to the masses. I think that's when I started including the constant to scale to ERA.
So, when talking about FIP, just be clear that I am well-aware of the self-imposed limitations. It's not so much that I was wrong, but rather, a choice I made in terms of presenting the idea. Had I used "PA/4" instead of IP in the denominator, it would look clunky. If I put the "/4" in the numerator instead, turning the coefficients from 13, 3, -2 to 52, 12, -8, that would look clunky. The presentation of FIP, and its use, is a feature. Or at least, it was back in the early days of DIPS.
Now that we're more enlightened, we can move forward to better presenting it.
Recent comments
Older comments
Page 1 of 151 pages 1 2 3 > Last ›Complete Archive – By Category
Complete Archive – By Date
FORUM TOPICS
Jul 12 15:22 MarcelsApr 16 14:31 Pitch Count Estimators
Mar 12 16:30 Appendix to THE BOOK - THE GORY DETAILS
Jan 29 09:41 NFL Overtime Idea
Jan 22 14:48 Weighting Years for NFL Player Projections
Jan 21 09:18 positional runs in pythagenpat
Oct 20 15:57 DRS: FG vs. BB-Ref
Apr 12 09:43 What if baseball was like survivor? You are eliminated ...
Nov 24 09:57 Win Attribution to offense, pitching, and fielding at the game level (prototype method)
Jul 13 10:20 How to watch great past games without spoilers