Doing Better Than FIP
Over the last decade or so, the baseball world has divided itself into two camps. In the first, we have the traditionalists. Traditionalists value counting statistics like RBI, homers, hits, batting average, ERA and pitcher wins to evaluate performance. To the old school, there isn ?t much of a story behind the statistics. You either produce or you don ?t. In the other corner, you have the analyticals. The analyticals think there is more to the story than how we typically have looked at the game and that there are other ways to measure success than what appears on the back of the baseball cards.
Often times, each statistic comes complete with a complicated formula and some like WAR don ?t even have a consensus for how to calculate it. Acronyms like ISO, wOBA, BABIP get thrown around, but in the end, they aren ?t relevant to this piece other than to include in an introductory paragraph.
The traditionalists think the analyticals are a bunch of dorks making up numbers. The analyticals think the traditionalists are backwards and too black and white. However, there is a particular statistic held in high regard that even most progressive fans might need to evaluate and update.
FIP stands for Fielder Independent Pitching. The premise behind the number is simple. Good defense helps make pitchers look better. If you have a Gold Glove shortstop and center fielder behind pitcher X in 90% of their outings while pitcher Y has normal fielders behind him, would it be fair to say pitcher X is performing better because his ERA is 3.60 where pitcher Y’s is 3.85?
Enter FIP to even the playing field. Figuring out FIP isn ?t simple, but it ?s not difficult either and it uses the traditional counting statistics that a pitcher can in theory control without the fielders influencing, walks, hit by pitch, strikeouts and homers. Each traditional statistics are weighted, and the good (strikeouts) are subtracted from the bad (walks, HBP, home runs) then divided by the number of innings pitched. That number is added against a FIP Constant for a given season that ?s normally around 3.2 to help it look like ERA. If you ?re interested, you can find the actual formula here.
However, FIP still has its limitations and the quality of batted ball data available now could improve it. Before we go too far down this rabbit hole though, I just want to clarify — I ?m not going to end this conversation with an exact formula for new FIP, but I want to discuss what a better formula would include and why. Maybe we can call it bFIP for better fielder independent pitching.
First and foremost, I would keep the walks, hit by pitch and strikeouts as measurables that gauge pitching performance. Of course, somebody could argue that pitch framing assists strikeouts, but I ?m trying to keep this new formula relatively simple. Adjusting for catchers doesn ?t lead us to that end. I would also keep the constant and continue to divide by innings pitched. For easy comparison and comprehension, the new stat needs to look like ERA. SIERA does, but it too might already be out of date. It ?s also incredibly complicated. One thing it does well is that it balances the weighting between strikeouts and walks, which could be something to consider for whomever decides to actually do the better FIP.
What SIERA does less effectively is that it uses batted ball types like line drives, ground balls and fly balls to serve as it foundation. Exit velocity provides an even more detailed picture. After all, hard hit ground balls will find a hole more often than a poorly hit one. xwOBA considers exit velocities and launch angle, but it doesn ?t look like an ERA. It doesn ?t consider innings and lacks the constant. Here is where I might get a little controversial to the analytical and traditional types alike. I would remove home runs from the formula. If the pitcher makes a mistake and a ball comes off the bat at 100 MPH, launch angle determines if it ?s a home run, double, line out or a single. That ?s on the hitter for partially capitalizing on the errant pitch and I think the difference between a single and a homer are almost arbitrary to evaluating the performance of the pitcher with a ball hit that hard. I would replace dingers with total balls in play over 100 MPH in the bFIP formula.
On the other end of the spectrum, good pitching can lead to poor contact. While BABIP against is generally seen as a luck statistic that should even out to the .300 over time, the best pitchers routinely generate BABIPs below that benchmark. Of the Cy Young Award winners over the last five years in both the NL and AL, only Rick Porcello has an career BABIP over .300. While this in and of itself is proof of BABIP against as a performance stat rather than a luck stat, I would caution people, myself included, to attribute without thought to the latter. To me, there appears to be a relationship between BABIP against and weakly batted balls. I ?m actually late to the party as 538 noticed it in 2015. Here ?s a quote that describes the connection:
At the same time, the pitcher ?s effect is not negligible. While the best batters increase batted ball velocity by as much as 7-8 mph, the best pitchers suppress it by 1.5 mph compared with the average pitcher. That has real significance: Such a decrease roughly equates to a 13-point decrease in batting average on balls in play (BABIP) for a given batted ball. Over the course of a game, the pitchers who can best decrease exit velocity save about a quarter of a run (on average). A quarter of a run doesn ?t sound like much? Multiplied over a season, all those quarters of a run add up to about one win of value.
As a result, I would add a new variable to the bFIP equation and it would credit the pitcher with the total soft balls in play. I ?m not sure what MPH threshold I would use to call it soft contact, but I would probably put it somewhere in the 70-80 range. My very rough formula for bFIP would like this. Note that I would evenly weight BB and strikeouts.
(X*Hardhit balls in play – Y*Soft hit balls in play + Z(BB+ HBP)-Z*K) / IP + Constant
Old school counting stats still have some value even though better ways of evaluating players have come along. Now some of the newer school stats are getting passed by as well. While FIP still explains baseball in a different way than traditional ERA, we can do better. FIP assumes that good fielders make pitchers better, but it misses that good pitchers make fielders better. Keeping most of the original framework and including exit velocity would go a long way to getting the statistics up to date.
Nice article.
You know all the teams probably are using this; we nerdy fans just need to catch up.