Why is Delino DeShields struggling?

The Texas Rangers offense has been a major disappointment this season. Though the bats have ticked up at late, Texas still has a well-below-average 87 wRC+ for the season, and an 86 OPS+. There has been much discussion as to why the offense has been so bad — Rougned Odor’s ongoing struggles, Joey Gallo chasing too many pitches, the injuries — but one of the areas that has not been talked about as much, and that is perhaps the most baffling of the 2018 offensive problems, is Delino DeShields.

In 2017, Delino DeShields put up a .269/.347/.367 slash line, good for a 90 wRC+, rebounding from a dismal 2016 campaign that saw him get relegated to the minors at one point. He was heralded as one of the positive signs of progress from last year, and was penciled in as not just the leadoff hitter for 2018, but as one of the most important pieces of the team, someone some in the media described as “untouchable” because of his importance to keying the offense.

In 2018, Delino DeShields has put up a .202/.295/.286 slash line. That’s a 59 wRC+. As a point of comparison, in 2016, DeShields put up a .209/.275/.313 slash line for the season, with a 55 wRC+, and was slashing .217/.294/.302 when he was demoted in May, 2016. DeShields has, this year, performed about as poorly, offensively, as he did in 2016 when he was sent down.

Now, I can’t imagine we are going to see a repeat of 2016, and see DeShields get sent to AAA anytime soon — among other things, and as I have harped on since the offseason, the Rangers have no other viable center field option, so they have to sink or swim with DeShields — but its a bit alarming that we have seen such a large regression, and it made me want to drill down into the numbers to see what the difference is between 2018 and 2017.

Looking at K and walk rates, there’s no real change. DeShields is a high strikeout player who draws a decent amount of walks -- that’s been his profile throughout his career, and has been consistent the past two years. In 2017, he struck out 24.8% of the time with a 10.0% walk rate. In 2018, those numbers are actually slightly improved -- he has a 23.8% K rate, and a 10.4% walk rate. Those numbers were both worse in 2016 (26.6% K/7.4% walk), but the improvement in 2017 has carried over to this year.

Similarly, we aren’t seeing a dramatic change in his power. DeShields doesn’t hit for much power, and while his ISO in 2018 is has dropped from 2017 -- .083 this year, compared to .098 last year — that’s not a big deal in the overall scheme of things. His batted ball profile hasn’t changed dramatically, either -- according to Fangraphs, he’s actually hitting a higher percentage of balls hard this year (26.4%) compared to last year (23.9%), and the same percentage of balls softly (20.8% in 2018 compared to 20.1% in 2017). His hard hit rate in 2016 was 20.5%.

In terms of balls in the air versus on the ground, there’s a slight rise in ground balls in 2018 compared to 2017 — DeShields has a 48.2% ground ball rate in 2018, compared to 45.1% in 2017, as well as a slight dip in line drive rate — but I don’t know that that is that significant a change. Conversely, DeShields’ infield fly ball rate — basically, pop ups to the infield — has dropped almost in half, from 10.2% last year to 5.4% this year. He’s pulling the ball less — 32.8% of the time, per Fangraphs, compared to 39.4% this year — which is interesting, but again, not necessarily all that meaningful.

Where we do see significant changes from 2017 are in two areas. First is in infield hit percentage -- in 2017, 16.7% of DeShields’ ground balls went for infield hits (19 out of 114). In 2018, DeShields has a 7.3% infield hit rate (4 infield hits out of 55 ground balls). Second, and related, is DeShields’ BABIP — in 2017, he had a .358 BABIP, compared to a .267 BABIP this year, a rather dramatic drop.

How dramatic? In 2017, DeShields was 11th in BABIP in the majors out of 216 major leaguers with at least 400 plate appearances. In 2018, he is 222nd out of 304 players.

My assumption until recently has been that DeShields’ poor offensive performance this year was a byproduct of his hand surgery, and I still think that the broken hand he suffered in the first week of the season could be an issue. However, if it is a problem, it isn’t manifesting in DeShields’ batted ball profile. He’s hitting the ball as hard and as often as last season, is walking the same amount, and is striking out the same amount. The difference between last season and this season is that more balls in play were falling in last year.

This is somewhat borne out by the Statcast data. Looking at Statcast’s xwOBA — a measurement of the wOBA a generic player would expect to have based on his walks, strike outs, and batted ball information as measured by Statcast, DeShields has a .264 wOBA this year, and an xwOBA of .251. Last year, DeShields had an actual wOBA of .315, but an expected wOBA of .263.

So according to Statcast, DeShields has been roughly the same hitter in 2018 as he was in 2017 -- the .251 v. .263 difference in xwOBA isn’t the big a difference. Where we see the difference, though, is that in 2017, DeShields outperformed his xwOBA by 52 points, and this season, he’s overperforming by just 13 points.

DeShields is likely going to overperform his xwOBA in the long term, given that his speed will generate hits on balls in play that for the normal player would be an out — the question is by how much. I don’t have a good answer for how much that should be, but that’s the critical question in evaluating what sort of offensive performer DeShields is going to be going forward. Here’s DeShields xwOBA by year:

2015 — .269

2016 — .235

2017 — .263

2018 — .251

While there has been variation from year to year in xwOBA, and 2016 and 2018 — his two bad years -- have had lower xwOBAs, his actual wOBA has consistently been higher, as one would expect, given his speed. It seems likely that what DeShields’ true offensive performance should be expected to be going forward is somewhere between his dismal 2016 and 2018 performances and his solid 2015 and 2017 performances.

There was some talk here last season about DeShields, about how much of his offense was driven by a BABIP that appeared to be sustainable, and about whether he should be expected to regress in 2018. He has, obviously, regressed this year, but by more than even his biggest detractors would expect. It is interesting, though, that at the end of the day, DeShields is hitting the ball this year much like he has most of his career. It suggests that he is likely to improve just due to regression to the mean, but that the mean he is going to regress to isn’t likely to be as good as 2017.