Poll: VAR or No VAR?
- Thread starter BaZ87
- Start date
I guess having a camera track the last defender might be a bit tricky because it has to distinguish who is a defender / attacker / official and whether they are actually on the pitch or not (i.e. not an assistant ref / sub etc) and having to change velocity quite rapidly as players move around in different directions. You'd also have to potentially customise it for different grounds as some have slopes just off the pitch etc.
I've not looked into this subject too deeply but the impression I'm left with is that perhaps there are too many 'marginal' decisions getting overruled rather than just 'clear and obvious errors'. If you take cricket LBWs basically they have the concept of umpires call whereby if ball tracking suggests the ball will barely clip the stumps/bails then the original decision stands. Occasionally people feel it might not be 100% but generally it is accepted and certainly no worse than if they were overruling the umpire on such decisions. To be fair the scenarios you get in football are a lot more complex and unpredictable but the ethos remains. If it takes more than 30s of people looking at the VT then presumably it can't be a 'clear and obvious' error.
The one downside to VAR is that annoying period after a 'goal' that subsequently gets disallowed where fans can't celebrate properly, we had a goal disallowed for a push or something once and it is a bit of a let down. But it's worth damaging the spectacle if it improves decision making.
I've not looked into this subject too deeply but the impression I'm left with is that perhaps there are too many 'marginal' decisions getting overruled rather than just 'clear and obvious errors'. If you take cricket LBWs basically they have the concept of umpires call whereby if ball tracking suggests the ball will barely clip the stumps/bails then the original decision stands. Occasionally people feel it might not be 100% but generally it is accepted and certainly no worse than if they were overruling the umpire on such decisions. To be fair the scenarios you get in football are a lot more complex and unpredictable but the ethos remains. If it takes more than 30s of people looking at the VT then presumably it can't be a 'clear and obvious' error.
The one downside to VAR is that annoying period after a 'goal' that subsequently gets disallowed where fans can't celebrate properly, we had a goal disallowed for a push or something once and it is a bit of a let down. But it's worth damaging the spectacle if it improves decision making.
I used Liverpool as an example of one of the flaws in these tables. How many other sides have won games despite a VAR decision going against them? You can't say a side has benefitted more from VAR simply because they've still gone on to win a game despite a decision going against them. Also as HangTime points out, you cannot assume that nothing else would have changed if these decisions weren't made.
The only measure that has any meaning is the net number of VAR decisions that have gone in favour or against a side and as I said, Brighton have had the most with Utd joint 2nd. Brighton and Utd, along with Palace and Watford are also the only sides that have had a VAR decision go in their favour that the PGMOL have since admitted were the wrong decisions too.
I'm done with VAR, I'd much rather things were how they were before. Wrong decisions all over the place and farcical offsides. The VAR ref who disallowed one of our goals when Chambers got fouled in the box had never reffed a premier league game. How is he overruling anyone??
Completely agree with this. I don't know if/how anyone can be defending the system as-is. It clearly needs work to fix the issues.
Perhaps the answer to to look at the example of speeding.
You don't get busted for doing 61 in a 60 zone, but you would get busted for doing 71 in a 60 zone.
Some decisions might still be "wrong" (ie detected doing 71 when you were doing 70), but you're unlikely (more unlikely) to be caught speeding when you were doing <60.
I do understand the argument that that is all of the bad of VAR (delays) with none of the good (accuracy), but I don't see improved accuracy with tons of disallowed goals as better than less accuracy with marginal goals allowed.
Ultimately, I believe it's totally crap to be disallowing marginal goals all the time. That comes down to preference, I guess. I prefer that the attacker has the benefit of the doubt. You need to be properly offside not marginally offside.
Last edited:
Offsides or penalty/red card decisions? We are getting correct offside decisions, we just don't like the offside rule. VAR for pens and red cards is a lucky dip though - you see a very soft foul given by VAR like the City penalty vs Wolves but then you see a nailed on penalty not given the next game.
No, I don't accept that.
If the operator moves his marker a pixel or two off the correct spot, or if he can't identify the correct spot on the monitor to the precise pixel, the system clearly allows wrong decisions to be made, in the most marginal offsides.
Anything to support this having happened?
A linesman might blink at the moment somebody plays a pass, we should get rid of linesmen.
Yeah a knowledge of HUDs and HIDs.
All positions being necessarily being resolved to pixels on the screen, by definition. A knowledge of pixels and sub-pixels, and how converting linear space into discrete space and discrete pixel coords inherently must lose accuracy.
Basic maths, in other words. The operator has to choose a pixel to move his line/marker.
If the offside is truly marginal (we've talked about 1mm or 2mm differences), that could translate to half a pixel on the screen. But you can't click between two pixels. You can only choose the one pixel or it's neighbour.
This is why a system with a human operator moving lines/choosing points on a discrete VDU cannot be accurate, in the most marginal situations, where the actual location of an element (a foot, arm) is between two pixels.
Also blowing up an image (digital zoom) resulting in less certainty, or interpolation (guesswork). So it's not like they can even just zoom into to make a more accurate determination.
The very act of asking the operator to mark a location on a screen means that you cannot be 100% accurate, and must mathematically only be able to resolve positions to within a margin of error, I'm afraid.
I wasn't asking you to explain why it's possible that they could make a mistake, I asked whether you had any evidence to suggest that they had selected the wrong frame or pixel. To my knowledge none of these close calls have been anything like 1 or 2mm on/offside. We exaggerate when we see the lines drawn on the TV pictures and as somebody explained before, the lines are widened so it's clearer for TV purposes, the problem with this is it makes close calls look even closer than they are. These close calls are likely to have been several cm's offside.
Is the technology 100% perfect? No. Is it 100x more accurate than a human linesman? Yes.
Based on the most accurate measure we have, the decisions have been correct. Again, the issue is people don't like the law.
Is the technology 100% perfect? No. Is it 100x more accurate than a human linesman? Yes.
Based on the most accurate measure we have, the decisions have been correct. Again, the issue is people don't like the law.
But my point is that these decisions are not mathematically provably correct in the most marginal situations.
They rely on a discrete display (a grid of pixels) and a human operator making an interpretation of what he is seeing on that discrete display.
A potentially more accurate system would be tracking sensors worn by the players (although likely to have its own issues).
But when you are making such marginal offisdes (how can you say they aren't 1mm/2mm btw? how can you prove that?), the current tech cannot mathematically provably provide a correct offside/not offside answer.
So to list accuracy as a plus point for VAR is not in the strictest sense true.
That is why in all models and most branches of statistics, anything involving measurements (etc), margin of error is an important consideration. The margin of error is always disclosed, so that people can decide if the results lie inside or outside the margin of error, and whether or not to give them meaning.
A marginal VAR decision that lies within the margin of error could be correct or it might be incorrect. Inside the margin of error, you do not have mathematical certainty.
VAR is not "accurate" all the time. Even with the best operators.
They rely on a discrete display (a grid of pixels) and a human operator making an interpretation of what he is seeing on that discrete display.
A potentially more accurate system would be tracking sensors worn by the players (although likely to have its own issues).
But when you are making such marginal offisdes (how can you say they aren't 1mm/2mm btw? how can you prove that?), the current tech cannot mathematically provably provide a correct offside/not offside answer.
So to list accuracy as a plus point for VAR is not in the strictest sense true.
That is why in all models and most branches of statistics, anything involving measurements (etc), margin of error is an important consideration. The margin of error is always disclosed, so that people can decide if the results lie inside or outside the margin of error, and whether or not to give them meaning.
A marginal VAR decision that lies within the margin of error could be correct or it might be incorrect. Inside the margin of error, you do not have mathematical certainty.
VAR is not "accurate" all the time. Even with the best operators.
That's why on marginal calls where the technology for numerous reasons may not be 100% accurate I think it's better to just go with the on field decision, they should just ban MSPaint and look (with the naked eye) for "clear and obvious" errors. The way VAR is being operated now we'd might as well just scrap linesmen as they're being undermined all of the time due to the way offsides are being intensely scrutinised unlike nothing else in football. Pulling and holding is a foul by the letter of the law but referees ignore it at corners and free kicks ALL of the time, why isn't VAR intervening in those situations if enforcing the rules at any cost to the sport is the way we're going? 20 penalties a match and 30mins reviewing things until players learn sounds fun to me.
I've not voted as I think VAR can be useful if it's used sparingly to correct serious mistakes, the problem as things stand now is it's taking over the officiating and causing just as much annoyance/grievance as before we had it. I still think a challenge system is best to limit its involvement and so that players can actually celebrate goals unless the other side ask for a review, amongst other things. Nearly all goals are being checked automatically as things stand and it's sucking the enjoyment out of scoring.
I've not voted as I think VAR can be useful if it's used sparingly to correct serious mistakes, the problem as things stand now is it's taking over the officiating and causing just as much annoyance/grievance as before we had it. I still think a challenge system is best to limit its involvement and so that players can actually celebrate goals unless the other side ask for a review, amongst other things. Nearly all goals are being checked automatically as things stand and it's sucking the enjoyment out of scoring.
Last edited:
Well it looks like everyone who would rather defer to the less accurate linesman for tight calls will be seeing their preference enacted and IFAB will be reissuing guidance in a couple of months to say VAR should only be used for clear and obvious in ALL circumstances, including offside.
I have my reservations whether this will be any better, you'll correct all the calls that are yards offside/onside but anything closer than that and we're just going to defer to the linesman and hope his guessing game is up to par.
I have my reservations whether this will be any better, you'll correct all the calls that are yards offside/onside but anything closer than that and we're just going to defer to the linesman and hope his guessing game is up to par.
This is mad. You cannot use VAR for offsides by just looking without the hawkeye technology. Camera angles will mean that just looking at a replay with the naked eye could potentially be even less accurate than the linesman making the decision in real time.
Dale Johnson does a weekly thing on twitter re VAR and he made the point today about how different camera angles can make the same call look completely different. He used the Harry Kane offside call vs Chelsea in last seasons League Cup as an example. The old VAR tech that was being used in last seasons Cups wasn't the same as the Hawkeye tech being used now, it was effectly using the naked eye from the replays with a straight line across the pitch - that day VAR decided that Kane was onside but post match Chelsea released an image from one of their in-house cameras that showed Kane to be offside. Another example would be Salah's goal vs City this season - just from looking at the replay numerous people said Salah was offside but the hawkeye tech which takes into account the camera angle showed he was comfortably onside. In that ocassion Salah was given onside but under your scenario, had the lino made the wrong call then VAR would have given him off when he was clearly on.
As for the stuff about a margin of error or giving the benefit of doubt to the on field officials decisions, it's just crazy. We're going to end up with all the bad side of VAR without more accurate decisions.
edit: In addition to the above, the back page of the Times has a story about FIFA supposedly considering changes to the offside law so that the attacker only had to have part of his body level with the defender to be onside.
I sincerely hope this never happens. It will do more damage to the game than anything we've seen before. Football will become 10x more defensive than it's ever been with teams camped on the edge of their 18 yard box to stop runs in behind.
Last edited:
Well let's face it the current situation is not sustainable.
I've heard the argument for it to stay as it is, made about as well as it could be made, but it really is intensely frustrating as things stand. It's not football. It's become farce.
Let's hope the new arrangement helps keep things fair but also fun.
The worst thing is that due to the slow camera frame rates in use, it's possible that a defender running one way and an attacker another, the distance between them can be 30cm per frame.
Permabanned
Being a Livarpool fan on the whole I am liking VAR. I just dont like the time it takes.
Bit more tweaking it should be fine.
Bit more tweaking it should be fine.
This is based on two players sprinting in the opposite direction and 50 FPS cameras. A) How often do we see this happen? B) How accurate do you think a linesman's call is in real time in this scenario?
Even with the limitations of the current technology, it's 100x more accurate than a human making the decision in real time. I've seen a few people claim that VAR are actually using much higher quality cameras than we're aware of but there's nothing concrete on that however one thing is for certain, they're coming and any margin of error will be much smaller by next season.