Although Earned Run Average has long been a standard measure of pitcher performance, it is limited by a couple of factors. First, earned runs are subject to the whims of different official scorers. More importantly, ERA does not take into account the ability of fielders to get to balls and this is a significant part of run prevention. A relatively new statistic – Fielding Independent Pitching ERA (FIP) –addresses the latter problem and is therefore a purer pitching performance metric than ERA.
In 2000, Voros McCracken, who later did statistical analysis for the Boston Red Sox, published the results of his study which showed that there is little difference among major league pitchers in their ability to prevent hits on balls put into the field of play. This conclusion was met with much skepticism, even in the sabermetric community but further analyses have shown that his theory is basically correct. That is, for most pitchers, hits allowed are more a function of fielders making plays than of the ability of the pitcher. In short, what Bill James has always said is true: "Much of what we think of as pitching is actually fielding".
With McCracken's discovery in mind, Tom Tango, author of The Book: Playing the Percentages in Baseball, created FIP from items which are independent of fielders: strikeouts, walks, home runs and hit batsmen. The formula for FIP is (HR x 13 + (BB+HBP) x 3 –K x 2)/IP plus a league specific factor to make it equivalent to ERA. That's a complex calculation but you can just think of FIP as the ERA that a pitcher "should have had" based on his strikeout, walk, homer and hit batsmen rates. One of the benefits of FIP is that it is better than ERA at predicting future performance.
A comparison between ERA and FIP helps to identify pitchers that were lucky or unlucky during a particular season. A pitcher with an ERA that was a lot higher than his FIP was probably a little unlucky. It could mean that his fielders did not make a lot of good plays behind him or that he was the victim of a lot of cheap hits that found holes. Conversely, a pitcher with an ERA that was substantially lower than his FIP was probably somewhat lucky and may have received very good fielding support in his games.
FIP can be applied to minor leaguer pitchers as well as major leaguers. The table below displays the ERAs and FIPs for the most frequently used starting pitchers at each level of the Tigers system in 2008 (data abstracted from Fangraphs.com). In this table, ERA is the pitcher's actual ERA, FIP is the Fielding Independent Pitching ERA described above and FIP-ERA is the difference between FIP and actual ERA.
Table1: ERA versus FIP for Tigers Starters in 2008
Player Team IP ERA FIP FIP –ERA Yorman Bazardo Toledo 130.0 6.72 5.12 -1.60 Chris Lambert Toledo 149.1 3.50 3.15 -0.35 Virgil Vasquez Toledo 159.0 4.81 4.87 0.06 Eddie Bonine Toledo 106.1 4.15 3.77 -0.38 Luke French Erie 170.0 4.02 4.55 0.53 Josh Rainwater Erie 132.0 4.09 4.37 0.28 Dan Christensen Erie 158.2 4.14 4.66 0.52 Rick Porcello Lakeland 125.0 2.66 3.83 1.17 Jonah Nickerson Lakeland 146.2 3.99 3.74 -0.25 Duane Below Lakeland 133.1 4.46 3.91 -0.55 Matt O'Brien Lakeland 107.1 4.95 4.52 -0.43 Jon Kibler West Mich. 154.1 1.75 2.58 0.83 Alfredo Figaro West Mich. 123.0 2.05 2.47 0.42 Manny Miguelez West Mich. 123.0 4.39 4.03 -0.36 Trevor Feeney Oneonta 72.0 3.25 2.88 -0.37 David Stokes Oneonta 64.1 5.74 4.89 -0.85 Luis Sanz GCL Tigers 51.2 3.66 3.97 0.31 Victor Larez GCL Tigers 56.1 3.67 3.00 -0.67
The table shows that four Tigers pitching prospects had ERAs that were more than a half a run higher than their FIPS. This group of pitchers was probably a little unfortunate and each has a good chance to improve their ERAS in 2009 if they pitch the same as they did in 2008:
- Yorman Bazardo (ERA=6.72, FIP=5.12) and David Stokes (5.74, 4.89) each had below average walk, strikeout and homer rates but did not pitch as poorly as their ERAs indicated.
- Victor Larez (3.67, 3.00) had one of the best K/BB rates (54/10) in the Gulf Coast League but his ERA was inflated by a high batting average against (.284). BAA is not generally as good of an indicator of future performance as strikeouts and walks and is not included in FIP.
- Duane Below (4.46, 3.91) had a high K rate (8.51 per 9 IP) but had the highest batting average on balls in play (BABIP) in the Florida State League. The high BABIP suggests that he may not have gotten much help from his defense.
On the flip side, four pitchers had ERAs that were more than a half run lower than their FIPs. This group of pitchers was probably fortunate and will likely have to pitch better in 2009 in order to match their 2008 ERAs:
- Rick Porcello (2.66, 3.83) had a fairly low k rate (5.2 K per 9 IP) but kept his ERA down with an extremely high ground ball rate (65%). It is difficult for a pitcher to maintain such a high ground ball rate every year. He will need to increase his strikeout rate in the future and scouts generally think that he will.
- Luke French (4.02, 4.55) and Danny Christenson (4.14, 4.66) had similar statistics including strikeout rates below 6.0. Such low strikeout rates do not bode well for their future ERAS.
- Jon Kibler (1.75, 2.58) had solid strikeout, homer and walk rates but did not pitch quite as well as his exceptional ERA. He had an unusually low .191 BAA which he is not likely to sustain in future years.
Since FIP is a better predictor of future performance than ERA, we should keep an eye on all pitchers in both groups described above in 2009.
Lee Panas is a special guest contributor to TigsTown - you can read more from Lee on his site Detroit Tiger Tales