clock menu more-arrow no yes mobile

Filed under:

Paul DeJong’s troubling profile

Randal Grichuk used to have the most extreme profile on the Cardinals. DeJong might have usurped that role.

St. Louis Cardinals v Philadelphia Phillies Photo by Hunter Martin/Getty Images

Seemingly every year, the Cardinals promote prospects to the major league team. It’s a Cardinals tradition at this point. Magneuris Sierra has already had a cup of coffee this year. He had good results, but I warned that regression was coming if he stayed. Evidently the Cards feel the same as they sent him back down, though they must like him considering they passed over Harrison Bader who started the year two levels higher than Sierra.

Next came Paul DeJong. He was originally called up after Kolten Wong first hit the D.L., and received a lot of starts to the confusion of V.E.B. editor John J. Fleming as well as myself. I think Greg Garcia would be the better choice. He holds a weak 70 wRC+ on the year, but it’s largely thanks to a .260 BABIP on the year. Only the slowest, most pull-happy fly ball hitters have true talent levels that low. Going forward, we can expect Garcia to be better than he has been.

Wong is hurt again, and DeJong is again starting. I’m writing this on Sunday, and going into the series finale with the Pirates he held a 100 wRC+, right at league average. It’s how he got there that’s the interesting part. Dejong has walked just once, which comes to 1.3% of his plate appearances. That ranks as 3rd lowest rate out of 375 players with at least 70 plate appearances on the year entering Sunday.

That would be one thing if he also didn’t strike out much, but that is far from the case. He also strikes out a ton, with the 18th highest strikeout rate (33.8%), inside the highest 5% of the league. Put them both together and you end up with a worse non-contact game than even Randal Grichuk.

To measure this, I separate out the non-contact and on-contact portions of wOBA. I’ve recently used this technique to compare Grichuk with Tommy Pham. Non-contact considers strikeouts, unintentional walks, and hit-by-pitches. Here’s the bottom 10 in the majors in non-contact wOBA, again working with the 70 plate appearance threshold:

Bottom 10 non-contact wOBA 2017

Player non-contact wOBA non-contact% on-contact wOBA on-contact% wRC+
Player non-contact wOBA non-contact% on-contact wOBA on-contact% wRC+
Arismendy Alcantara .025 37.8% .369 62.2% 39
Paul DeJong .026 35.1% .494 64.9% 100
Alexi Amarista .036 21.6% .401 78.4% 78
Cheslor Cuthbert .053 28.6% .283 71.4% 25
Javier Baez .059 28.4% .414 71.6% 89
Rene Rivera .067 34.1% .396 65.9% 74
Nick Hundley .067 32.0% .369 68.0% 67
Kevan Smith .069 20.0% .316 80.0% 60
Rey Fuentes .069 25.6% .337 74.4% 55
Colby Rasmus .069 39.4% .563 60.6% 131

For additional context, .200 is league average. On-contact wOBA is shown as well, of which average is .367. wRC+ shows that even though these players mostly are better than average on contact, their non-contact game puts them in such a hole that they can’t climb out. Tony La Russa’s pal Colby Rasmus is the only above-average hitter shown here. I didn’t pick ten just to make my point either, only 2 of the 11th through 23rd lowest ranked scores have managed an above-average overall performance at the plate (Tyler Moore and Michael Taylor).

DeJong has the third lowest non-contact wOBA, and needed the 29th best on-contact wOBA to make up for it. More on that later. Back to non-contact skills, DeJong gets there with a high O-Swing% (39.9%, the 16th highest rate) and a low Contact% when he does swing (68.2%, the 26th lowest). He’s aggressive in general, with the 14th highest swing rate. But since he makes so little contact, he’s piling up swings and misses. His SwStr% (swings and misses as a percentage of total pitches) ranks 9th highest at 17.8%.

You might want me to pump the brakes. I mean it’s only been 70 plate appearances. But these plate discipline stats don’t need much time at all to stabilize. And there’s also the fact that this is just a continuation of what we’ve seen since he’s started facing more advanced competition. At Double-A last year, he held a 7.2% walk rate and 26.1% strikeout rate. That would be below average but fine in the majors, but translating that performance from the minors isn’t great. Overall the numbers got worse at Triple-A this year, with a 4.7% BB rate and 24.2% strikeout rate. Yeah, he cut out some strikeouts, but the drop in walks is more concerning than the strikeouts are relieving. The projections expect improvement going forward, but not much: a 5.4% walk rate and a 27.6% strikeout rate. They also, however, see a huge regression in his on-contact numbers, projecting just a 81 wRC+ going forward, or 19% worse than average.

Let’s investigate how pitchers are achieving this against DeJong. We’ll start with curveballs and sliders. What follows is three strike zone heat maps, all from the catchers perspective. The first heat map represents all curves and sliders thrown to him. The second represents all swings taken on curves and sliders. Finally, the third represents swings and misses on curves and sliders:

Pitchers aim for the low outside corner when throwing breaking balls against DeJong, which isn’t surprising. The problem though, is that he is showing little ability to control the strike zone. His swing heat map is very similar to how he’s pitched, which includes a ton of pitches well off the plate.

In the third heat map, you can see that most his swings and misses come well off the plate. Pitchers are getting DeJong to expand his zone at will, and reaping the benefits. Overall, he’s seen 101 breaking balls, swung at 57, and missed 23.

Let’s look at the same thing, but with all fastballs:

It’s not just breaking balls that DeJong is having trouble with. Pitchers are pounding the zone with fastballs, and he’s missing a lot of them. Overall, he’s seen 178 fastballs, swung at 95 of them, and missed 29. He’s only making contact with 70% of fastballs he swings at, which is the easiest pitch to make contact on.

In the Swing heat map, you can see that DeJong is at least selecting for balls he should be able to drive, about a third of the way off the inside corner. He’s also swinging at a good amount of pitches up in the strike zone over the outer half, which the third heat map indicates is a hole in his swing.

Non-contact is just one part of the game though. It’s also of less importance. As I mentioned, DeJong had the third lowest non-contact wOBA, but the 29th best on-contact wOBA got him back to average. Most plate appearances end in contact, and players do more damage overall when making contact. So let’s grade out DeJong’s contact quality, using a radial chart:

If you haven’t seen one of these before, this is presented by, using MLB’s statcast data. The protractor shaped image above captures all combinations of Exit Velocity and Launch Angle, and each dot represents one of DeJong’s batted balls. The six shaded regions represent the six qualities of contact. For a full breakdown, check out this post on Aledmys Diaz earlier in the year (or the Pham/Grichuk post I linked earler)

This image is great, but it lacks context. Here’s a breakdown of DeJong’s contact quality, with the percentage of each type shown instead of the raw counting numbers. I’ll also show the average wOBA for each batted ball type, and the league average rate of each.

Paul DeJong contact quality breakdown

Player Barrels% Solid Contact% Flares and Burners% Under% Topped% Weak%
Player Barrels% Solid Contact% Flares and Burners% Under% Topped% Weak%
Avg wOBA 1.433 .692 .630 .095 .206 .046
Lg Avg 6.7% 5.6% 24.6% 24.5% 33.0% 5.7%
Paul DeJong 6.5% 8.6% 25.8% 19.4% 34.4% 5.4%

If you’ve read these types of posts from me before, you’ll know that DeJong’s contact quality doesn’t indicate anything special. He’s very close to average in all the categories, and being average on-contact isn’t enough to make up for how terrible he’s been in non-contact situations.

xwOBA tells the same story. xwOBA is also provided by, and replaces the on-contact portion of wOBA with an estimation based on the average performance of the player’s batted balls in the statcast era. Despite ignoring player speed and how easy it is to shift against a hitter, Craig found that xwOBA is more predictive of future wOBA than wOBA itself.

For DeJong, that’s bad news. His xwOBA is .037 points lower than his wOBA, in the bottom fourth among those with at least 50 at-bats this year. Correspondingly, his wRC+ should have been around a 90 instead of 100, or about 10% worse than league average at the plate. Over a full season, that’s a difference of about 2/3rd of a win. Your reaction might be to say that DeJong is only 23, and he can improve. Well he can, sure, but on average hitters decline from that point.

This isn’t set in stone. Maybe DeJong makes the appropriate adjustments, and improves going forward. That would be great. Or maybe like Grichuk, this just is who he is. Grichuk isn’t quite as bad in non-contact situations, but he’s still bad enough that when he’s not crushing the ball, it looks really ugly. The good news is that a nice bench piece isn’t a bad outcome for a late 4th round pick. Just don’t bet on him to be a starting-caliber player.