Born Under A Bad Sign: John Lackey and the Impact of Luck

Submitted for your consideration, two pitchers

Pitcher A: 5.75 K/9, 3.43 BB/9, 8.4% HR/FB, 4.44 FIP, 4.72 xFIP, 2.95 ERA

Pitcher B: 6.08 K/9 3.15 BB/9, 9.7% HR/FB, 4.71 FIP, 4.70 xFIP, 6.41 ERA

Difficult as it might be to believe, these are real numbers from two very real players’ 2011 seasons. There is an astonishing 3.46 ERA difference between these two despite an xFIP difference of just .02. By the best measures of what a pitcher can control, these two pitchers were equal effective, yet by ERA, Pitcher A was the 17^th best pitcher in baseball, while Pitcher B was the worst by nearly a full run*. Can we really attribute this massive discrepancy to luck?

As you may have surmised by now, Pitcher B is our own John Lackey. Sox fans’ favorite punching bag was baseball’s unluckiest pitcher- judging by the difference between ERA and FIP- by a strong margin. The luckiest pitcher by this measure is Pitcher A from the line above, Tampa Bay’s Jeremy Hellickson. The Rays’ rookie managed a 2.95 ERA despite a FIP of 4.40 and an xFIP, 4.72, just a hair worse than John Lackey. Neither pitcher was very effective; Lackey was 11% worse than the league by FIP, Hellickson 15%. However, their results are so extraordinarily different that it warrants a closer look.

 *Ranking numbers here are for all pitchers with more than 140 IP.

The first important difference comes in their batted ball tendencies. John Lackey has been a ground ball heavy pitcher throughout his career. Even this season, when he posted his lowest GB% since 2003 (40.5%), he still got more ground balls than fly balls. Hellickson is an extreme fly ball pitcher, with only four other pitchers seeing a higher percentage of batted balls taking flight. Getting ground balls is usually seen as a good sign for a pitcher, since they rarely ever result in multiple base hits. However, fly balls also have a  lower batting average on balls in play and BABIP constitutes the most noticeable difference between Hellickson and Lackey. Hellickson had baseball lowest BABIP, just .223, while Lackey had the highest, .339.

What is strange about Lackey’s high BABIP is that it comes in a season where he was actually getting more fly balls than usual. Earlier in the year, I looked at the positive effect an increased number of fly balls had on Josh Beckett, who was among the luckiest pitchers in baseball by ERA-FIP. Two negative signs accompany Lackey’s declining ground ball rate, a rise in his HR/FB and his LD%. 

In Beckett’s case, both his LD% and his home run rates went down, meaning he essentially just traded ground balls for fly balls. Lackey had noticeable jumps in both areas, possibly suggesting that the extra fly balls were not just coming at the expense of ground balls, but as a result of hitters making better contact overall. These variations are not wildly abnormal though, so it is hard to make that case on numbers alone. One odd side note should also be included here; Lackey had an extremely high infield fly percentage, which should have helped him out. His IFFB% of 16.5% was second only to Ted Lilly’s 16.7. Third on that list is Jeremy Hellickson, 16.4%.

And there is the rub. Lackey and Hellickson struck out nearly the same number, with Lackey being slightly better. They walked nearly the same number, with Lackey being slightly better. They got almost precisely the same number of automatic-out infield flies. They both pitched in front of good defenses (by UZR) with Hellickson getting the slight edge. Yet, when it comes to runs scored, they are miles apart.

You could go mad trying to determine what exactly differentiates these two pitchers in skill, because the difference in their luck is so enormous. The only noticeable difference is their secondary stuff. John Lackey has the lower overall swinging strike % and that is entirely due to his secondary pitches. His most oft-thrown pitch, his cut fastball gets a higher whiff rate than Hellickson's primary pitch (a four seamer), but Hellickson has two pitches he can throw with whiff rates over 10%, an elite change (18.7% whiff rate) and a good curve (12.7%); Lackey has just one, his slider and it isn’t as good as either of Hellickson’s pitches, getting 11.4% swinging strikes.

This seems to support the idea that Lackey, while unlucky, was also working with lesser stuff than Hellickson. Beyond his slider, Lackey had little he could throw that hitters couldn’t handle. We recently learned that Lackey will need Tommy John surgery and that may explain the loss in stuff. His fastball and curveball had both been far less effective than in past seasons, without any major change in velocity. While Hellickson isn’t any kind of ace that this point, he does have some tools with which he can keep hitters guessing. It is hard to imagine that this alone could explain the difference between an elite ERA and the worst in baseball, but it is the only thing other than luck that really separates them.

Worst of all for Boston, the massive disparity between these two pitchers was even more dramatic during September, when Lackey improved his FIP (4.17) but posted an even more horrendous ERA (9.13!) in his 23 innings. Conversely, Hellickson’s incredible good fortune in September suggests divine intervention; he was terrible by FIP, with a 5.34 mark driven by a BB/9 of 4.28 and a K/9 of just 2.94, yet somehow he was blessed with a sparkling 2.67 ERA in 33.3 innings. The difference between these two pitchers September results was 14 runs total, or almost 1.5 wins in Hellickson’s favor. On the basis of FIP, we would have expected a difference of 1.17 runs per nine innings in favor of John Lackey, or in a stretch of between 23-33 innings, around half a win, at the most.

With the playoff spot being decided by a single win, the difference in batted ball outcome between John Lackey and Jeremy Hellickson alone was enough to put Tampa Bay over the top. The difference in their underlying skills does not suggest that Hellickson was any better than Lackey, in spite of the massive difference in results. It may be that Hellickson actually did induce weaker contact than Lackey regularly, but if so, he must had done so at an absolutely historic level. More realistically, this difference is largely due to random variation.