Game 14, Angels at Mariners: What’s Going on with Grounders?

marc w · August 6, 2020 at 12:24 pm · Filed Under Mariners 

Taijuan Walker vs. Dylan Bundy, 1:10pm

It’s out first day game of the season, and another game against Dylan Bundy, the apparent anchor of a scuffling Angels rotation and staff. Bundy pitched well against the M’s last week, striking out 8 in 6 IP, but he made one mistake that proved decisive: an early mistake that Jose Marmolejos hit for a three-run bomb in the M’s 3-2 win. Tai Walker was even better HIS last outing, tossing 7 scoreless against Oakland in an M’s win.

I spent a lot of time discussing Bundy’s arsenal and its evolution last week, so let’s talk about a global trend in 2020. This morning, MLB Data Architect, former M’s advisor, and friend of this…uh…enterprise, Tom Tango, mentioned an early-season oddity:

Since start of 2020 season, fielders have converted into outs 71.1% of the 7630 balls of hit into play

In 2019, there was NO STRETCH where fielders converted as many plays into outs. The average was 69.1%

2020 is 3.8 standard deviations from 2019

DER is essentially 1 – BABIP

Defensive efficiency is way, way up in the early going. That is, fewer and fewer balls in play are finding empty spaces and becoming hits. It’s only two weeks, of course, but Tango says there was no individual stretch in 2019 in which DER was this high. This is not just a small-sample oddity.

I decided to break this down by batted-ball type, and the thing that jumps off the page is ground balls. Tango mentions that the same basic pattern holds for liners and fly balls, too, but ground balls…well, just look. Here’s a table of the batting average and MLB’s expected batting average (xBA) by year. See if you can spot the outlier:

2017 0.249 0.25
2018 0.246 0.246
2019 0.242 0.245
2020 0.215 0.261

This is odd; xBA has never been higher, but BA has never been *close* to this low. What happens when we add exit velocity (the speed off the bat) and launch angle (the angle off the bat)?

2017 0.249 0.25 83.8 -11.8
2018 0.246 0.246 86 -11.8
2019 0.242 0.245 86.1 -11.6
2020 0.215 0.261 84.3 -12.7

Launch angle is *lower* than any other year, and speed-off-the-bat is much lower than the prior two years. Yes, 2017 was even lower, but I believe there were more “missed” balls in play back then. Focus mostly on 2019 to 2020: these are pretty sizable changes. And because we’re comparing to 2019, you can’t argue it’s because teams are suddenly shifting more; they shifted all the time in 2019!

What kinds of pitches turn into ground balls, and is there anything odd there? Well, the drop in sinker usage has meant that whereas the plurality of grounders were hit on sinkers back in 2017, that’s not true any more: in 2019 and 2020, the most grounders are hit off of four-seam fastballs. That has accelerated here in 2020, with 28% of grounders coming on four-seamers compared to 25.7% last year, with sinkers’ share dropping from 24% in 2019 to 22% this year. Breaking balls’ share of ground balls had been steadily moving upwards (just as breaking ball usage had), but that’s stopped so far. Instead, the share of sliders+curves is at 24% this year, down from 26% last year, while off-speed pitches (change-ups and splitters) are filling the gap. Off-speed pitches’ share is up to 18% from 16% last year. (Note: percentages don’t add up to 100% because I’m leaving off the “everything else” category like knuckleballs, eephus pitches, and cutters).

Balls in play hit off of four-seamers have the highest average exit velocity, so a larger share of four-seam grounders should *increase* the average exit velocity, I would think. Instead, exit velocities on all types have just cratered. But nowhere is that drop more pronounced than on fastballs. Here’s four-seamers:

2017 0.282 0.279 86.7 -10
2018 0.283 0.278 89 -9.8
2019 0.287 0.28 89.3 -9.5
2020 0.24 0.295 88.3 -10.4

And here are sinkers/two-seamers:

Sinkers BA xBA EV LA
2017 0.246 0.244 85.2 -13.1
2018 0.246 0.248 87.8 -13.2
2019 0.242 0.238 87.7 -13.3
2020 0.207 0.251 86 -14.9

Again, these results hold for breaking balls (where BA dropped from .228 last year to .211 this year) and change-ups (BA of .197 in 2019, and .177 in 2020). But look at that drop for sinkers! And look at the .047 drop in four-seamer BA! In all cases, the launch angle is lower in 2020 than it’s ever been, and exit velo’s lower than the previous two years – in some cases by nearly 2 MPH. I don’t know what’s causing this, but it’ll merit some attention through the rest of the season. Pitchers’ stuff has gotten exceedingly nasty, so maybe it’s understandable we’d see more mis-hits (like the one Jo Adell used to pick up his first big league hit), but that was true in 2019, and we didn’t quite see….this.

So: any theories? What could be causing this? And is the rise in expected batting average just some sort of data gremlin, or is it *better* to hit slow/topped ground balls? In this Q and A with the Dome and Bedlam blog (which you should check out), M’s analyst and ex-Baseball Prospectus guy John Choiniere mentioned that part of his role in infield positioning, so maybe he and, uh, all of his colleagues throughout the league figured something out this offseason?

Anyway, today’s line-up:
1: Crawford, SS
2: Moore, RF
3: Seager, 3B
4: Vogelbach, DH
5: Long, 2B
6: White, 1B
7: Gordon, LF
8: Smith, CF
9: Hudson, C
SP: Taijuan Walker

Kind of a B-team line-up with Austin Nola and of course Kyle Lewis sitting out. Shohei Ohtani’s in the Angels’ line-up at DH for the first time since his diagnosis with a forearm strain that’ll keep him from pitching this season.

Jose Marmolejos isn’t available for important dingers in today’s rematch with Dylan Bundy. Instead, he and RP Bryan Shaw were optioned to Tacoma today, as teams need to trim their rosters from 30 to 28.


One Response to “Game 14, Angels at Mariners: What’s Going on with Grounders?”

  1. Stevemotivateir on August 6th, 2020 3:50 pm

    I thought Cortes might have been a casualty, but he survived another day. Kind of. Maybe he gets replaced with Delaplane before too long.

Leave a Reply

You must be logged in to post a comment.