Video review (at least in the NFL) has a good answer for this. Clear and convincing. Anything that's borderline will stay with the ruling on the field. Only those situations where the live call missed something are overturned.
Last night's Rams/Seahawks game had a perfect example of this. A late-game pass from Jared Goff was tipped by the intended receiver and hit the ground behind him. The Seahawks' safety picked it up right away but the play was whistled dead (incomplete pass).
Except that's not what happened. To the refs on the field and to us at home, it looked like the ball hit ground. But a slow motion video replay showed that the safety never allowed the ball to touch the turf. He intercepted it.
This was a huge turn of events, and probably changed the outcome of the game. It wasn't borderline or vague. It was just too fast and the distances (between ball and grass) were too small to rely on a real-time judging of the result.
> But here’s an objection: You might think these problems can be avoided if video review is used only to correct “clear and obvious” errors. This is the supposed standard for use of VAR in soccer. Things aren’t so simple, though, because there aren’t just borderline cases. There are borderline borderline cases. That is, it can be vague when the borderline cases start. It isn’t just that there is unclarity, but that there is unclarity about when the unclarity even begins.
I don't think that section invalidates my point. Those borderline-borderline cases can also be dealt with by leaving the call on the field in place. Or not.
What I'm saying is that video review doesn't make the judgements worse. It either illustrates the vagueness, or discovers a concrete fact that was missed during live action judgements.
Just because video review can't solve the Sorites Paradox (or any of the meta-paradoxes at the margins of the excluded middle), doesn't mean that it's useless. What replay review does do is mitigate the fallibility of human judges seeing the action from limited angles in real time.
It seems Pareto superior to still have the replay, though... with no replay, obviously wrong calls exist and borderline calls will go whichever way was called on the field... with replay review, the obviously wrong calls are fixed and the borderline calls still go whichever way it was called on the field.
Borderline borderline doesn't change this... by definition, a borderline borderline call is neither is neither super-right or super-wrong, so a overturn is not a travesty of justice.
No Replay: People get salty because of subjectivity of person.
Replay: People get salty because of subjectivity of technology/rules.
Result: Some people will always be salty.
Why: There is an underlying subjectivity in nature. Nature itself is probabilistic. There should be general agreement on rules of competition with the intention to minimize the degree of contention. The creation/modification of rules to have a higher/lower degree of precision should be measured by the degree of utility they provide in minimizing that contention, as well as degree of attractiveness to convince people to agree to compete with them.
That being said, I'd be more concerned about head injuries than I would be about Replay review for the NFL.
This is a philosophically interesting line of reasoning made no more interesting by applying it to football. The argument for replay review is simple if you look at its history: before replay review was introduced fans were often in the situation where a ref made a call and then they saw a replay that showed the ref was completely incorrect. This is a terrible viewing experience that leads to people hating refs, and ultimately not wanting to watch. It's incredibly stupid for viewers at home to have information available to them, that refs, who control the outcome of the game don't. Replay reviews weren't introduced as a way to eliminate vagueness, they were introduced to eliminate cases where the call is clear upon review, but the ref didn't see it, because flagrantly incorrect calls degrade viewers interest.
Interesting read. I think the fact that these fuzzy, "Serites-susceptible" situations are so hard to grasp for us is also why societal change is so difficult: each individual person feels like they don't make a difference (like the one additional hair being added on the bald head), but the individual decisions are necessary for a critical mass to be reached, to cross the threshold from 'clearly false' through the fuzzy zone into 'clearly true'.
That’s especially true when it leads to an underdog winning a game.
Yes, on the day, that’s unfair to the better team, but in the long term, it makes the sport more popular relative to sports with ‘duller’ refereeing, leading to a net benefit even for that better team. (Edit: one should realize that incentives aren’t perfectly aligned in sports. Clubs compete with other clubs, so they want to win games, but sports leagues such as the NBA and the NFL compete with each other and Netflix, so they want interesting games)
Open problem is to define “doesn’t happen too often”. American football, IMO, needs video review; soccer doesn’t.
I like that the MLS (USA association football league) fines players after the fact by reviewing later. The viewers get the suspense now, but later the players can be suspended for a game or fined for taking a fake dive.
> is a feature because it increases spectator engagement
That sound like the kind of optimization (aiming for audience “engagement”) which gave the media its current culture of clickbait and striving for maximum enragement and polarization. I.e. even though the audience could be said to (in some sense) prefer it, this is not necessarily a direction which you want to allow.
There is another problem of vagueness that the article didn't cover, vagueness in intent of the rule. This can come up when a rule was written assuming a level of vagueness in observation that was removed by the introduction of this video review technology. Tag plays in baseball are one example of this.
A baseball player is out whenever they aren't touching a base and are tagged by a defender holding the ball. There is very little vagueness there. However that specificity is assuming a certain vagueness in observation. Some players are running over 20mph toward bases that are fixed in place and don't have much give. Simple physics results in the players often bouncing off the base momentarily. This bouncing generally isn't something that is visible in real-time with the naked eye so for roughly 100 years everyone who initiated contact with the base before being tagged was ruled safe. Now in comes high def slow motion video to prove that there are fractions of a second in which it is clear the player isn't touching the base. Do you call that player out? The rule itself says yes, but all precedent of officiating says no. Do you go by the text of the rule or by what was expected by the people who wrote the rule? You might end up with a discussion about original intent that would be at home in a debate in the Supreme Court.
Watch a slow motion punch or other collision. At high speeds, the colliding surfaces will often deform faster than they're colliding and / or there is a slight rebound which separates them for a tiny, imperceptible amount of time, but you can see it on high speed camera.
MLB posts a ton of video to Youtube, but the descriptions on those videos are rather minimal. That makes it hard to find a perfect example without combing through hundreds of videos, but this one is close [1]. The call here was more a problem with the transfer from sliding to standing, but his front leg does the "bounce" I was talking about that is hard to notice in the real-time replay at 30 seconds but is obvious in the slow motion replay at 60 seconds. Regardless of whether the problem is the bounce or the transfer, this player clearly "beat" the tag which is in my opinion the spirit of the rule. He would have likely been called safe for 100+ years of baseball history. Yet with video review, he is clearly out based on the letter of the rule.
So just because one cannot reach 100% certainty one should withhold judgement? Bullshit. The ref's job is to make the call. The only difference that replay makes is that now the ref, instead of having a tiny fraction of a second once to get all the information, he can review the replay, in slow motion (in some sports), from different angles before doing his job of making his call. It might not be 100% accurate, but it is likely to be a lot more accurate than before and this proves out in real life. Also, sometimes there are clear, non-debatable errors that need to be fixed. In both football and soccer video review has been a god-send to the games. This article is just bullshit and as a philosophy major underscores the giant problem with a lot of philosophy: it's completely useless and impractical. That isn't to say that there isn't very useful philosophy, but idiotic ideas like this give a bad name to philosophy that tries to deal with real important questions in practical terms.
Maybe tangential but I notice that faith in machines seems to create consensus out of ambiguity (whether appropriate or not).
For example, in Tennis, replay shown by the Hawkeye system is not video but reduced to a computer animation. The animation shows the ball impact with an unreasonable shape and with sharp outlines of infinite resolution of both ball and line on every surface, even grass. Yet the result is accepted with little or no argument. So while ambiguity still obviously exists to the same degree, the final decision has been delegated to software and machinery rather than publicly reviewing the actual event via video.
I find two interesting things about this. 1) apparently having a human review the video and rendering the final verdict would cause more argument than black-box software and 2) Hawkeye has become part of the rules and so some players have found ways of using the psychology involved to their advantage.
15 years ago the referee had the best view of what was happening on the field. Today they have possibly the worst. Sitting on my couch I have real time access to super HD broadcasting, ability to zoom into a blade of grass across the field, multiple viewing angles, slow motion and replay on demand.
Fans don't want 100% accurate rulings, they just don't want obvious preventable errors in refereeing. It's just that the obvious errors are far easier to spot now, and so technology has to catch up.
VAR has certainly not created more confusion in the Italian Serie A (that’s how their first-level football/soccer league is named). Yeah, it isn’t perfect but it certainly highly improved things compared to pre-VAR times, and I say that as an Juventus fan (i.e. a team that has won the title a few years ago based on a blatant referee error, just search for “Muntari goal” on YT). I personally stopped reading the article right at that second sentence where that sweeping and false generalization was made.
The author discusses 3 concepts of vagueness but none of them really seem helpful. To use his example of baldness, if you survey people with images of heads with varying amounts of hair and forced bald/not bald choice, there will be a clear pattern of bald - random - not bald. For a hair amount in the middle, supervaluationism says it's borderline true that it's bald, epistemic reasoning fails because people disagree on whether it's bald, and ontic vagueness seems to be some thought experiment type thing about how even if everybody did agree on whether it's bald they wouldn't necessarily be right.
Fuzzy logic seems much more useful, there you can use the graph and say something like it's 50% true that they're bald.
But it doesn't really help football games, since they still have to make a play and go on. Audience voting on plays seems a little cheesy but probably could work. The rules are arbitrary in the first place, it's a game after all...
After laying out an "ontic vagueness" view, which holds that the world itself is fundamentally vague, rather than the vagueness resulting from an inadequacy of our means of describing the world, piece continues:
> The implications for video review should be obvious here as well. Because if vagueness is metaphysical, then some state of affairs under review might be fully indeterminate. There is no fact of the matter as to whether someone controlled the ball or not. And so looking frame by frame won’t help. There is nothing there for a referee to discover.
Well, if we have "ontic vagueness" and "there is no fact of the matter as to whether someone controlled the ball or not" then what is the status of the referee in facilitating adherence to rules relating to ball controlling?
Since an objective approach is ruled out, is the suggestion that a refereeing praxis based on rolling dice or tossing coins to determine decisions would be just as valid?
Or is it more the case that there is an existing praxis, where the referee applies their accumulated experience and learning to new situations and sincerely attempts to reach a decision which is most beneficial in terms of the nexus of social relations which surround playing the game and making referee's decisions, where "ontic vagueness" might more reasonably and advantageously be applied to the said praxis, accumulated experience, and social relations?
And given that, is there any remaining reason, on the "ontic vagueness" view, to resist the notion that the viewing of replays may be a legitimate or desirable part of a referee's praxis, or that of those who determine what resources are available in the cause of refereeing?
A bit of a rub with the examples. Baldness isn't a binary condition, it's a continuous set of outcomes, ie levels of baldness.
Same with the heap of sand. There's no definite number of grains, but it's reasonable to say a heap occurs once the sand cannot pile higher without getting wider.
Football catches are binary outcomes. There are no "mostly caught" (like mostly bald) cases. Either it was or it wasn't.
But like with the sand, it's hard to put a number on each of the variables. Time of control, amount of wiggle, number of steps etc.
25 comments
[ 3.6 ms ] story [ 49.8 ms ] threadLast night's Rams/Seahawks game had a perfect example of this. A late-game pass from Jared Goff was tipped by the intended receiver and hit the ground behind him. The Seahawks' safety picked it up right away but the play was whistled dead (incomplete pass).
Except that's not what happened. To the refs on the field and to us at home, it looked like the ball hit ground. But a slow motion video replay showed that the safety never allowed the ball to touch the turf. He intercepted it.
This was a huge turn of events, and probably changed the outcome of the game. It wasn't borderline or vague. It was just too fast and the distances (between ball and grass) were too small to rely on a real-time judging of the result.
> But here’s an objection: You might think these problems can be avoided if video review is used only to correct “clear and obvious” errors. This is the supposed standard for use of VAR in soccer. Things aren’t so simple, though, because there aren’t just borderline cases. There are borderline borderline cases. That is, it can be vague when the borderline cases start. It isn’t just that there is unclarity, but that there is unclarity about when the unclarity even begins.
> The motivation for using video review in sports is obvious: to get more calls right.
The writing seems to ignore the very premise, by arguing the existence of edge cases to the edge cases makes any improvement impossible.
Seems the sort of ridiculous navel gazing that earns philosophy a bad reputation.
What I'm saying is that video review doesn't make the judgements worse. It either illustrates the vagueness, or discovers a concrete fact that was missed during live action judgements.
Just because video review can't solve the Sorites Paradox (or any of the meta-paradoxes at the margins of the excluded middle), doesn't mean that it's useless. What replay review does do is mitigate the fallibility of human judges seeing the action from limited angles in real time.
Borderline borderline doesn't change this... by definition, a borderline borderline call is neither is neither super-right or super-wrong, so a overturn is not a travesty of justice.
Replay: People get salty because of subjectivity of technology/rules.
Result: Some people will always be salty.
Why: There is an underlying subjectivity in nature. Nature itself is probabilistic. There should be general agreement on rules of competition with the intention to minimize the degree of contention. The creation/modification of rules to have a higher/lower degree of precision should be measured by the degree of utility they provide in minimizing that contention, as well as degree of attractiveness to convince people to agree to compete with them.
That being said, I'd be more concerned about head injuries than I would be about Replay review for the NFL.
That’s especially true when it leads to an underdog winning a game.
Yes, on the day, that’s unfair to the better team, but in the long term, it makes the sport more popular relative to sports with ‘duller’ refereeing, leading to a net benefit even for that better team. (Edit: one should realize that incentives aren’t perfectly aligned in sports. Clubs compete with other clubs, so they want to win games, but sports leagues such as the NBA and the NFL compete with each other and Netflix, so they want interesting games)
Open problem is to define “doesn’t happen too often”. American football, IMO, needs video review; soccer doesn’t.
I like that the MLS (USA association football league) fines players after the fact by reviewing later. The viewers get the suspense now, but later the players can be suspended for a game or fined for taking a fake dive.
That sound like the kind of optimization (aiming for audience “engagement”) which gave the media its current culture of clickbait and striving for maximum enragement and polarization. I.e. even though the audience could be said to (in some sense) prefer it, this is not necessarily a direction which you want to allow.
A baseball player is out whenever they aren't touching a base and are tagged by a defender holding the ball. There is very little vagueness there. However that specificity is assuming a certain vagueness in observation. Some players are running over 20mph toward bases that are fixed in place and don't have much give. Simple physics results in the players often bouncing off the base momentarily. This bouncing generally isn't something that is visible in real-time with the naked eye so for roughly 100 years everyone who initiated contact with the base before being tagged was ruled safe. Now in comes high def slow motion video to prove that there are fractions of a second in which it is clear the player isn't touching the base. Do you call that player out? The rule itself says yes, but all precedent of officiating says no. Do you go by the text of the rule or by what was expected by the people who wrote the rule? You might end up with a discussion about original intent that would be at home in a debate in the Supreme Court.
https://scholarship.law.upenn.edu/cgi/viewcontent.cgi?articl...
I forgot if people have tried to write rejoinders to this based on other schools of legal interpretation.
[1] - https://www.youtube.com/watch?v=t3J1QEwDmmg&t=30
For example, in Tennis, replay shown by the Hawkeye system is not video but reduced to a computer animation. The animation shows the ball impact with an unreasonable shape and with sharp outlines of infinite resolution of both ball and line on every surface, even grass. Yet the result is accepted with little or no argument. So while ambiguity still obviously exists to the same degree, the final decision has been delegated to software and machinery rather than publicly reviewing the actual event via video.
I find two interesting things about this. 1) apparently having a human review the video and rendering the final verdict would cause more argument than black-box software and 2) Hawkeye has become part of the rules and so some players have found ways of using the psychology involved to their advantage.
Fans don't want 100% accurate rulings, they just don't want obvious preventable errors in refereeing. It's just that the obvious errors are far easier to spot now, and so technology has to catch up.
Fuzzy logic seems much more useful, there you can use the graph and say something like it's 50% true that they're bald.
But it doesn't really help football games, since they still have to make a play and go on. Audience voting on plays seems a little cheesy but probably could work. The rules are arbitrary in the first place, it's a game after all...
> The implications for video review should be obvious here as well. Because if vagueness is metaphysical, then some state of affairs under review might be fully indeterminate. There is no fact of the matter as to whether someone controlled the ball or not. And so looking frame by frame won’t help. There is nothing there for a referee to discover.
Well, if we have "ontic vagueness" and "there is no fact of the matter as to whether someone controlled the ball or not" then what is the status of the referee in facilitating adherence to rules relating to ball controlling?
Since an objective approach is ruled out, is the suggestion that a refereeing praxis based on rolling dice or tossing coins to determine decisions would be just as valid?
Or is it more the case that there is an existing praxis, where the referee applies their accumulated experience and learning to new situations and sincerely attempts to reach a decision which is most beneficial in terms of the nexus of social relations which surround playing the game and making referee's decisions, where "ontic vagueness" might more reasonably and advantageously be applied to the said praxis, accumulated experience, and social relations?
And given that, is there any remaining reason, on the "ontic vagueness" view, to resist the notion that the viewing of replays may be a legitimate or desirable part of a referee's praxis, or that of those who determine what resources are available in the cause of refereeing?
Same with the heap of sand. There's no definite number of grains, but it's reasonable to say a heap occurs once the sand cannot pile higher without getting wider.
Football catches are binary outcomes. There are no "mostly caught" (like mostly bald) cases. Either it was or it wasn't.
But like with the sand, it's hard to put a number on each of the variables. Time of control, amount of wiggle, number of steps etc.