Jeff Locke’s Great Regression


PNC PARK – Jeff Locke was an All-Star. Now you wonder if he’ll be in the playoff rotation, assuming the Pirates make the postseason.


That’s quite a potential fall.


The statistical community long predicted Locke was a prime candidate for serious regression when he was boasting a near sub 2.00 for much of the first half of the season. Analysts noted Locke’s ERA was likely not sustainable due to his lacking strikeout rate and middling walk rate. The prediction was his ERA would soon resemble his fielding-independent ability (FIP). He’s getting there as his ERA is now 2.90, rising nearly a run since the break, and his FIP rests at 3.85.

Locke continued to defy projections for much of the season mostly due to an MLB-best strand rate, which hovered at an other-worldly 83 percent earlier this month. (It dipped below 80 percent Saturday)


Which Jeff Locke is the real Jeff Locke? 


The truth is Locke had been regressing long before his season-worst outing Saturday (2.2 IP, 10H, 8ER, 3BB, 2K). Locke’s ERA was slowly inching up in large part because of his walk rate:




May: 3.3


June: 4.2


July: 5.2


August: 6.4


Quite simply the numbers suggest Locke has continually pitched away from contact as the season has gone along. He’s been allowing more baserunners and not stranding as many in August.


Maybe he’s just tired having already pitched a MLB-high 139 innings, and nearing the most work he’s had in a professional season.


My sense is he’s trying to be too fine.


Perhaps he’s changed from a pitcher not afraid of inducing contact with his sinking fastball, to one seeking perfection on the edges of the plate. Perhaps he believed his strand rate was a real skill and sustainable. (You’ll hear a chorus of “Nays” from the sabermetric crowd.) Perhaps he was trying to be a 2.00 ERA pitcher.


Locke is not Clayton Kershaw (who is?) whose ERA neighborhood Locke resided in in the first half of the season.


He’s also not the guy we saw Saturday. He’s better than that.


Locke is a pitcher who must pitch to contact and avoid issuing walks. He doesn’t have monster stuff but he’s showed elite command in the past, posting sub 2.0 walk rates at three different minor league stops. His increased groundball rate is the biggest difference in his profile from last season to this season. He should be pounding the lower part of the strike zone, especially with the Pirates’ elite defense behind him.


Locke can still be part of the postesason rotation – if the Pirates get there. To get there he must stop living on the edge.


– TS