Can Josh Johnson Fix The Stretch?

There have been plenty of good signs in Josh Johnson’s return from a season-ending shoulder injury this year. His 2.98 FIP is excellent; he’s striking out nearly eight batters per nine innings and walking well under three. His 9.35 strikeout rate is an improvement over last year’s mark of 8.5%. His 46.9% ground ball rate is healthy, just under his career average. And yet, despite all this, runs keep crossing the plate against Johnson. His 4.14 ERA would easily be a career high.

As is so often the case, the issue is an elevated BABIP. Johnson checks in at .338 so far this season, a career high and 36 points over his career mark. To make matters worse, the most damage has come with runners already on — Johnson has allowed a .305 BABIP with the bases empty but a whopping .383 mark with runners on. The league typically allows more hits with runners on, but the league split is seven points higher with runners on, not well over 70.

Specifically, Johnson has had a big issue with runners on first — in 81 plate appearances, hitters have a .373/.405/.485 line off Johnson in the split with a .450 BABIP. This, at least to me, raises the question: are Johnson’s pitches losing their effectiveness out of the stretch?

Consider: Johnson is allowing a .245/.296/.354 batting line with nobody on base, a slight improvement on his .249/.307/.355 career mark. But typically, Johnson has gotten better with men on base, posting a .238/.310/.335 line (.295 BABIP) in the split. Nor has he exhibited a specific weakness with runners on first (.670 OPS, .303 BABIP).

The problem doesn’t appear to be with the fastball. Looking just at clear stretch situations — runner on first, first and second, or first and third — Johnson’s fastball has a velocity of 92.9 against 93.0 for all fastballs. No change there. Nor is there a significant deviation on any of his other pitches.

Instead, the difference comes in his two most frequently used off-speed pitches, the slider and the changeup. At 16.7%, the slider is still generating whiffs but significantly fewer than its 22.4% norm. And instead of swinging strikes, the result is hits in play — 7.2% of Johnson’s sliders from these stretch situations have resulted in hits or run-scoring outs.

The changeup has been even worse. Of 49 thrown, nine have resulted in hits or run-scoring outs, the second most common result after “ball.” Only one of the 49 changeups has generated a swinging strike, and just four have generated in-play outs.

These pitches appear to lose whatever makes them so dangerous as soon as Johnson hits the stretch. In the stretch, whiffs are becoming hits, and that’s been the difference between us talking about Josh Johnson, Cy Young candidate and the Josh Johnson who could be dealt as the Marlins sell at the deadline.