Heading into the 2012 season, Brandon Morrow was seen as something of an enigma. His peripherals had been solid across his young career, contributing to a dignified FIP- of 91 since his rookie campaign in 2007, but his ERA never quite fell in line with his fielding independent numbers. In fact, his "E-F" over that time was 5th worst among qualified pitchers (per Fangraphs), bested only by a handful of bust-outs that included Andrew Miller, Manny Parra, Luke Hochevar, and Andy Sonnanstine. With over 400 innings under his belt at that point, it seemed as though Morrow was headed toward that dark, desolate, post-apocalyptic corner of the Baseball world: Ricky Nolasco Town.
Part of Morrow's struggles over that period can be attributed to an unusually large discrepancy between his BABIP from the stretch and his BABIP from the windup. His 2010 splits were particularly damaging in that regard, as Morrow's BABIP with Runners On ranked 4th highest over the last 10 years among starting pitchers that at least matched his 140 IP.
WORST R.O. BABIP SEASONS SINCE 2002
|#||First||Last||Year||BABIP Runners On||BABIP Bases Empty||IP||FIP||ERA||E-F|
What I find most interesting here, though, is that we have a very strange mix of pitchers in this group of 15. Several of these starters we can reasonably assume were suffering from nothing more than bad fortune from random variation in small sample sizes. Clemens, Maddux, Pettitte, Felix, and Glavine are all hugely successful pitchers that would never have achieved their supreme status in the game of baseball if they had genuine troubles locating pitches from the stretch.
Yet, we also have pitchers like Glendon Rusch, owner of a career .75 E-F, Casey Fossum, who finished his career with a rather high E-F of .58 in over 700 IP, and Luke Hochevar who is currently the not-so-proud owner of an .86 E-F in almost as many innings.
So when we had a young pitcher like Brandon Morrow in 2010, struggling with runners on in a limited sample, how were we to know if we were looking at a case of bad luck or a genuine issue with mechanics from the stretch?
We need to keep in mind what a tiny sample we are dealing with when using Runners On splits, but at the same time we cannot forget the enormous impact it can have on a pitcher's ERA. BABIP tends to stabilize only after about 3729 balls in play. And before the 2012 season, Morrow's Runners On sample (as both a reliever and a starter) was only around 525 BIP.
After Morrow endured the historically unkind R.O BABIP in 2010, it didn't get much better in his next season when opposing batters hit for a .342 BABIP against him. But sure enough, here in 2012, Morrow is seeing some very positive regression with an R.O. BABIP of just .267 this season. As a consequence, his E-F has likewise swung in his favor as well, with an ERA now .51 points below his FIP as we head into September.
This is something nearly every other pitcher from our original table experienced as their BIP sample grew. The group's average R.O. BABIP dropped from .379 to .315 in year two, and finally down to just .289 in Year 3. Their E-F normalized from .99 in year one to a -.15 in year three:
|#||First||Last||Year||BABIP R.O. Y1||E-F Y1||BABIP R.O. Y2||E-F Y2||BABIP R.O. Y3||E-F Y3|
*with season-samples below 60IP deleted
BABIP with Runners On can have a profound effect on a player's E-F either way. And Brandon Morrow was unfortunate to have been bitten by it two seasons in a row, but he's also now experiencing the sweet, delicious fruits it can yield here in 2012. I took a sample of 1039 pitcher-seasons with at least 140 IP dating all the way back to 2002, and found a high correlation of .70 between R.O.BABIP and E-F. This might be something to think about next time you think you've found the next case of Ricky Nolasco disease.
All RO/BE splits are from games used as a stater only, include post-season data, and are courtesy of Retrosheet. FIP constants used from Fangraphs' "Guts".
For a good time call @JDGentile on twitter.