James Shields had a rough year. What went wrong?

Things fell off the rails for James Shields in 2016. In 33 starts for the Padres and White Sox he posted a league-worst (among qualified pitchers) 5.85 ERA and 6.01 FIP. Both those marks are considerably worse than the much maligned Jered Weaver (5.06 ERA, 5.62 FIP), who finished second to last in both categories. He combined a career low K/9 of 6.69 (just edging out 2008’s 6.70) with a career high BB/9 of 4.06. It was a rough year for the former ace, to put it mildly.

The season’s exclamation point came in his last start against the Twins when Shields allowed two home runs, bringing his season total to 40, a mark that hadn’t been reached by anyone since 2011.

What happened to “Big Game James?” The 2015 season saw Shields slip a little bit from his previous four-year run of excellence, but it wasn’t nearly this bad. First let’s see if he suffered from some unusually bad luck. Maybe we can pin this disaster season on some bad breaks.

Nope.

His BABIP was once again right around league average. His LOB% was down about seven percentage points, again just about league average. Shields ran a very high HR/FB rate and led the league in home runs allowed for the second year in a row, but in 2015 he still managed to have respectable rate stats despite the uptick in dingers.

Since nothing stands out that allows us to place the blame on luck, we need to investigate his pitch arsenal to see if anything has changed. Let’s take a look at his usage and pitch type run values from the PITCHf/x data available at FanGraphs.

Shields used the knuckle-curve a little bit less and the two-seamer a little bit more, but overall his usage has barely changed. The run values however, oh boy, the run values are a sight to behold. In 2016 Shields did not have a single positive pitch based on run values. Every pitch, with the exception of his two-seamer, was worse in 2016 than it was in 2015. That’s not to say the two-seamer was good. It was not.

There were no drastic differences when looking at the velocity and movement on the pitches either, which leads to a look at his command of the strike zone. The changeup was always Shield’s marquee pitch, but it seems that he had some trouble locating it this year. As you can observe, it was leaking out over the plate more in 2016 than in 2015.

We can also see that Shields gave up more contact on his hard pitches in the strike zone this year than he did in 2015. In this instance hard pitches are made up of his four-seam fastball, two-seam fastball, and cutter.

It wasn’t just one pitch that led to home runs for Shields. He gave up most of his home runs on four-seamers and cutters but saw balls fly over the fence on all of his pitch types.

Baseball Savant

There’s no clear answer as to how Shields should proceed in the 2017 season, but improvement will have to begin with renewed command of the changeup. The pitch was once his calling card and needs to become that once again. It might also be time for Shields to change up his pitch usage. Predictability is the enemy when trying to get big league hitters out, and giving batters a different look might be a good idea.

We’ll see next season whether or not Shields can reclaim his former pitching glory. He has the pedigree of course, but based on 2016 we can’t give him the benefit of the doubt and just assume it will happen. Time for James Shields to make some adjustments.