clock menu more-arrow no yes mobile

Filed under:

When FIP Won't Buy Wins

June 29, 2012; Miami, FL, USA; Philadelphia Phillies starting pitcher Cliff Lee (33) looks on from the dugout during the fifth inning against the Miami Marlins at Marlins Park. Mandatory Credit: Steve Mitchell-US PRESSWIRE
June 29, 2012; Miami, FL, USA; Philadelphia Phillies starting pitcher Cliff Lee (33) looks on from the dugout during the fifth inning against the Miami Marlins at Marlins Park. Mandatory Credit: Steve Mitchell-US PRESSWIRE

There's been an unhealthy amount of discussion in the mainstream media recently over the quality of Cliff Lee's 2012 season. Heading into yesterday's game against the Mets, Lee was still winless after 13 starts despite fantastic underlying peripherals which contributed to a highly impressive 3.01 FIP and a 3.11 SIERA. With Lee's victory yesterday he improved to 1-5 on the season, but many of his detractors still remain.

Certainly, readers of Beyond the Boxscore are acutely aware of the noise and randomness involved in a pitcher's W/L record, but that doesn't necessarily mean that Lee's situation is commonplace. Such a disparity between a pitcher's W/L record and his FIP has occurred only a handful of times in recent history. But it clearly happens to the best of them-- Roger Clemens, Curt Schilling, and Mike Mussina are among the names who have also suffered under the cruel, undiscriminating hand of an undeserving W/L record at one time or another.

Here's is a look at all player-seasons since 1993 in which the pitcher produced a W/L record below .500 despite a FIP+ of at least 125. (min 150IP)

First Last Age Year Team W L W% IP K% BB% HR% ERA+ FIP+ Team W%
Bob Tewksbury 36 1997 MIN 8 13 0.38 168.7 12.76 4.3 1.66 109 128 0.42
Darren Dreifort 26 1998 LAN 8 12 0.40 180 22.34 7.58 1.6 101 125 0.51
Chuck Finley 39 2002 2 11 15 0.42 190.7 21.51 9.64 1.61 105 128 0.46
Mike Mussina 31 2000 BAL 11 15 0.42 237.7 21.28 4.66 2.84 125 128 0.46
John Burkett 32 1997 TEX 9 12 0.43 189.3 16.79 3.62 2.42 105 131 0.48
Gavin Floyd 27 2010 CHA 10 13 0.43 187.3 18.92 7.27 1.75 106 127 0.54
Roger Clemens 33 1996 BOS 10 13 0.43 242.7 24.9 10.27 1.84 139 141 0.52
John Thomson 23 1997 COL 7 9 0.44 166.3 14.7 7.07 2.08 110 127 0.51
Ben Sheets 25 2004 MIL 12 14 0.46 237 28.18 3.42 2.67 161 164 0.42
Doug Fister 27 2011 2 11 13 0.46 216.3 16.69 4.23 1.26 136 128 0.41
Curt Schilling 36 2003 ARI 8 9 0.47 168 28.83 4.75 2.53 159 175 0.52
Curt Schilling 29 1996 PHI 9 10 0.47 183.3 24.86 6.83 2.19 134 136 0.41
Shane Reynolds 27 1995 HOU 10 11 0.48 189.3 22.1 4.67 1.89 112 135 0.53
Curt Schilling 33 2000 2 11 12 0.48 210.3 19.49 5.22 3.13 125 124 0.52
Erik Hanson 28 1993 SEA 11 12 0.48 215 18.15 6.68 1.89 127 128 0.51

Curt Schilling's 2003 season with the Diamondbacks is clearly the best of the bunch with a nearly-immaculate 175 FIP+. Schilling led the league with a stellar K/BB ratio of 6.1 that season despite finishing with a pedestrian 8-9 record after 168 innings. I can only imagine that if Schilling's season would have occurred today it would be one of considerable controversy considering his superlative saber numbers. Schilling was no stranger to this sort of disparity, however, as he had been its victim in both his 2000 and 1996 seasons while wearing the uniform of the perpetually-disappointing Philadelphia clubs of that era.

Just behind Schilling's 2003 season is Ben Sheets' magical 2004-- often overlooked as one of the better pitching performances in recent memory. In that year, Sheets became one of just 3 pitchers in history to post both a K-rate above 28% and a Walk-rate below 4% with at least 150 IP. (The other two being Schilling and Pedro). But what is almost just as incredible, is that Sheets still managed to lose more games than he was able to win in that remarkable year, finishing the season with a paltry 12-14 record. It leads one to speculate, could Sheets have been able to muster enough votes for the Cy Young award had his 8-WAR showing occurred just a few seasons later, after the sabermetric revolution really began to take hold?

Gavin Floyd and Doug Fister represent the most recent cases of this phenomenon. For those with short memories, 2011 was the Tale of Two Fisters, one which produced a 161 FIP+ and an 8-1 record with Detroit and another which posted a 116 FIP+ with a 3-12 record. When you combine column A with column B, you get a rare and fascinating season.

Floyd has a reputation on the south side of Chicago for being one of the more frustrating pitchers you can find. This frustration may have reached critical mass a few years ago when Floyd posted some of the best numbers of his career, just a fraction away from matching his career high in fWAR, but could not find the victories to match at just 10-13.

John Burkett saw the biggest E-F of the group, losing over 25 points between his ERA+ and his FIP+ in his 1997 season. A wicked .346 BAbip and a merciless 68.3 LOB% won't ever do anybody any favors, but for Burkett this sort of thing was par for the course throughout his playing days. Burkett finished his career in 2003 with a career ERA nearly half a run above his career FIP, a mark which still leads all pitchers with at least 2500 IP since his rookie year.

But perhaps no player's W/L record was more unforgiving than that of Bob Tewksbury in 1997. Tewksbury's offense was especially stingy that year, granting him an average of just 3.7 runs in games that he pitched. Meanwhile, that very same Minnesota line-up curiously managed to generate a much more generous 5.5 runs per game for teammate Brad Radke. Radke went on to win 20 games that season while Tewksbury finished the year just 8-13 despite a FIP 30 points lower than the staff ace.

Bob kept his walk-rate lower than most (4.3%) and his HR-rate even more lower than most at 1.6%. This was an especially impressive feat in 1997 as the steroid-era was just beginning to transition into the Longball Circus. Tewksbury famously threw an eephus pitch to Mark Mcgwire twice that year, both times successfully inducing groundball outs. Maybe Bob would have preferred the W's, but at least his W% of just .38 while still holding a FIP+ above 125 makes him the winner of our contest here today. (WTG, Bob!)

Cliff Lee, obviously, is also dealing with run-support issues in 2012, similar to most of the players in the query. His Phillies have scored an average of just 3.2 runs per game in those that he has pitched, and are on pace to win just 44% of their games. Only Shilling's '96 Phillies, Sheets' '04 Brewers and Tewksbury's '97 Twins finished with lower winning percentages among the qualifiers. Yet with Lee's peripheral strengths, his W/L record is absolutely all but a guarantee to improve significantly in the second half. Whether the media and the fans of Philadelphia choose to appreciate their ace, however, is still very much in doubt.

[Zack Greinke's 2010 season just missed the cut at 124 FIP+, as did another Clemens season, 1993. Tom Gordon's 1997 season also returned at 133 FIP+ with a 6-10 record which I omitted on account of his 17 relief appearances which I felt went against the spirit of the inquiry. I used FIP+ in order to differentiate from Fangraph's 'FIP-' which, as far as I know, still penalizes pitchers for IBB. I used 1993 as the cut-off after seeing some data that suggests FIP rapidly loses its effectiveness prior to that year.]