Printer-friendly

Contact Author

December 21, 2012

Overthinking It

The Mike Minor Mystery

by Ben Lindbergh

Through his first four starts and 26 1/3 innings of 2012, Braves starter Mike Minor allowed one home run, striking out 21 and walking five. He had a 3.42 ERA, and the Braves were 3-1 when he pitched.

Then came his next six starts. In those six starts (four of which Atlanta lost) and 31 2/3 innings, Minor still struck out 30, but he walked 16 and gave up 12 home runs—as many as Tim Hudson allowed all season. Minor’s outings got so ugly that on May 21st, after the fifth of those sixth starts, Fredi Gonzalez defended him—sort of—by saying, “he only gave up four solo home runs.”

The next day, Atlanta Journal-Constitution columnist Mark Bradley wrote a post entitled, “Numbers suggest Minor hasn’t been quite as bad as all that.” In that post, Bradley pointed out that Minor “has, at least according to one key stat-geek metric, pitched almost as well as Brandon Beachy.” Beachy, at the time, led the majors with a 1.33 ERA; Minor’s ERA was 6.96. Predictably, frustrated Braves fans left 100-plus comments helpfully suggesting that stat geeks start watching Minor’s starts instead of their spreadsheets.

Through that point of the season, there was one obvious difference between Beachy and Minor*: very few of Beachy’s fly balls had left the ballpark, while many of Minor’s had (roughly 28 percent over those six ugly outings, or about three times the league-average rate). The stat Bradley based his post on was xFIP, a spin-off of FIP that replaces the actual percentage of a pitcher’s fly balls that left the ballpark with a league-average rate. xFIP assumes that a pitcher who’s allowed more homers than his fly-ball rate indicates has been “unlucky” and adjusts his stats accordingly, penalizing him only for how many homers he “should” have allowed. But just as some pitchers consistently allow low BABIPs and beat their FIPs, some pitchers consistently allow a higher-than-league-average rate of home runs on fly balls, either because of the way they pitch or because of their ballparks. For them, xFIP is deceptive, like Spanx for pitching stats. It makes their lines look less bloated, but it can't remove the bulges beneath.

*Actually, there were a couple other important differences: Minor also had a higher BABIP and an awful strand rate (because he’d pitched poorly with runners on base). I’m just focusing on the homers here.

At that point in the season, Minor’s and Beachy’s xFIPs were almost identical. The question was whether Minor had been the victim of bad luck on fly balls—in which case his 3.82 xFIP was a better reflection of how he’d pitched—or was doing something to make himself more susceptible to homers. In his first two partial seasons, Minor hadn’t shown any tendency to give up more homers than expected (for a fly-ball pitcher), which suggested bad luck might be to blame. Still, 12 homers was a lot to wipe away with one stat.