FA Cup Semi Final | Coventry City v United

[SkyBlueFlux]

It must be said that the offside tech is there as we've seen it work accurately in major international tournaments and even the Club World Cup where it confirmed a very close offside call in one of our games within a handful of seconds.

I also believe the current format of drawing lines is an issue of process and execution, since for the 50/50 decisions, the VAR can basically choose if an incident is onside or offside by moving the lines by a pixel. Nobody can ever prove otherwise.

I think the tech is almost there with semi-auto offsides but it's still not quite good enough for me. I think a core underlying principle that should have been packaged with VAR is that any technology used should in no way alter the way the game plays out. If it does, then it is fundamentally flawed because the technology is interfering with open play. You don't notice good technology, it should assist but not interfere.

With semi-auto offside, I agree it is fast, but you still have linesmen allowing play to go on until the decision system kicks in. That means potential goals being scored when a linesman without technology would have otherwise flagged before play has even developed. The fact these two outcomes are different is unacceptable to me.

The ideal would be a system similar to semi-auto but it instantaneously pings a message to the ear of the linesman, something like "10 Off", and then all he has to do is look to see if number 10 is interfering and flag as per usual. To all of us sat at home we don't even notice the tech is happening because it all looks exactly like it did before VAR was a thing. With the exception that they can render some fancy graphics to show the offside after the fact. It needs to be effectively instantaneous though, like net calls in tennis. I think this is possible, I really do, but whether we're there yet I'm not sure. Goal line tech is the perfect example of a system like this working as intended.

PGMOL when implementing VAR should have set out core principles, it would not have been hard to come up with something like:
1. The technology must provably improve decision-making.
2. The technology should not change in-game outcomes or the way the game is played.
3. The fans should understand the decisions made by the technology through transparent communication at all times.

They've managed to mess up 2 and 3 monumentally. 1 is debatable, I would suggest it's improved accuracy in offsides generally, and probably avoids most huge errors, but it has done pretty much nothing for most contentious red cards/handballs and in some cases it has taken us backwards because referees now don't give decisions hoping VAR will bail them out.

 

[halfcenturyup]

I think the tech is almost there with semi-auto offsides but it's still not quite good enough for me. I think a core underlying principle that should have been packaged with VAR is that any technology used should in no way alter the way the game plays out. If it does, then it is fundamentally flawed because the technology is interfering with open play. You don't notice good technology, it should assist but not interfere.

With semi-auto offside, I agree it is fast, but you still have linesmen allowing play to go on until the decision system kicks in. That means potential goals being scored when a linesman without technology would have otherwise flagged before play has even developed. The fact these two outcomes are different is unacceptable to me.

The ideal would be a system similar to semi-auto but it instantaneously pings a message to the ear of the linesman, something like "10 Off", and then all he has to do is look to see if number 10 is interfering and flag as per usual. To all of us sat at home we don't even notice the tech is happening because it all looks exactly like it did before VAR was a thing. With the exception that they can render some fancy graphics to show the offside after the fact. It needs to be effectively instantaneous though, like net calls in tennis. I think this is possible, I really do, but whether we're there yet I'm not sure. Goal line tech is the perfect example of a system like this working as intended.

PGMOL when implementing VAR should have set out core principles, it would not have been hard to come up with something like:
1. The technology must provably improve decision-making.
2. The technology should not change in-game outcomes or the way the game is played.
3. The fans should understand the decisions made by the technology through transparent communication at all times.

They've managed to mess up 2 and 3 monumentally. 1 is debatable, I would suggest it's improved accuracy in offsides generally, and probably avoids most huge errors, but it has done pretty much nothing for most contentious red cards/handballs and in some cases it has taken us backwards because referees now don't give decisions hoping VAR will bail them out.

Why do you think SAOT will be more accurate than the current system?

 

[SkyBlueFlux]

Why do you think SAOT will be more accurate than the current system?

I don't think I said that, did I? I just said it will be faster which the evidence shows it absolutely will be.

Importantly, it might not be any more precise but it will at least be objective. It's an algorithm, not a human being moving pixels about on a screen. Goal line technology has a margin of error but everybody just accepts that and nobody questions it because there's no human being behind the wheel to influence it. This is really no different.

It has a margin of error, like all technology, which should be carefully measured by the technologists and independently verified (Hawkeye do this with all their products). They should take that into account in its implementation and only call offsides that fall outside of that margin of error. Common sense really. Not that PGMOL or IFAB are particularly great at enacting common sense policies.

 

[halfcenturyup]

I don't think I said that, did I? I just said it will be faster which the evidence shows it absolutely will be.

Importantly, it might not be any more precise but it will at least be objective. It's an algorithm, not a human being moving pixels about on a screen. Goal line technology has a margin of error but everybody just accepts that and nobody questions it because there's no human being behind the wheel to influence it. This is really no different.

It has a margin of error, like all technology, which should be carefully measured by the technologists and independently verified (Hawkeye do this with all their products). They should take that into account in its implementation and only call offsides that fall outside of that margin of error. Common sense really. Not that PGMOL or IFAB are particularly great at enacting common sense policies.

Fair enough.

I did wonder if the goal line cameras have a higher fps? It's only two cameras, so I was hoping they are more accurate than the cameras used for offside. If not, those beautiful graphical representations don't mean much.

The IFAB VAR protocol, btw, applies clear and obvious to offsides as well. It's just FIFA, UEFA and the PL that try to tell us 100% accuracy is the way to go.

 

[SkyBlueFlux]

Fair enough.

I did wonder if the goal line cameras have a higher fps? It's only two cameras, so I was hoping they are more accurate than the cameras used for offside. If not, those beautiful graphical representations don't mean much.

The IFAB VAR protocol, btw, applies clear and obvious to offsides as well. It's just FIFA, UEFA and the PL that try to tell us 100% accuracy is the way to go.

To my knowledge, the goal line cameras are a higher frame rate and almost certainly will be more accurate than any semi-auto offside tech will be able to accomplish, given it's just a harder problem to solve. That said, a semi-auto offside system accurate to within about quarter of a meter is probably better than a potentially biased human being so it doesn't have to be perfect to be an improvement.

Like I said in my original post, I don't think the technology is there yet to be usable, in my view it will only be there when it is 1. Near instantaneous for the layman (maybe <1 second to come to a decision would be sufficient), 2. Only given power that's commensurate with its accuracy. Like DRS in cricket that others have mentioned, don't overturn decisions it's not capable of making with 100% certainty because it's within a margin of error.

 

[halfcenturyup]

To my knowledge, the goal line cameras are a higher frame rate and almost certainly will be more accurate than any semi-auto offside tech will be able to accomplish, given it's just a harder problem to solve. That said, a semi-auto offside system accurate to within about quarter of a meter is probably better than a potentially biased human being so it doesn't have to be perfect to be an improvement.

Like I said in my original post, I don't think the technology is there yet to be usable, in my view it will only be there when it is 1. Near instantaneous for the layman (maybe <1 second to come to a decision would be sufficient), 2. Only given power that's commensurate with its accuracy. Like DRS in cricket that others have mentioned, don't overturn decisions it's not capable of making with 100% certainty because it's within a margin of error.

Had a look. GLT cameras (14 of them!) operate at 320 fps. Even I can't really argue with accuracy too much at that speed. Apparently, they have to guarantee accuracy to within 5 mm and have the result relayed within half a second.

John Stones's Liverpool clearance was 11 mm iirc. Pretty close.

 

[Stoned Rose]

I said it's fair enough to question it, not that the wrong frame was used.

You’ve said it’s offside. You’ve said it’s fair to question it.

Ffs Bob.

 

[youngbob]

You’ve said it’s offside. You’ve said it’s fair to question it.

Ffs Bob.

It's fair to question evolution.