Liverpool vs Wolves - Premier League (19/20 - Game 19)

[Exiled_red]

We've benefited from it today, but I hate the way offsides are being measured right now. I know their cameras are higher definition that they can show on TV but unless we have technology that can tell conclusively, we cannot measure to that level of accuracy.

If it isn't clear and you need to pull out those tools to measure, just call it onside.

The ref giving handball against Lallana is why we need VAR, they are ****ing clowns. The general display from the ref today has been really poor. They can't be trusted to come to the correct decision, so we need to make this **** work.

The issue with this is determining exactly when the ball is played, this doesn't necessarily correspond to a frame of the VAR camera and players are moving relatively quickly. I don't believe the system is accurate enough to make such tight calls.

[danperkins]

Bollocks - by any normal definition, that was level.

Was the error by the linesman clear and obvious? Of course not.

The width of a pixel on a computer screen is not within any reasonable margin of error considering the technology available and those operating it with mouse clicks.

We've got away with it, which great. But goals like that should stand. Or change the rules to be governed by pixels accordingly.

Absolutely wrong

[Fuzzy]

Btw, if our goal was scored against us, we'd be going mental about that VVD handball and how it wasn't looked at properly at all just saying.

[Exiled_red]

Btw, if our goal was scored against us, we'd be going mental about that VVD handball and how it wasn't looked at properly at all just saying.

They said at HT that it was looked and and it was inconclusive as to whether it was handball or not

[Exiled_red]

Wijnaldum heads over from a corner

[dom9]

You're wrong. Offside is offside.

How big an error would you like them to introduce in the name of fairness?

The people developing the software would be better placed to answer that than me or you.

But there are many variables to consider.

1. The timing of screen shot (when the ball was kicked). How is that decided? And by whom? Is the frame rate fast and accurate enough to determine this accurately?

2. The manual manipulation of the grid lines by a person. Is the image a high enough resolution to accurately determine how far along the grid each body part is?

3. In the case of subjective body parts being the cause of offside, how does a VAR ref decide which pixel on the screen represents part of say an armpit, and which one is part of an arm? It's guesswork.

4. What happened to being level. Is that a thing of past, consigned to history by the arbitrary use of the above?

5. Clear and obvious - we keep hearing that only clear and obvious mistakes should be overturned. That happened correctly for our goal. Not for Wolves goal.

Margins of error and probabilities absolutely have to be taken into account with any algorithms. No algorithm is perfect. Any mathematician or computer scientist will tell you that.

[RichC]

Bollocks - by any normal definition, that was level.

Was the error by the linesman clear and obvious? Of course not.

The width of a pixel on a computer screen is not within any reasonable margin of error considering the technology available and those operating it with mouse clicks.

We've got away with it, which great. But goals like that should stand. Or change the rules to be governed by pixels accordingly.

We’ve benefited from one today and lost out on others earlier in this season but that’s not a “clear and obvious error”.

Implementation of VAR by the PL is awful

[Zapater]

Btw, if our goal was scored against us, we'd be going mental about that VVD handball and how it wasn't looked at properly at all just saying.

The same people defending it would be the most mental in their protests. Although our goal was definitely legal and rightfully stood.

It is pretty hilarious that their goal was overruled considering the scale of the celebration.

[Fuzzy]

They said at HT that it was looked and and it was inconclusive as to whether it was handball or not

Yeah I know, and I don't think that answer would please us if it was on the other end, that's all I wanted to say

Btw Hendo is playing his 308th PL game since 2010, which makes it the most of any player this decade.

[Buzzo]

Hendo has been very good again today. Patrolling the midfield.

[Irishnev]

We’ve benefited from one today and lost out on others earlier in this season but that’s not a “clear and obvious error”.

Implementation of VAR by the PL is awful

I think the fact that we all disagree on the VAR implementation clearly shows there are issues with it. According to the current rules he was offside as was Pukki but both seem 'harsh'. I do think the FA need to fine tune it somehow and make it more transparent- right now the red lines etc make it open to criticism

[Fuzzy]

Lallana has played really well, but I don't think he'll be able to go much longer. He's already doing that 'head-down' thing when he tries to sprint. Strangely I think I only saw Jones, Elliott and Williams warming up, and no Keita, Origi or Milner.

[Roboklopp]

Done well Lallana I think



FFS

[Exiled_red]

The people developing the software would be better placed to answer that than me or you.

But there are many variables to consider.

1. The timing of screen shot (when the ball was kicked). How is that decided? And by whom? Is the frame rate fast and accurate enough to determine this accurately?

2. The manual manipulation of the grid lines by a person. Is the image a high enough resolution to accurately determine how far along the grid each body part is?

3. In the case of subjective body parts being the cause of offside, how does a VAR ref decide which pixel on the screen represents part of say an armpit, and which one is part of an arm? It's guesswork.

4. What happened to being level. Is that a thing of past, consigned to history by the arbitrary use of the above?

5. Clear and obvious - we keep hearing that only clear and obvious mistakes should be overturned. That happened correctly for our goal. Not for Wolves goal.

Margins of error and probabilities absolutely have to be taken into account with any algorithms. No algorithm is perfect. Any mathematician or computer scientist will tell you that.

My thinking on this is that the biggest source of error comes from the timing, the fastest VAR camera is 120 frames per second, meaning that there is about 8 milliseconds between frames. The exact moment that the ball is played almost certainly won't be caught by the camera so the error is determined by the speed that the players are moving at. The fastest PL player has been clocked at ~33km/h if we assume a speed of ~20km/h then between frames that corresponds to a movement of ~5cm. The error would be about half of this (offset by the choice of which frame to use). So for these speeds which are not at all unresonable (as it really represents the combined speed of the attacker running in behind and the defender stepping up) the error would be ~2.5cm or 1 inch, if we go for faster speeds the error is bigger.

This is assuming everything else is either done accurately or the margin of error is insignificant.

IMO the easiest way to offset this under the current rules is that you have to be offside in the two frames either side of the ball being played for the player to be deemed offside, I don't know if they do it like this though...

[dom9]

My thinking on this is that the biggest source of error comes from the timing, the fastest VAR camera is 120 frames per second, meaning that there is about 8 milliseconds between frames. The exact moment that the ball is played almost certainly won't be caught by the camera so the error is determined by the speed that the players are moving at. The fastest PL player has been clocked at ~33km/h if we assume a speed of ~20km/h then between frames that corresponds to a movement of ~5cm. The error would be about half of this (offset by the choice of which frame to use). So for these speeds which are not at all unresonable (as it really represents the combined speed of the attacker running in behind and the defender stepping up) the error would be ~2.5cm or 1 inch, if we go for faster speeds the error is bigger.

This is assuming everything else is either done accurately or the margin of error is insignificant.

Yep.

And then you have humans manually manipulating the gridlines, adding further subjectivity to a flawed process.

Use it as a guide, but not as the final arbitrator.

BTW, if anyone still disagrees with this, then by all means state why.