Tommy Hunter's Curveball and Regression

A month ago on Fangraphs, RAB author Joe Pawlikowski profiled Tommy Hunter.  When he wrote the post, July 28th, Hunter had a 2.31 ERA, best among starting pitchers with at least 60 innings.  Hunter doesn’t profile as an ace, and Joe noted that the primary way he had gotten success in 2010 was by maintaining two nearly fluky features: “Hunter is currently experiencing one of the greatest combinations for pitcher performance: low BABIP and high strand rate.”  At the time, Hunter’s BABIP was .234, second lowest in the AL among qualified pitchers, and his strand rate was 86.2%, highest in the AL. Joe astutely notes that Hunter wasn’t necessarily doomed to see his ERA rise drastically.  Hunter frequently featured a high curve which frequently induces poor aerial contact, thereby reducing the pitcher’s HR/FB ratio and creating a discrepancy between the ERA and the xFIP.

One of the best arguments for the use of advanced stats is not just their internal logic or theoretical simplicity but when they, you know, work. This time they worked.  Since Joe’s Fangraphs’ piece, Tommy Hunter has started 8 games and thrown 43.2 innings.  During that time, he’s allowed 54 hits and walked 14, striking out 21.  His ERA is 6.39, and batters are OPSing .948 against him.  His BABIP during that time was .283, still low but certainly higher than his former mark of .234.  As a result, his BABIP has risen to .257 and his strand rate has fallen to 80%, while his ERA has skyrocketed over a run and a half to 3.99.  Both of those marks are still very high.  Clearly, Tommy Hunter’s dance with good fortune has come to an end.

Yet there is something else at work, something brought up in the comments of the original Fangraphs piece.  As Joe noted, part of the key to Hunter’s success is the way his high curveball induces weak contact.  Yet once the Rangers acquired Cliff Lee, Tommy Hunter’s use of his curveball dropped dramatically.  A commenter noted that this was the result of Cliff Lee’s advice, who recommended that Hunter rely more heavily on his fastball.  I can’t substantiate this claim, but Hunter’s arsenal certainly has changed since the Rangers acquired Lee:

Pre-trade: Curveball 35.0%, Fastball 39.3%

Post-trade: Curveball 19.3%, Fastball 63.2%

This is obviously a big change.  At the end of the day, the cause of Tommy Hunter’s regression is multifaceted.  For one, it’s nearly impossible to sustain a BABIP south of .240 for an entire season.  By the same token, a strand rate of nearly 90% is par for the course for Mariano Rivera, but he’s the ultimate outlier.  It also seems to be the case that a decline in his usage of the curveball may result in better contact from batters.  Cliff Lee and Tommy Hunter may think that his fastball is his best pitch, but it seems that throwing it at the expense of the curveball has resulted in far worse results.  One last thing to note is that Hunter’s BABIP and strand rate currently stand at .257 and 80.7%, respectively, and his FIP is about a run and a half higher than his ERA.  He’s probably not done coming back down to Earth.  Hopefully the Yankees can assist in aiding his regression this evening.