Saberseminar 2019 in Review

August 20, 2019

This past weekend, Dan Brooks and Chuck Korb hosted another Saberseminar, an annual conference focused on "Sabermetrics, Scouting, and the Science of Baseball," held at Boston University. The schedule was packed with more than 50 research presentations and Q&A sessions over the course of two days. This week’s Stat of the Week will give a recap of the big themes, findings, and ideas from the conference, and link to either the presentation itself or related publications when available.

Baseball Info Solutions (BIS) had a presentation of its own, a review of the landscape of four-man outfields by Andrew Kyne. BIS started tracking the alignment in 2018, and to this point in 2019 teams have already more than doubled their use of the tactic, as measured on balls in play (37 in all of 2018, 80 in 2019 through Aug. 1st).

Andrew defined a player as a "four-man outfield candidate" if he hits at least two-thirds of balls that wouldn’t be fielded by a typical first baseman to the outfield. That’s because if you want to put four of six non-P/C/1B fielders in the outfield, you should be expecting at least four of six balls to be hit there. Early returns suggest the tactic works much better on those candidates, but teams haven’t fully refined who to use it on yet.

Four-Man Outfield Performance on Non-Bunts by Candidacy 
Through 8/1/19

Candidate 60 15 3 .250 .317
Non-Candidate 48 24 5 .500 .604


The conference was themed around diversity, so there were several sessions related to getting more women and minorities involved in baseball. That included a Q&A session with Yankees and Red Sox executives Jean Afterman and Raquel Ferreira, and a panel called "Dirt in the Skirt" featuring women who have umpired, played, and worked for the league.

An unofficial theme that you could have seen coming was the physics of baseball, particularly with respect to the ball itself. Lloyd Smith analyzed dozens of 2017 balls and found no significant difference in flight characteristics from previous seasons. Meredith Wills presented her research suggesting that the ball changed in 2019 to have smoother leather and lower seams, leading to less drag and more home runs. And Bart Smith used high-speed cameras to show just how seam height affects the drag.

Another common thread was pitch and pitcher analysis. Glenn HealeyZach Aldrich, and Jesse Jeter each used pitch characteristics to evaluate pitches on their objective quality independent of the results. Both Healey and Vicente Iglesias discussed how to optimize a player’s pitch mix, with Iglesias suggesting that finding the right balance of fastball and breaking ball could be worth about 1 run per 1,000 pitches. Aldrich took his metric and evaluated pitches on the 20-80 scale, and Jeter calculated an expected Win Probability Added by pitch type to create a WAR-like metric.

In addition, there was a good deal of discussion on player health and safety.Brooks Platt showed that after Tommy John Surgery, pitchers tend to lose movement on their fastball. Kristen Nicholson detailed a robust study that Wake Forest is conducting that combines various biomechanics measurements with psychological evaluation to see which factors affect pitcher injuries and rehabilitation. Chris Ewanik did some preliminary research on players who use the C-flap helmet, and through case studies of a few notable hitters showed there might be some plate discipline improvement with the safer helmet, but the effect would be small.

One last finding from one of the presenters: Wataru Ando used home run rates by launch angle and spray angle for each park to estimate how many home runs each player would hit in each park. Yankees shortstop Didi Gregorius would lose the most home runs of any hitter if he played in any of 23 different parks. He must be thankful that he plays in Yankee Stadium.

Saberseminar is always a great event, and it’s growing each year! We highly recommend attending if you haven’t already.

For more baseball content, check out the Sports Info Solutions Blog or the SIS Baseball Podcast.


COMMENTS (15 Comments, most recent shown first)

Weird. As a retired programmer, there are few things that annoy me more than nondeterministic software.​
3:28 PM Aug 26th
Steve -- You're right, at least one of them was my fault (I think a couple), but I don't know why others didn't 'take.'

BTW, it's not consistent one way or the other. Sometimes it does, sometimes it doesn't. (Some on this page did.)

Anyway, notice, the bold worked; let's see if the underline also works.....

And let's see if the italic will work now....​
7:17 PM Aug 24th
What is wrong with this website, that it ignores the italic tags? (The one at the bottom is your fault, Maris, but the one in the middle looks correct to my eye.)
4:17 PM Aug 24th
.....Sorry, hate to tell y'all, but Gregorius just hit a second HR in this game, which is still not being held at Yankee Stadium. :-)
12:20 AM Aug 24th
.....And, as long as we're taking a further look at this (or at least I am)... :-)

I'm sometimes criticized for how I criticize the articles. I say it's an issue of what we [i]expect[/i] from "Articles," and more broadly and importantly. what we expect of serious sabermetrics and serious sabermetricians -- what degree of knowledge and thought we think there should be.

I realize that the casual yuk-yuk about Gregorius was merely casual. But, as I noted, the thought seems to have required an unsophisticated mindset about how players operate -- which, I would offer, becomes even more striking when it's about a player who seems (IMO) to be among the more intelligent and adaptable players.

[i]But in this instance there's even more.[/i]
I don't know about you, but when I see a yuk-yuk like that about a player, written by someone who I assume to be a serious sabermetrician or even a half-serious one, I would think we could take it as a given that whatever else might be the story about the player, he DOESN'T hit particularly great on the road, because otherwise, [i]surely[/i] it would never occur to the writer to put something that way.

Well, in this case, the writer clearly didn't even care to check. He didn't care. That, in my book, is poor sabermetrics -- embarrassingly poor sabermetrics. Forget anything about whether it's fair to Didi Gregorius -- that's nothing. What this is about is what we expect of serious people in this field.

The message here, to writers of the articles and to whoever it may be who vets the articles before they get put on here (if there be such) is, when you write an article, at least sort of KNOW THE SUBJECT. Don't just mechanically work with some limited set of data and hold forth on it. That's not sabermetrics, it's robotic drivel.

[/i]Oh -- by the way, here's Didi Gregorius''s home-road split this year, without counting tonight and the grand slam against the Dodgers at Dodger Stadium:

Yankee Stadium: 114 plate appearances, 5 HR, 16 RBI, .213/.246/.380
Not at the Stadium: 125 plate appearances, 6 HR, 23 RBI, .311/.336/.538[/i]
11:51 PM Aug 23rd
In case anyone's still looking here.....

I think maybe the Dodgers read Alex's article, but not our comments. :-)

Dodgers and Ryu vs. Yanks right now, Dodgers at home, Yanks had runners at 2nd and 3rd with 1 out.
Gary Sanchez up, Didi Gregorius on deck.

I guess if you read this article but not the comments, it's an easy call since this isn't Yankee Stadium, more so since the pitcher is a lefty and Gregorius hits lefty, although I wouldn't have thought they'd much worry about Ryu pitching to anybody, but especially since it's not Yankee Stadium and so Gregorius has nothing to be thankful for in this game (sorry to be laying it on thick, but hey, we're entitled to have some fun over it if an article says something a little stupid).... :-)

So yeah, if they did read the article but not the comments, the obvious and clearly correct move is to walk Sanchez to load the bases and pitch to Gregorius.

I, personally, was delighted that they walked the guy to get to Gregorius, but what the hell do I know.....

I won't lay it on even thicker by saying what Gregorius then did on the first pitch from Ryu.
10:58 PM Aug 23rd
Shouldn't that be 'Skirt in the Dirt'?
8:21 PM Aug 21st
(Thanks for taking a second look!)
6:10 PM Aug 21st
Maris, I see it now -- I failed to fully scroll down and missed the last two paragraphs of your first comment.
5:33 PM Aug 21st
Dan: You misread what I said.
2:16 PM Aug 21st
Maris, your earlier comment below says that data shows that players don't adjust their batting approach to a shift, but your later comment cautions against assuming that players won't adjust their approach to a stadium. Those are two different things, but it seems odd that players would adjust to one, but not the other.
12:07 PM Aug 21st
I'm interested in the changes to the baseball. One question I have is about the variability between baseballs. Is that reduced? If so, how does that impact the game?

I can imagine if baseballs are exactly the same that pitchers can throw more repeatably. How many bad pitches by good pitchers in previous years were just because the ball construction was a little off.

Anecdotally, the best pitchers are doing very well even though HRs are at a rate that has gone through the roof. Could that be because of more uniform baseballs? Was there any analysis along those lines?
8:47 AM Aug 21st
Maris, I agree with your PS, because we both operate on the assumption that players are not robots.
6:42 AM Aug 21st
P.S. I would refrain from saying stuff like "Yankees shortstop Didi Gregorius....must be thankful that he plays in Yankee Stadium" (re "would lose the most home runs of any hitter if he played in any of 23 different parks"), even just flippantly.

It assumes his approach doesn't depend on his ballpark and that he'd be hitting the same way in any other home park.

I mean, it's not like there's anything evil about putting it that way :-) but I'd think it would never enter the mind of a sophisticated observer to put it that way, because we'd be assuming that the player's approach would be different in a different park.
1:13 AM Aug 21st
Interesting info.
About the "two-thirds" rule for 4-man-outfield candidate: Of course it could be said that the bar should be somewhat higher than 2/3 because doing such an alignment would change the incentives for the hitter and he would change his approach accordingly to take advantage of that alignment. And of course I don't know if that's a good point, because, as we've seen, most hitters don't seem particularly to change their approach according to the alignments.

But some do, and presumably if any team were to put such a formula into their toolbox, they'd know to tweak if according to the hitter.

If we were talking, say, Ichiro, or even D.J. LeMahieu, I think they'd know to set the bar a fair amount higher. If it's Chris Davis or Giancarlo Stanton, not particularly.
5:56 PM Aug 20th
©2019 Be Jolly, Inc. All Rights Reserved.|Web site design and development by|Terms & Conditions|Privacy Policy