Friday, August 14, 2009

In my recent analysis of the Baseball America top ten prospects, I listed a statistic called Fielding Independent Pitching (FIP) as applied to the two pitching prospects listed in the top ten.  FIP will be further explained here and will be applied to the top 26 starting pitchers in regards to ERA in the PEBA to examine how each pitcher is performing independent of defense and luck factors.

FIP was developed by Tom Tango for now-defunct Major League Baseball as a way to filter out the factors/variables not directly under the pitchers control.  That is, to filter out all non-HR batted balls and other luck factors (to be explained further).  The FIP formula is as follows:

FIP = 13 * HR + 3 * (BB + HBP) – 2 * K/IP + 3.2

 (Add 3.2 to bring the number to an approximate ERA number assuming average defense supporting the pitcher and to allow comparison to ERA)

Unfortunately hit-by-pitch data is not available for starting pitchers in our league.  This is not that drastic of a problem as the original formula for FIP developed by Tango did not include HBP in the equation.  Thus we'll only use BB data in developing these formulations here.

To compare FIP to a pitcher’s ERA we simply subtract ERA from FIP (FIP-ERA), and the resulting number will either be positive or negative (or right on the dot).  A positive number indicates that the pitcher has been the beneficiary of some luck and/or above average defensive play behind him; a negative number indicates that that the pitcher has suffered some bad luck and/or poor defensive play behind him.  The larger the number, the stronger the defense and luck has played a part in influencing the pitcher's ERA.

FIP is a valuable stat in that, by filtering out the conditions not under the pitchers control, we receive a clearer picture of the pitcher's true performance.  FIP is not to be taken to replace using ERA or to be an exact equivalent to ERA when comparing the two.  There are a few topics to note before applying FIP to PEBA pitching statistics.  FIP does not credit pitchers for being able to induce ground balls at higher rates than other pitchers (a repeatable skill).  This may not necessarily be too much concern, as modern scouting philosophy in the PEBA suggests HR rates for pitchers appear to be primarily a function of the movement pitchers attain on their pitches.  Fargo Scouting Director Scotty Ross explains, "Movement is a measure of the movement on a pitcher's pitches.  It is harder for batters to make good contact with pitches that have good movement.  As a result, pitchers with high movement ratings tend to give up fewer home runs.”  The other factor that can be construed as a criticism of FIP is park factors (i.e. not all parks are the same dimension and thus it is easier or harder to hit HR in one park compared to another park).

Many PEBA scouts (including Ross) also believe that high fly ball rates don’t factor into how well a pitcher avoids giving up HR, a break from previous lines of thought on the importance of inducing ground balls.  "Ground Ball % (GB%) is a measure of how many balls hit off this pitcher are ground balls, as compared to fly balls.  Pitchers with high GB% ratings tend to get more ground outs and double plays.  Pitchers with low GB% ratings tend to have a lower BABIP (batting average on balls in play), because fly balls are generally more likely to turn into outs than ground balls.”  FIP does not give credit for pitchers who are able to induce higher double play numbers as this is also dependent on the ability of the infielders to turn the double play.

Finally, BABIP and left on base percentage (LOB%) are not factored.  BABIP is a good stat to use when looking to figure out why there might be a large variance between a pitcher's FIP compared to their ERA.  Unfortunately the PEBA does not track LOB%, which indicates how often a base runner is unsuccessful at making it home against that pitcher (league average is usually around 70%).

The following is the top 26 pitchers in the PEBA ranked by ERA.  Listed will be the pitcher's name, ERA, FIP, FIP-ERA, HR, BB, K and IP.

There are only four pitchers that have a FIP-ERA at +/- 1.00 or more, so we can conclude that the majority of these ERA numbers are close to reflecting the true performance of these pitchers.  It shouldn't be too surprising that the majority of FIP-ERA numbers are in the positive.  ERA will spread out much further on a bell curve than FIP simply because there are more variables composing ERA, and thus an increased variance in the final numbers.

Conan McCullough is just flat-out amazing at not allowing HR.  One homerun in 163 innings pitched!  He owns the lowest FIP (2.12) on the list accordingly even though he does not have the best K or lowest BB numbers.  The ERA is a bit of a fluke for him, but at such a level of dominance it’s easy to see him keeping a sub 1.75 ERA the rest of the way out this season.

A bit further down the list we find our second ranked FIP pitcher, George Thompson (2.17).  Thompson's FIP-ERA (0.54) indicates a bit of bad luck.  His 204 strikeouts in 129.2 innings pitched are impressive and are what’s shooting him up the charts.  A good bet to finish stronger and move up the ERA list by season's end.

Here's an interesting case.  Adrian Reséndez was injured not too long ago and is out for the rest of the season.  He was having a fantastic year based on ERA, but his FIP indicates that he was the beneficiary of quite a bit of luck or defensive play.  The lower number of innings pitched (87.2) is playing a part in the variance, but the FIP is over 4.  HR and BB were the cause of that, and over a larger sample size of innings may have started to creep that ERA up closer to his FIP.

– Steve Youngblood

Fargo Gazette