clock menu more-arrow no yes mobile

Filed under:

Dillon Gee's expected rates and mysterious BABIP

Is Gee a good pitcher, a bad one, or an average one? Let's dig into the stats to find out.

Al Bello

Dillon Gee has a 3.84 ERA through 16 starts, but only -0.2 fWAR. That 3.84 ERA is right around league average this season, although it would have looked better in 2012, when the average was 4.01. In 2012, Gee recorded a 4.10 ERA despite posting the highest strikeout and ground ball rates of his career. His walk rate dropped from 10.1% in 2011 to 6.3% in 2012. His 3.54 xFIP in 2012 suggested that Gee could lower his ERA if he kept up his rate improvements.

Gee has lowered his ERA since 2012, but not because he sustained his strikeout and ground ball rates. Gee's strand rate increased dramatically since 2012 and he continued limiting walks in 2013. His pedestrian walk rate in 2014 has been offset in part by a BABIP of .235, which seems low considering his unimpressive ground ball and infield fly ball rates. Gee has a 1.15 WHIP this season, but a 4.28 xFIP. It's hard to believe that he can continue to succeed without regaining his previous ground ball, strikeout, and walk rates.

In 2012, Gee's strikeout rate rose from 16.2% in 2011 to 21%, which was supported by an increase in his swinging-strike rate from 9.1% to 10.6%. It was only a temporary gain, as his swinging-strike rate has now dropped for two seasons in a row and his strikeout rate has fallen with it. According to Chris Carruthers's system for determining expected strikeout rates (xK%), Gee could continue to see a decline.

Gee's strikeout rate this season stands at 17%, but his expected rate is 14.32%, down from a 2013 mark of 16.8%. Carruthers's system for xK% differs from Mike Podhorzer's system in that it eliminates pitch framing as a factor. Carruthers argues pointedly that pitch framing is ostensibly a catcher's skill, so it should not affect a pitcher's xK%. Armed with an 89-mph fastball and a 7.2% swinging-strike rate, Gee's strikeout rate is not likely to rise.

A league-average walk rate for Gee this season is disappointing, as he posted above-average rates in 2012 and 2013. Gee has a 7.08 xBB% by Carruthers's formula, which discards intentional walks and umpire bias. That mark is lower than his current 8% walk rate, but still higher than the rates he posted in 2012 and 2013.

Among starters with at least 100 innings pitched this season, Gee's .235 BABIP ranks as the fourth lowest. In his August 12 article, "The Starting Pitcher xBABIP Overperformers," Podhorzer identified Gee as a pitcher whose expected BABIP is much higher than his actual rate. Podhorzer suggested that Gee could see a decline over the rest of the season due to a mediocre batted ball profile.

"Dillon Gee looks to be a prime example of an average pitcher. He offsets the lowish strikeout rate with better than average control and he owns a league averageish batted ball profile. But, that league averageish batted ball profile has somehow yielded a .230 BABIP."

Podhorzer points to a decreased line drive rate for Gee (17.1 LD%) as a possible reason for his low BABIP. That certainly seems like a factor, though Gee's low 1.5% infield hit rate must play a role as well, albeit a small one. Prior to 2014, Gee had never posted an infield hit rate below 6.4%. This year, he has given up only two infield hits in 16 games. While infield hits are a minor factor, Gee's luck in limiting them has been so obvious that it can't be discarded offhand. Given the volatility of the Mets' shortstop position this season and the defensive limitations of Daniel Murphy and Lucas Duda, Gee's miniscule infield hit rate looks like an anomaly.

Gee's low infield fly ball rate has not helped his BABIP, though he has increased his ground ball rate slightly from 42.6% in 2013 to 43.8% in 2014. A look at Baseball Heat Maps' current pitcher home run and fly ball distance leaderboard shows that Gee's 279.05 average distance is pretty mediocre and really not alarming enough to imply that his bloated home run rate (1.25 HR/9) is going to continue in 2015. The only real cause for optimism in Gee's batted ball profile is his line drive percentage, which has dropped consistently since 2012.

In March 2013, Brett Talley predicted that Dillon Gee would likely perform much better than expected. Talley seemed particularly impressed with Gee's slider, which was indeed impressive in 2012.

"The extra swings-and-misses were probably due to the increase in use and effectiveness of his slider. He threw it just 35 times in 2011 and got zero whiffs. But last year he threw it 249 times and got 23 whiffs (9.2% SwStr%). If he keeps using that slider and stays healthy, Gee has more than top 60 upside."

According to PITCHf/x, Gee started using his slider more often in 2013, but it was much less effective. He stopped throwing the slider as much in 2014, and it has produced better results. Gee is still tweaking his arsenal, which is evident in the fluctuation of his slider usage and the addition of a knuckle-curve. Gee's curveball produced a 60.9% ground ball rate in 2012, so it's encouraging that his knuckle-curve has resulted in a 57.1% ground ball rate this season. His changeup has improved since 2011, but it has been accompanied by a drop in fastball velocity and effectiveness.

Though Talley was once one of Gee's biggest supporters, his enthusiasm deflated after 2013.

"The main reason he was able to post a sub-4.00 ERA was a reversal of strand rate fortune. His strand rate went from 68.9% to 77.9%. To repeat his 2013 ERA he’ll either have to regain the above average strikeout and groundball rates or get lucky again."

Thus far, Gee has posted another sub-4.00 ERA with a 76.5% strand rate. The Mets will have a glut of pitching in 2015 and Jacob deGrom's breakout season could mean that there's one less chair left in the rotation when the music stops playing. While no one expected Gee to lead the rotation, he did look capable of producing quality innings while rookie pitchers went through growing pains. Gee can still achieve that in 2015, but he's much more likely to do it by increasing his strikeout and ground ball rates than by relying on a high strand rate and low BABIP.