Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Baseball Newsstand > Baseball Primer Newsblog > Discussion
Baseball Primer Newsblog
— The Best News Links from the Baseball Newsstand

Tuesday, September 05, 2017

Astros, Luhnow at center of baseball’s scouts versus data debate – The Athletic ($)

Posting the latest Rosenthal because the Athletic is running a 50% flash sale for the next hour. Paying about $60 for the Athletic was too high for me. Paying half that is more to my liking.

Jim Furtado Posted: September 05, 2017 at 06:41 PM | 14 comment(s) Login to Bookmark
  Tags: astros

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. Bring Me the Head of Alfredo Griffin (Vlad) Posted: September 05, 2017 at 07:03 PM (#5526559)
because the Atlantic is running a 50% flash sale for the next hour


You mean the Athletic?
   2. RMc's Unenviable Situation Posted: September 05, 2017 at 08:03 PM (#5526601)
Paying about $60 for the Atlantic was too high for me. Paying half that is more to my liking.

RMc don't pay nuthin' for no internet writin'. (Or boobies.)
   3. Jim Furtado Posted: September 05, 2017 at 09:28 PM (#5526655)
Fixed.
   4. Walt Davis Posted: September 05, 2017 at 09:39 PM (#5526677)
I probably just missed it in earlier reports but this is the first mention I've noticed that half of the scouts dismissed were from the pro scouting department. In this day and age, I'm not sure there is much value in pro scouting. Between stats, video, statcast you've got enough objective information that a scout's insights are unlikely to add much. In terms of assessing attitude, commitment, locker room persona, etc. that's a scout talking to opposition FO, managers, players ... I can't imagine they get very far and no reason I can see they should be better at acquiring that inside info than the GM and his staff.

That said, I was thinking a bit about the Hosmer play from a couple of years ago. The play was supposedly the result of Royals scouts saying that Duda didn't throw well so take advantage if you can. Presumably one can find the velocity of Duda's throws on statcast, you can look at video to see how long he takes to catch and throw, etc. so it seems you could derive the same information from non-scout information. But would you look for it? This presumably was the result of a scout at a particular game seeing Duda screw up. Maybe that scout then kept looking for it or reviewed video or just acted on that hunch but, when faced with all of that data, is a team of analysts ever going to ask the question "can we take an extra base on Duda?" It's far from an obvious question, I can't think of any obvious search algorithm that would seem likely to turn it up.

At least one flip side of that is hinted at: "a scout at a particular game." 1B don't have to make that or a similar play all that often. There's a good chance you scouts doing advance work on the Mets wouldn't witness the play that sparked the original observation that Duda doesn't handle that sort of play well. What insights a scout has any chance to come up with is a largely random process.
   5. cardsfanboy Posted: September 05, 2017 at 10:38 PM (#5526720)
About the only real value I can see a major league scout that they can provide would be to determine if a pitcher is tipping his pitches someway or if there is a mechanical flaw with a potential acquisition that they think can be fixed.
   6. Bhaakon Posted: September 06, 2017 at 02:55 AM (#5526831)
Well, I assume scouting can be useful for spotting injuries that players are attempting to play through. If an outfield has excellent throwing stats and such, but tweaked a muscle in his shoulder a week before and is holding back because of it, that's something that might be seen with the eye but not the stats. It might not even show up in the game video if there isn't a relevant play, but a scout on-site might notice that the guy is holding back in warmups and the various non-notable (and hence untelevised) throws that occur during a game.

Yeah, that's pretty thin, but it's all I've got.
   7. McCoy Posted: September 06, 2017 at 08:01 AM (#5526859)
This presumably was the result of a scout at a particular game seeing Duda screw up. Maybe that scout then kept looking for it or reviewed video or just acted on that hunch but, when faced with all of that data, is a team of analysts ever going to ask the question "can we take an extra base on Duda?" It's far from an obvious question, I can't think of any obvious search algorithm that would seem likely to turn it up.

I don't see why you wouldn't look for that. If the data exists I don't see why a team wouldn't crunch the numbers to see what was there. Perhaps the not so bright or the not so up to date might not look but the best would and that would cause a trickle down effect so that eventually everyone would.

The real challenge would be getting your coaches and players to remember it all in the heat of the moment. Perhaps you can drill it into them for a postseason series but I would think it could get to information overload during the season. I also think Hosmer was running regardless of what the report was on Duda and he probably wasn't thinking about that report at all when the play happened.
   8. Rally Posted: September 06, 2017 at 09:09 AM (#5526893)
This presumably was the result of a scout at a particular game seeing Duda screw up. Maybe that scout then kept looking for it or reviewed video or just acted on that hunch but, when faced with all of that data, is a team of analysts ever going to ask the question "can we take an extra base on Duda?" It's far from an obvious question, I can't think of any obvious search algorithm that would seem likely to turn it up.


I don't think an algorithm would find that. What you could do from a database/video is first identify all the times a first baseman has to throw, excluding tosses back to the pitcher and flips to the pitcher covering the bag. That probably gives you a relatively small number of plays. Then you can find those plays on video and see what it looks like.

The value of a scout is that when you start the process you don't know what you're looking for. A 1B with throwing problems was what they found. The scout(s) would have started out just trying to find any small detail that could help down the road. Maybe it's an outfielder who doesn't go after the ball quickly enough and you can stretch a single into a double. Maybe it's a pitcher who has a tell before he throws over to first. Or a million other little things.
   9. Rally Posted: September 06, 2017 at 09:11 AM (#5526896)
I also think Hosmer was running regardless of what the report was on Duda and he probably wasn't thinking about that report at all when the play happened.


Good question. Sometimes I wish I could bend space and time just to answer hypotheticals like that. Say you've got the exact same situation, score and everything, except the Mets' 1B is the 1986 version of Keith Hernandez. I'd like to see if Hosmer still runs.
   10. Tim D Posted: September 06, 2017 at 11:04 AM (#5527048)
"Or a million other little things."

Precisely. With so much $$$ on the line I believe it would be hard to overstate the value of ML scouting. How many you need with all the data available is open to debate, but you need them. Does a curveball have the same bite it had when we saw him before? Is he hiding an injury? Does he fall asleep on the field sometimes? Video can tell you a lot but being there live with a trained pair of eyes is better than video. Is there something he does before he makes that borderline balk move? Etc, etc. Video can tell you a lot but 3D is better than 2D, especially for things like depth on a slider.
   11. Bug Selig Posted: September 06, 2017 at 12:37 PM (#5527174)
Precisely. With so much $$$ on the line I believe it would be hard to overstate the value of ML scouting. How many you need with all the data available is open to debate, but you need them. Does a curveball have the same bite it had when we saw him before? Is he hiding an injury? Does he fall asleep on the field sometimes? Video can tell you a lot but being there live with a trained pair of eyes is better than video. Is there something he does before he makes that borderline balk move? Etc, etc. Video can tell you a lot but 3D is better than 2D, especially for things like depth on a slider.
And I can't imagine they pay those guys six figures. That salary is a popcorn fart in terms of an MLB budget.
   12. Jim Furtado Posted: September 06, 2017 at 01:14 PM (#5527213)
Didn't the Astros talk about grading their scouts as well as players a few years ago? The idea made/makes sense to me. Why pay scouts who aren't actually providing actionable info? Why not shift resources from travel to video analysis? Scouts are a valuable resource and can, if you have some of the best of them, give you an edge. Without a systematic system in place to measure them, however, you're just spending money willy-nilly. Although we don't have details, I suspect their reorganization was well considered and not extreme in its execution.
   13. What did Billy Ripken have against ElRoy Face? Posted: September 06, 2017 at 01:24 PM (#5527224)
Without a systematic system

Is there any other kind?
   14. Walt Davis Posted: September 06, 2017 at 07:16 PM (#5527530)
This is an excellent piece on the Hosmer play with comments from all of the participants. You never know what's 20/20 hindsight but Hosmer mentions both a play in G4 and the Royal scouting reports as a foundation for his decision.

Yes, I think it's unlikely anybody would design an algorithm to look for those sorts of things. I suspect you would pull out the throwing velocity of every fielder and obviously you've got errors and things like that. But "does the 1B make good throws home on plays from 3B to 1B?" ... I'm not sure there are any examples even if you looked for it and you're certainly dealing with a small sample size. The play from G4 that Hosmer cites was a dropped pick-off throw that Duda dropped again when he went to pick it up. That's an indicator of general sloppiness, maybe slow reactions. That play is not going to show up in statcast, the scorebook and I'm doubtful any analyst has thought "I'll review the video of all Mets pickoff throws to see if Duda dropped one and then see how quickly he recovered." (Note, "hey, intern, review video of all Mets pickoff throws, see if you notice anything" is in the realm of possibility ... and maybe they'd see it, maybe they wouldn't.)

I suppose you might be able to pull out all "unusual" plays the Mets defense committed (errors, guys stealing successfully on pickoff attempts, guys scoring from 2nd on infield singles, etc.) and try to notice a pattern of who messed up most often.

And of course, on the play, Duda did nothing wrong but throw the ball 8 feet off target. I assume he doesn't regularly throw it 8 feet off target but that would be easily discernible from his 15 throwing errors every year. So other than planting the idea that there's a good chance Duda would do something to mess the play up (general sloppiness), it's not clear the scout's observation, Hosmer's observation or anything a video/stat analyst could have come up with actually had any value for that play.

This isn't necessarily a defense of scouts. An intern watching every Mets game on video (repeatedly) is probably more likely to pick up on some little things than a scout watching the Mets for a week. This is an issue with some of the items in #10. Does the curveball have the same bite? First, that's what statcast is supposed to pick up -- velocity, movement, spin rate. Second, you need to have before and after to assess whether it's the same. And your advance scout is seeing a given SP probably no more than once so even if they decide it doesn't have the same bite, you don't know if that was just a one-game thing. "Does he fall asleep sometime?" is actually "did he fall asleep (to an unusually high extent) in the 5-7 games the scout actually saw and was the scout awake at the time?" I suspect it was pure random luck that the Royals scout happened to see a play where Duda messed up -- i.e. if he saw a different set of 5 games, there would have been no play to test Duda's reactions or he would not have messed up on that play.

If a scout for the Royals got to follow the Mets for 162 games, hang out in their clubhouse, view their workouts, etc. then I'm sure they'd turn up lots of useful inside info. It is a fair point that given how little these guys are paid relative to the value of the players, it might well be worth it to have scouts out there looking at all 29 teams every game but that doesn't seem to be how they do it.

You must be Registered and Logged In to post comments.

 

 

<< Back to main

News

All News | Prime News

Old-School Newsstand


BBTF Partner

Support BBTF

donate

Thanks to
Dingbat_Charlie
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogTexas Rangers: Shohei Ohtani or not, Rangers still considering nontraditional 6-man, 4-man rotations next season | SportsDay
(10 - 9:36pm, Dec 11)
Last: PASTE, Now with Extra Pitch and Extra Stamina

NewsblogOTP 11 December, 2017 - GOP strategist: Moore would have 'date with a baseball bat' if he tried dating teens where I grew up
(349 - 9:35pm, Dec 11)
Last: You Know Nothing JT Snow (YR)

NewsblogYankees in talks on Giancarlo Stanton trade
(198 - 9:32pm, Dec 11)
Last: ERROR---Jolly Old St. Nick

NewsblogJack Morris, Alan Trammell elected to Hall | MLB.com
(136 - 9:03pm, Dec 11)
Last: Baldrick

Sox TherapyA Container of Milk, A Loaf of Bread and a Dude Who Can Hit Home Runs
(10 - 8:48pm, Dec 11)
Last: the Hugh Jorgan returns

NewsblogDerek Jeter's defense of Giancarlo Stanton trade was weak | SI.com
(3 - 8:34pm, Dec 11)
Last: JJ1986

NewsblogPrimer Dugout (and link of the day) 12-11-2017
(20 - 7:53pm, Dec 11)
Last: Walt Davis

NewsblogRyan Thibs has his HOF Ballot Tracker Up and Running!
(340 - 7:29pm, Dec 11)
Last: cardsfanboy

NewsblogOT - NBA 2017-2018 Tip-off Thread
(1906 - 6:52pm, Dec 11)
Last: If on a winter's night a baserunner

NewsblogMets have talked to teams about trading Matt Harvey: sources - NY Daily News
(7 - 6:49pm, Dec 11)
Last: JJ1986

Gonfalon CubsLooking to next year
(318 - 6:48pm, Dec 11)
Last: Andere Richtingen

NewsblogOT: Winter Soccer Thread
(313 - 5:03pm, Dec 11)
Last: SPICEY WITH A SIDE OF BEER ON A BABYYYYYYY

Hall of Merit2018 Hall of Merit Ballot Discussion
(324 - 3:41pm, Dec 11)
Last: Kiko Sakata

Hall of MeritBattle of the Uber-Stat Systems (Win Shares vs. WARP)!
(375 - 3:36pm, Dec 11)
Last: Carl Goetz

NewsblogPeter Gammons: Splashes and notes from the Orlando Winter Meetings - GammonsDaily.com
(11 - 3:22pm, Dec 11)
Last: Batman

Page rendered in 0.4098 seconds
47 querie(s) executed