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

Monday, August 13, 2018

Fangraphs: Jackie Bradley, Jr. and the Error Rule

To summarize:

1. The MLB rules state that the part of the error rule about ordinary effort is an objective standard.

2. The part of the error rule that says it’s all dependent on the judgment of the official scorer establishes a subjective standard.

3. Statcast is more 1. than 2.

IMO the standard of “ordinary effort” applies to any part of the fielding process, not the sum total. If a fielder gets an extraordinary jump, runs with extraordinary speed on an unusually efficient route, and then botches the catch, it doesn’t matter that ordinary effort wouldn’t have put him in position to make the catch. A fielder in position to make the catch will, in ordinary effort, make the catch. Failing to catch it, after getting into position, is an error. Failing to get into position to make the catch - through ordinary effort - is not an error.

villageidiom Posted: August 13, 2018 at 12:36 PM | 39 comment(s) Login to Bookmark
  Tags: boston, boston red sox, defense, defensive metrics, errors, jackie bradley, jr., red sox, statcast

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. Zach Posted: August 13, 2018 at 01:59 PM (#5725826)
Is that ball really a 6% catch probability? Bradley has to run a long way because he's playing short to the pull side when it's slightly to the right of being straight away. He's at maybe 10 o'clock, and the ball is closer to 1:30.
   2. Zach Posted: August 13, 2018 at 02:00 PM (#5725831)
I'm closer to villageidiom's definition. It's a good run to get into position, but muffing the catch like that is an error.
   3. villageidiom Posted: August 13, 2018 at 02:46 PM (#5725856)
Is that ball really a 6% catch probability? Bradley has to run a long way because he's playing short to the pull side when it's slightly to the right of being straight away. He's at maybe 10 o'clock, and the ball is closer to 1:30.
It might not be a 6% catch probability. Per Baseball Savant's catch rate breakdown page, for balls having a hang time of 4.2 seconds and a fielder 80 feet away, 69 of 71 balls hit this year have landed safely. As far as this season goes 6% is generous; it's not even half that.
   4. Lyford Posted: August 13, 2018 at 08:03 PM (#5726066)
Regardless of what the probability was when the ball left the bat, a fielder running that long and getting in that position to have the ball hit the center of the glove is going to catch the ball, or be charged with an error. Bradley may make a similar drop one more time in his career, or not, but when the ball arrived at his glove, he should have caught it. Clearly an error, despite the fact that most guys don't get there.

If the point is the errors are a horrible way to evaluate defense - well, yeah. But that's neither new nor controversial nor particularly interesting...
   5. If on a winter's night a baserunner Posted: August 13, 2018 at 08:39 PM (#5726093)
I think the point was that the rules are ambiguous about whether a single part of the act of fielding can independently result in an error regardless of how extraordinary the effort as a whole may be, or whether the entire fielding attempt has to fall within "ordinary effort".
   6. Textbook Editor Posted: August 13, 2018 at 09:25 PM (#5726113)
Asking because I honestly don't know the answer to this...

If only ~6% of players even GET to a ball in the first place, does even getting there to be able to make the error count as an overall + in the defensive metrics?

Intuitively I would guess it does not, but... it sort of feels like it should? Though of course I confess I don't know how you'd "count" that in any statistically valid way...
   7. Jose is an Absurd Force of Nature Posted: August 13, 2018 at 09:34 PM (#5726119)
TE - That’s a good question.

As for the play in question it says a lot about Bradley that I was stunned by the drop. It never dawned on me that he had done anything special just getting to it, when the ball was hit I figured he’d make the play. He makes the occasional diving catch but the plays he makes that impresss me are the ones like this one where he just gets there so easily. He’s not a burner, he runs well but he’s not super fast but his first step is just otherworldly. I love to watch defense when I’m at games and JBJ is special out there. He so rarely gets a bad read, it’s like he’s got a radar or something and as soon as the ball is hit he is moving in the right direction.
   8. Chip Posted: August 13, 2018 at 09:43 PM (#5726121)
Kind of mystified by the Statcast baseline #s that would put this catch probability at 6%, and the diving catch-and-tumble-into-the-wall at 42%.

Also mystified how someone intent on trashing the latter catch in a recent game thread could argue that Bradley took a bad step in the wrong direction to start that play.
   9. Gold Star - just Gold Star Posted: August 13, 2018 at 10:43 PM (#5726154)
If I'm scoring that game, it's an E8.
   10. Carl Goetz Posted: August 14, 2018 at 09:45 AM (#5726254)
Someone needs to explain to MLB what the word "objective" means. The Ordinary Effort part of the rule is clearly a subjective decision. Now they've given some good guidelines by which to assist the scorer in making his subjective decision. But the objective fact is that an "average" centerfielder making the same play with "ordinary effort" is an entirely theoretical scenario being played out in the official scorer's head (as well as any of us watching the game). What Jackie Bradley did on the play is an objective fact; what the theoretical "average" player would have done in an alternate universe is the official scorer's opinion.
   11. villageidiom Posted: August 14, 2018 at 10:01 AM (#5726264)
Kind of mystified by the Statcast baseline #s that would put this catch probability at 6%, and the diving catch-and-tumble-into-the-wall at 42%.
Baseball Savant has the former one at a 62% chance, not 6%. (Search in the link for "result = Field Error". It has a hit probability of 38%.)

So I guess the next question is this: if Statcast says that a ball with this hang time and a fielder this distance away has been caught only twice in 71 opportunities, then why is the catch probability for this ball 62%? I assume it has to do with exit velocity. But does that make any sense?

Same hang time + higher exit velocity = lower launch angle, right? Likewise, same hang time + lower exit velocity = higher launch angle. But in either case we're talking about balls not just with the same hang time but the same initial fielder distance. The launch angle and exit velocity really play into what happens after the fielder doesn't catch it, but in any case the fielder has the same amount of time (the hang time) to cover the same amount of ground (the distance to the ball).

Something is wrong here.
   12. Carl Goetz Posted: August 14, 2018 at 10:02 AM (#5726266)
"If I'm scoring that game, it's an E8."

Yup, and that's reason #8,653 that errors are a useless stat.
   13. The Interdimensional Council of Rickey!'s Posted: August 14, 2018 at 10:10 AM (#5726272)
Watched the video. That's an error on the centerfielder. He arrived under the ball on his feet (wasn't unbalanced from a dead sprint run forward) with his glove up. Don't care anything about how many other players fail to get there due to getting a worse jump, etc. JBJ got there in a position to make the catch, and clanked it. E8.

The takeaway here is that FLDG % is a stupid way to judge defenders. The number of these plays that a batter might get as a "hit" vs a "reached on error" isn't significant enough to worry about for batting analysis.
   14. SoSH U at work Posted: August 14, 2018 at 10:14 AM (#5726275)
The takeaway here is that FLDG % is a stupid way to judge defenders. The number of these plays that a batter might get as a "hit" vs a "reached on error" isn't significant enough to worry about for batting analysis.


And we don't do that any longer.

But errors have a vital place in the game, even if they don't have a vital place in the value equation. The problem is thinking that stats only exist for how they help us determine value. Getting rid of them, as some have suggested, would be profoundly stupid.
   15. Carl Goetz Posted: August 14, 2018 at 10:19 AM (#5726279)
"But errors have a vital place in the game, even if they don't have a vital place in the value equation. The problem is thinking that stats only exist for how they help us determine value. Getting rid of them, as some have suggested, would be profoundly stupid."

So, if it doesn't determine value, what's the point? Just Nostalgia? I don't have a problem with the error stat still existing (I can ignore it), but I fail to see how it would be stupid to get rid of it.
   16. PreservedFish Posted: August 14, 2018 at 10:33 AM (#5726287)
So, if it doesn't determine value, what's the point? Just Nostalgia?


Explaining what happened during a game.

Which is the most important job of statistics. A record of what happened.

I get that everyone plays fantasy baseball, and everyone pretends they're a GM, but come on.
   17. bunyon Posted: August 14, 2018 at 10:33 AM (#5726288)
Sure, that's an error. But not getting to a ball that the fielder should get to should also count as an error and, yet, hardly ever does.

I see errors as relatively unimportant for measuring fielders (unless they really pile up) but very important for measuring hitters. If a guy hits a ground ball to short that the SS muffs, he shouldn't get credit for a hit.

In this play, I could see it either way. Sure, JBJ should have caught it once he got to it. But, if few would have gotten to it, the hitter probably deserves credit for a hit.
   18. The Interdimensional Council of Rickey!'s Posted: August 14, 2018 at 10:35 AM (#5726290)
So, if it doesn't determine value, what's the point? Just Nostalgia? I don't have a problem with the error stat still existing (I can ignore it), but I fail to see how it would be stupid to get rid of it.


Seconded on this question. I understand the continued use of error tracking for aesthetic reasons. Fans, as well as anyone else, know that that JBJ play was an error. No analytics department worth its salt - so no one except probably the Mets - worries about it for analysis. So as far as I can see, the only real "debate" here is between folks who think we should care about the play JBJ made to able even make the error, and the folks who don't. So another twirl around the "stats vs scouts" thing, but just in regard to who is being the most proper fan. I see no value in that, personally.
   19. SoSH U at work Posted: August 14, 2018 at 10:38 AM (#5726297)
So, if it doesn't determine value, what's the point? Just Nostalgia?


Explaining what happened during a game.

This. If you think it would be better to describe the the Buckner play, for example, as BIP-BR instead of E-3, then Brian Kenny has a crappy newsletter he'd like you to subscribe to.

   20. Carl Goetz Posted: August 14, 2018 at 11:16 AM (#5726322)
"This. If you think it would be better to describe the the Buckner play, for example, as BIP-BR instead of E-3, then Brian Kenny has a crappy newsletter he'd like you to subscribe to."

E3 does a pretty good job of explaining what happened on the Buckner play. E8 does a terrible job of explaining the JBJ play that this article is about.
   21. SoSH U at work Posted: August 14, 2018 at 11:23 AM (#5726332)
E3 does a pretty good job of explaining what happened on the Buckner play. E8 does a terrible job of explaining the JBJ play that this article is about.


What would do a better job? If you don't call this an error, you're basically limiting the number of errors a good outfielder can make compared with a less rangy fielder? Does that really make more sense?

He dropped a fly ball. His inherent ability made it possible that he was in position to drop a fly ball that others wouldn't have reached, but he still dropped a fly ball that he catches more than 99 times out of 100.

To bunyon's point - this ought to be a hit because of how few batters would have gotten to it. If we're going to go there, why should we even care if he catches or drops it?
   22. PreservedFish Posted: August 14, 2018 at 11:33 AM (#5726353)
Obviously he should get a .06E8
   23. jacksone (AKA It's OK...) Posted: August 14, 2018 at 11:43 AM (#5726381)
It does seem wrong that JBJ gets penalized for otherwise doing well. It's an error on him, but wouldn't be on most any other CF, simply because they wouldn't even be there to begin with. I don't really have an issue with JBJ getting the error, and this isn't a good example, but there are plays where no fielder touches a ball that clearly should be caught and no error is awarded.


Explaining what happened during a game.

Which is the most important job of statistics. A record of what happened.



And even if we all "know" errors don't matter, and don't really tell us a lot about fielding ability, and all they do is tell us what happened in the game, remember that the *lack* of errors changes how we remember players. 'No error on balls not gotten to' is basically why Jeter won gold gloves.
   24. bunyon Posted: August 14, 2018 at 11:49 AM (#5726390)
To bunyon's point - this ought to be a hit because of how few batters would have gotten to it. If we're going to go there, why should we even care if he catches or drops it?

My lack of knowledge on advanced metrics comes out here. Does the batter get credit for production on this play?

I agree, to record "what happened" - E8. But it's a bad description of the play. But, certainly, JBJ shouldn't get a deduction and the hitter should get some credit in terms of assigning value.

Maybe we just have to admit the world is complicated and there is no single designation or number that can properly evaluate some things.
   25. Carl Goetz Posted: August 14, 2018 at 11:52 AM (#5726393)
"What would do a better job? If you don't call this an error, you're basically limiting the number of errors a good outfielder can make compared with a less rangy fielder? Does that really make more sense?"

Question#
1) I don't know, but probably would require a more complicated notation. But when complicated things happen in real life, they are complicated to explain.
2) If you're just describing what happened, who cares? If using for evaluation, I would expect a good stat to rate a better player more highly than a worse one.
3) Yes. See #2 above.

"To bunyon's point - this ought to be a hit because of how few batters would have gotten to it. If we're going to go there, why should we even care if he catches or drops it?"
"Obviously he should get a .06E8"
If we're evaluating what happened exactly, this play and the final result should at some level being apportioned to the hitter (positively in this case) and to JBJ (Negatively) and probably to the pitcher as well. I doubt we'll ever see this applied to traditional stats .94H and .06E8 as PreservedFish jokingly (I assume) suggested. I do expect that sort of performance attribution to come more and more into WAR calculations as we get more and more Statcast data. Partial performance attribution is already being used by Tom Thress in calculating his Player Won-Loss records. Basically, he takes the change in win probability after each play and attribution that change to the relevant batter, pitcher, baserunners, and fielders.
   26. PreservedFish Posted: August 14, 2018 at 11:53 AM (#5726395)
Echoing some of the first comments, I'm not buying that there was a 6% chance of making this play. The gif doesn't show JBJ's jump so I can't be confident, and I have no doubt that it was a great run he made, but 6%? JBJ is very good but I don't think he often makes EXTRAORDINARY plays look literally routine. There are two guys in his division that might have better range than he does.
   27. Jose is an Absurd Force of Nature Posted: August 14, 2018 at 11:57 AM (#5726404)
'No error on balls not gotten to' is basically why Jeter won gold gloves.


Which is one of many many examples of why we shouldn't worry too much about subjectively voted on awards.
   28. If on a winter's night a baserunner Posted: August 14, 2018 at 02:53 PM (#5726628)
Obviously he should get a .06E8
Taking as a given the validity of the 6% thing, I'll cosign this.
   29. cardsfanboy Posted: August 14, 2018 at 03:16 PM (#5726657)
Echoing some of the first comments, I'm not buying that there was a 6% chance of making this play. The gif doesn't show JBJ's jump so I can't be confident, and I have no doubt that it was a great run he made, but 6%? JBJ is very good but I don't think he often makes EXTRAORDINARY plays look literally routine. There are two guys in his division that might have better range than he does.


And it also makes me wonder if that "6%" is for centerfielders, outfielders or all fielders who ever had to make a run that far. If it's all outfielders/fielders, it's pretty useless in evaluating a centerfielders defense. There are about a dozen centerfielders who could have made that play, but probably not a half dozen corner fielders who could have. Which of course brings the question does statcast differentiate between forward, left, right and back in their stats or is it just a broad range if so that seems to be a missed opportunity(and it's possible the data just isn't there yet for them splice it more than they are right now, but that seems unlikely after a couple of years)
   30. jacksone (AKA It's OK...) Posted: August 14, 2018 at 03:34 PM (#5726683)
And it also makes me wonder if that "6%" is for centerfielders, outfielders or all fielders who ever had to make a run that far.


I am fairly certain that corresponds to the % of times a ball with that hangtime, requiring a fielder to run that distance, gets caught. Position does not come into play. Individual players are then compared to each other by reviewing the % of catches the player makes at varying difficulties, also taking into account the number of chances the player has.

Based on the statcast rankings, Bradley is 6th among all fielders with 10 outs above average. He adds 5% catch rate to the expected catch % for all the balls hit to him (he makes 91% of catches, when the average would be 86%) - Harrison Bader is apparently a freakishly good outlier at an added 10% catch rate!, he just doesn't have the # of chances.
   31. villageidiom Posted: August 14, 2018 at 03:57 PM (#5726711)
It does seem wrong that JBJ gets penalized for otherwise doing well. It's an error on him, but wouldn't be on most any other CF, simply because they wouldn't even be there to begin with.
See my #11. It's a 62% chance of being caught, not 6%. Most CF would have gotten there and made the play.
Echoing some of the first comments, I'm not buying that there was a 6% chance of making this play.
See my #11. It's a 62% chance of being caught, not 6%. Most CF would have gotten there and made the play.
2) If you're just describing what happened, who cares?
Do you ever check the scores of yesterday's games? They're just describing what happened. How about the standings? They're just describing what happened.
And it also makes me wonder if that "6%" is for centerfielders, outfielders or all fielders who ever had to make a run that far.
See my #11. It's a 62% chance of being caught, not 6%. Most CF would have gotten there and made the play. That aside, it's based on how far the fielder had to move, which direction he had to move, and how much time he had to do it. IIRC it's also split into infield vs. outfield, but I haven't found where I think I'd read that.
   32. Carl Goetz Posted: August 14, 2018 at 04:06 PM (#5726721)
"Do you ever check the scores of yesterday's games? They're just describing what happened. How about the standings? They're just describing what happened. "

I was specifically replying to SoSH's question:
"If you don't call this an error, you're basically limiting the number of errors a good outfielder can make compared with a less rangy fielder?"

I was not arguing against describing what happened.
   33. Carl Goetz Posted: August 14, 2018 at 04:15 PM (#5726730)
"See my #11. It's a 62% chance of being caught, not 6%. Most CF would have gotten there and made the play."

So to clarify what this means; the theoretical average centerfielder would make this play 62% of the time?
   34. villageidiom Posted: August 14, 2018 at 05:04 PM (#5726781)
So to clarify what this means; the theoretical average centerfielder would make this play 62% of the time?
The theoretical average outfielder starting 80 feet away from where a play could be made on the ball, having 4.2 seconds to cover that territory (from the moment the pitcher released the ball), and running in that particular direction (to his left and back a little bit, as opposed to running in or running back), made the play 62% of the time.

What I'm not sure of is whether "outfielder" is restricted to just CF or all outfielders, and what I was trying to confirm is that it really is "outfielder" and not just "fielder". There are a few flares back of 1B for which a 2B would face the same conditions; I'm pretty sure they're not included but as I mentioned I haven't been able to trace back to confirm that.
   35. villageidiom Posted: August 14, 2018 at 05:06 PM (#5726784)
Which of course brings the question does statcast differentiate between forward, left, right and back in their stats
They do. They originally didn't, but now they include direction in the determination of catch probability.
   36. Zach Posted: August 14, 2018 at 05:15 PM (#5726792)
I think the difference between the two probabilities has to do with the defender's positioning. Bradley is playing short and to the pull side, and the ball is hit medium length to the opposite field. Just looking at the ball coming off the bat, it doesn't strike you as a certain hit. It's well short of the wall, so that's not a factor. So a metric built on the sector of the field and hang time will say that's a routine play. The only way to get 6% is to take into account the distance Bradley has to run.
   37. Zach Posted: August 14, 2018 at 05:27 PM (#5726798)
The Ordinary Effort part of the rule is clearly a subjective decision.

It is and it isn't. The scorer is the one making the call, but he's being asked to apply an objective standard. So he's not allowed to say "I don't give errors on running catches," or "I don't think it's possible to make an error where you have to run 50 feet to get to the ball." Instead, he's being asked to say "a fielder of average skill in this league should catch a ball that hits him in the glove like that."
   38. Baldrick Posted: August 14, 2018 at 07:05 PM (#5726870)
Do you ever check the scores of yesterday's games? They're just describing what happened. How about the standings? They're just describing what happened.

This is a very bad analogy.

If someone compiles a list of times that each player scratches himself during a game, that would also just be a description of what happened, but surely you'd acknowledge that such a list would not be very interesting to people trying to have a conversation about which players are contributing more and less to the team' success?
   39. villageidiom Posted: August 15, 2018 at 08:46 AM (#5727035)
The only way to get 6% is to take into account the distance Bradley has to run.
Another way to get 6% is to drop the 2 out of 62%. See my #11.

I appreciate that a lot of people are attempting to understand what they've been told is a 6% catch probability, because (I hope) of a desire to give the benefit of the doubt to an advanced statistic in spite of what their eyes tell them. Statcast actually says it's a 62% catch probability. You can look it up. I did. I linked to it in #11.

This is a very bad analogy.

If someone compiles a list of times that each player scratches himself during a game, that would also just be a description of what happened, but surely you'd acknowledge that such a list would not be very interesting to people trying to have a conversation about which players are contributing more and less to the team' success?
After "This is a very bad analogy" you need a colon, not a period.

Failure to catch a catchable ball - the reason why this play was scored an error - is a clear demonstration of contribution (or failure thereof) to team success. The scores, and the standings, are measures of team success. It's all relevant to the discussion. You might not find it interesting to discuss, but your lack of interest in matters of relevance is not my concern.

I don't have the breakdown of the 62% catch probability, but I'm going to guess it's something like a combo of {62% of fielders have enough range to get into position to catch that ball} and {fielders in position to catch the ball actually catch it near 100% of the time}. Bradley failed at the ~100% part of the equation. This is more than adequately described as an error. Not to describe it that way, on the notion that most* other outfielders wouldn't have gotten there in the first place and it would have landed as a hit, is like saying we shouldn't describe a 5-star catch as an out under the same pretense that other outfielders wouldn't have gotten there to make the play.

If what y'all are saying is that the error statistic, either in definition or in application, is an imperfect measure, you'll get no argument from me. But it is a perfect record of the actual result of the BIP, in this case. It might not be satisfactory for projection, and it might not feel like a satisfactory explanation of what led to the result, but the fact is the ball hit the ground because Jackie Bradley, Jr. failed to secure a ball that his glove was in position to secure. If "error" seems inadequately simple as a label for a result, then so is "hit", or "double", or "double play", or "sacrifice hit", or "passed ball", or "flyout", or "strikeout", or "caught stealing".

* Using the 6% probability, which is not actually what happened.

You must be Registered and Logged In to post comments.

 

 

<< Back to main

News

All News | Prime News

Old-School Newsstand


BBTF Partner

Dynasty League Baseball

Support BBTF

donate

Thanks to
BDC
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogOT - NBA Thread (2018-19 season kickoff edition)
(2140 - 2:03am, Nov 13)
Last: Athletic Supporter is USDA certified lean

NewsblogOTP 2018 2018 Nov 5 - The Trump dilemma hits the Boston Red Sox amid a wave of White House boycotts in sports
(2342 - 2:01am, Nov 13)
Last: Gonfalon Bubble

NewsblogOT - November* 2018 College Football thread
(165 - 1:28am, Nov 13)
Last: there isn't anything to do in buffalo but 57i66135

Hall of MeritMock 2018 Today’s Game Hall of Fame Ballot
(29 - 11:21pm, Nov 12)
Last: Dr. Chaleeko

NewsblogNo MLB rule changes likely until eve of spring training
(28 - 11:16pm, Nov 12)
Last: PreservedFish

NewsblogJoe Mauer Retires After 15 Seasons
(58 - 10:40pm, Nov 12)
Last: Sweatpants

NewsblogOT - Catch-All Pop Culture Extravaganza (November 2018)
(106 - 9:32pm, Nov 12)
Last: ERROR---Jolly Old St. Nick

NewsblogRed Sox: Bill James “is not an employee nor does he speak for the club”
(180 - 8:47pm, Nov 12)
Last: PreservedFish

NewsblogDodgers release Utley to facilitate retirement
(42 - 8:27pm, Nov 12)
Last: cardsfanboy

NewsblogReport: Nationals Ownership Rejected Bryce Harper Trade to Astros
(21 - 8:06pm, Nov 12)
Last: Adam Starblind

NewsblogOT: Soccer Thread (2018-19 season begins!)
(1189 - 7:26pm, Nov 12)
Last: Mefisto

NewsblogIt's not just the worst teams not spending money in baseball free agency — it's the richest, too
(41 - 7:02pm, Nov 12)
Last: the Hugh Jorgan returns

NewsblogOT Gaming: October 2015
(914 - 3:18pm, Nov 12)
Last: Zonk is Just the Right Amount of Wrought

NewsblogMLB owners expected to unanimously approve extension for Rob Manfred next week, report says
(14 - 2:44pm, Nov 12)
Last: shoewizard

Gonfalon CubsNow what?
(216 - 2:28pm, Nov 12)
Last: What did Billy Ripken have against ElRoy Face?

Page rendered in 0.4236 seconds
46 querie(s) executed