Rules Score Review (ARC) Thread - Lions v Tigers Elimination

[My_Username]

What are your qualificatiosn in this matter? Experience?

Lots of work in a related field.

 

[Chief]

Lots of work in a related field.

Of course.

 

[Meteoric Rise]

Of course.

Nobody on this forum with any claims to knowing for sure has posted anything to cast doubt on what My_Username is posting about this.

In any event, the AFL should be making public why they believe the ARC is the most accurate available option for these cases, so that those with expertise can come forward and confirm or ask informed questions.

But one thing we should know for certain about the AFL is it is in no way safe to assume they have it all sorted in the best possible way. As far as we know until informed otherwise, the cameras and hardware could be totally unsuitable for the task of judging whether the ball travels over the post, and the ARC operators could have no extra expertise than you or I in judging a decision like this.

As matters stand, I would bet that Lynch goal decision in isolation would not stand up to a legal challenge if the AFL needed to prove the accuracy of the method they use for it to do so.

 

[My_Username]

Of course.

I'm not going to post my resume here.

The other (related) point is I can do the maths.

e.g. How big is the area covered by a pixel in a HDTV camera at 100M?
How far does a ball travel between frames on that camera?
etc.

 

[Chief]

Nobody on this forum with any claims to knowing for sure has posted anything to cast doubt on what @My_Username username is posting about this.

Pretty sure they did far as I recall. People with direct knowledge and experience said "this is how it is done, and it is accurate".

Hand waving with "oh its the technology and frame rates and stuff" is ridiculous without proof.

Too bad. The rules were followed. It was a point.

 

[Chief]

The other (related) point is I can do the maths.

e.g. How big is the area covered by a pixel in a HDTV camera at 100M?
How far does a ball travel between frames on that camera?
etc.

It doesn't matter when the the ball is shown over the post. If it was over the post between frames you'd have a point.

 

[Meteoric Rise]

Pretty sure they did far as I recall. People with direct knowledge and experience said "this is how it is done, and it is accurate".

Hand waving with "oh its the technology and frame rates and stuff" is ridiculous without proof.

Too bad. The rules were followed. It was a point.

People in the public cannot supply proof of the accuracy of the system one way or another if the AFL don’t reveal detail about the equipment used and the method used. That is the point I am making, it is for the AFL to convince people they have a system that works better than a goal umpire’s judgement. So far, as far as I am aware, we have no public explanation containing any great detail.

 

[Chief]

it is for the AFL to convince people they have a system that works better than a goal umpire’s judgement.

If you can't understand that multiple camera angles - modern high def TV cameras - are better than one human's eyeballs then you will never be convinced.

 

[Meteoric Rise]

It doesn't matter when the the ball is shown over the post. If it was over the post between frames you'd have a point.

For triangulation - the method we now know is used to make the decision as per Hardwick’s comments - to work, it does matter. You need the ball to be over the post from the 3 different angles at the same point in time.

 

[Meteoric Rise]

If you can't understand that multiple camera angles - modern high def TV cameras - are better than one human's eyeballs then you will never be convinced.

Rubbish, I have posted what would convince me. If the AFL open their method and equipment up to scrutiny and people with the expertise to judge the merits of their system are convinced, then I am convinced. What doesn’t convince me is people like you assuming that just because a system is in use that it works sufficiently well to do the job is purports to do.

 

[My_Username]

It doesn't matter when the the ball is shown over the post.

But was it?

Factor in the error margins, and 'over the post' should be 'umpires call' (i.e. insufficient certainty to overrule).

 

[Chief]

For triangulation - the method we now know is used to make the decision as per Hardwick’s comments

It was OBVIOUS they triangulated the ball's position. Who needed confirmation from the AFL?

Hardwick has professional football knowledge. His knowledge doesn't extend into the domain of looking at a football on a TV screen, it seems.

- to work, it does matter. You need the ball to be over the post from the 3 different angles at the same point in time.

You don't need three cameras. It's not radio triangulation. It's not an abstract mathematics problem.

If you can't see it, you need a third point to establish the direction, eliminating one of the points of intersection.

Point X is 10 metres from point A and 5 metres from point B, you need a distance from point C to eliminate the two points of intersection between the radius around A and B.

What is triangulation?

You've got the post that you can see. The ball that you can see. You've got those points of reference. You don't need a third camera.

The point of attack here, for MU, is that the cameras aren't accurate enough. Everyone else who has any experience says yes they are syncronised and accurate enough for the purpose.

 

[Chief]

Factor in the error margins

What are the error margins?

You have experience, apparently. You know this stuff.

 

[JyeDye]

It was sad that people were still arguing about this a week ago. Not sure what we're at now. Move on.

 

[My_Username]

What are the error margins?

You have experience, apparently. You know this stuff.

I don't know their specs to say exactly, but as a guesstimate based on standard HDTV cameras, I'd say +-40-50cm on all 3 axes.

So I'd want the ball to appear to be (at least) 40cm on the wrong side of the line to overrule the ump. (to keep it simple, say a full ball width to the 'point' side of the post).

Less than that and you're dealing with degrees of probability rather than being certain. (and to go back to what I've been saying all along, I think it was probably a point, but don't think it was certainly so.)

 

[Bird Man]

If you can't understand that multiple camera angles - modern high def TV cameras - are better than one human's eyeballs then you will never be convinced.

professional camera guy here.

Taking into account frame rate and shutter really affect and image. Ideally, the AFL would have a camera pointed directly downwards and shooting at a high frame rate so it can be slowed down for a clear image.

This whole idea that the images are sufficient is bogus. Look at the EPL and NBA.. they have so many cameras pointed around the pitch and they STILL use digital sensors to be 100% certain that a ball crossed the line... ALSO even in the EPL where they have Billions more invested in the game they still can't get onside calls to bang on because the technology is there but it's not viable to be used at every game and stadium

At the end of the day, the Umpire said he thinks its a goal and the video CLEARLY doesn't not make it 100% certain it wasn't... go with the guy who is standing directly under the goal post

 

[My_Username]

professional camera guy here.

Taking into account frame rate and shutter really affect and image. Ideally, the AFL would have a camera pointed directly downwards and shooting at a high frame rate so it can be slowed down for a clear image.

This whole idea that the images are sufficient is bogus. Look at the EPL and NBA.. they have so many cameras pointed around the pitch and they STILL use digital sensors to be 100% certain that a ball crossed the line... ALSO even in the EPL where they have Billions more invested in the game they still can't get onside calls to bang on because the technology is there but it's not viable to be used at every game and stadium

At the end of the day, the Umpire said he thinks its a goal and the video CLEARLY doesn't not make it 100% certain it wasn't... go with the guy who is standing directly under the goal post

Exactly.

As I've said before, just look at hawkeye in the tennis, and the reviews in cricket.

Both are (technically) far simpler to sort out (mostly due to cameras focused on much smaller areas), and both still have their error margins.

Tennis doesn't really advertise it, but it's there...the graphic they show is 'adjusted' to suit the decision rather than the other way around, but their error margins are pretty small anyway (specialized cameras they spend a lot of time setting them up, and they 'only' need to look at a narrow strip that is relatively close).

My favorite is Cricket because they're just honest about it...Ball tracking graphic shows 1/3 of the ball hitting off stump...that's probably out, but because it's in the error margin it's not good enough to over rule, so it's umpires call. Everyone accepts that and they move on.

 

[Fadge]

It was sad that people were still arguing about this a week ago. Not sure what we're at now. Move on.

Yeah, those of us with teams who play finals are concerned a similar event will again directly impact the result of another finals game.

And whether your team is on the right or wrong side of the decision, it is a poor way to decide a finals match.

 

[JyeDye]

Yeah, those of us with teams who play finals are concerned a similar event will again directly impact the result of a finals game.

And whether your team is on the right or wrong side of the decision, it is a poor way to decide a finals match.

Nah I said stop so you have to.

 

[Chief]

professional camera guy here.

Taking into account frame rate and shutter really affect and image. Ideally, the AFL would have a camera pointed directly downwards and shooting at a high frame rate so it can be slowed down for a clear image.

This whole idea that the images are sufficient is bogus. Look at the EPL and NBA.. they have so many cameras pointed around the pitch and they STILL use digital sensors to be 100% certain that a ball crossed the line... ALSO even in the EPL where they have Billions more invested in the game they still can't get onside calls to bang on because the technology is there but it's not viable to be used at every game and stadium

So how does that make them worse than human eyeballs?

At the end of the day, the Umpire said he thinks its a goal and the video CLEARLY doesn't not make it 100% certain it wasn't... go with the guy who is standing directly under the goal post

I can't take it seriously...

 

[Meteoric Rise]

It was OBVIOUS they triangulated the ball's position. Who needed confirmation from the AFL?

Hardwick has professional football knowledge. His knowledge doesn't extend into the domain of looking at a football on a TV screen, it seems.



You don't need three cameras. It's not radio triangulation. It's not an abstract mathematics problem.

If you can't see it, you need a third point to establish the direction, eliminating one of the points of intersection.

Point X is 10 metres from point A and 5 metres from point B, you need a distance from point C to eliminate the two points of intersection between the radius around A and B.

What is triangulation?

You've got the post that you can see. The ball that you can see. You've got those points of reference. You don't need a third camera.

The point of attack here, for MU, is that the cameras aren't accurate enough. Everyone else who has any experience says yes they are syncronised and accurate enough for the purpose.

It wasn’t obvious to me the ARC triangulated their way to the decision, I had never had cause to even think about triangulation before I saw it mentioned on here. I thought upon reading about it that it stood to reason this is what the ARC had based their decision upon. But that doesn’t mean it was the one and only possible way they arrived at the decision.

Hardwick is a footy coach. He should mainly only be interested in understanding how these mechanisms affect the way he instructs his team, if at all.

My lack of understanding of how triangulation works is also immaterial to my post to which you were responding. I understand it well enough to know that it requires multiple cameras to position the ball above the post simultaneously. If that is required it is a pretty easy leap to the next step. The cameras need to be sufficiently synchronised or you cannot be certain they are catching the ball in a certain position at the same time. There are still shots I have seen that cast doubt on whether the 3 cameras are actually synchronised. So if 3 are not synchronised how would we know that 2 are?

So I agree that is where the discussion is now. Why are we left to assume the cameras are more accurate than the naked eye? For all you or I KNOW for certain, the ARC method may not be as accurate as the umpire’s naked eye in this case. If it is for some reason obvious to the AFL that the ARC is more accurate than the umpire’s naked eye then it should be a very easy matter for them to come out and explain why they think that. What do they base that belief on? Don’t you agree nothing valuable can be lost by them doing that?

 

[Soggy2112]

So basically what he was saying that he wasn't swayed by their explanation about triangulation, without coming out and saying as much, in the same way as many on here haven't been swayed by the triangulation argument.

The bottom line is, the vision had to be CONCLUSIVE, and it wasn't.

Hence his continued issue with 'the process'. Which is reasonable and logical.

If the triangulation thing (which I’ll admit I was unaware of before this), is a standard part of their process though, then according to their process they did have conclusive evidence to overturn.

You can argue until the cows come home about whether it was a goal or point, but if the triangulation method is a part of their process then the process was in fact followed correctly, and Hardwick has no right to suggest that it was not.

 

[Fadge]

If the triangulation thing (which I’ll admit I was unaware of before this), is a standard part of their process though, then according to their process they did have conclusive evidence to overturn.

You can argue until the cows come home about whether it was a goal or point, but if the triangulation method is a part of their process then the process was in fact followed correctly, and Hardwick has no right to suggest that it was not.

This is a good point.

But it is still a reasonable question as to whether or not a determination using the triangulation method is 'conclusive'.

Based on the pictures we have seen and the positioning of the cameras, I can't see how it could possibly be deemed 'conclusive'.

 

[Soggy2112]

This is a good point.

But it is still a reasonable question as to whether or not a determination using the triangulation method is 'conclusive'.

Based on the pictures we have seen and the positioning of the cameras, I can't see how it could possibly be deemed 'conclusive'.

It is definitely fair to argue over whether or not it’s conclusive evidence and if it should be a part of the process. It’s one that the AFL should potentially review in the off season as they presumably look to continually improve the ARC.

Frankly it’s beyond my knowledge to know if it is or isn’t. However in this specific example my opinion from watching is it looks like a point to me so I think the ARC got it right.

 

[Brian_45]

So the technical director does it by a visual check?

That might tell you that frame #684357 on camera A and frame #25764 on camera B happened at the 'same time'.

There is sync and there is sync.

At 25fps, there is a frame every 0.04 of a second. If frame A is taken at 0.000 seconds, and frame B is taken at 0.01 seconds, then the cameras are probably regarded as synced, but the frames aren't fully in sync. (you certainly couldn't tell from a visual check)

Mightn't sound like much, but with a ball moving at ~10-15M/s, that 0.01s means the ball has moved 10-15cm between when those 'synced' frames were taken, and considering that a goal post is only 7.5cm wide, that's significant in this context.

Tell me, what duration does that light flash they sync off last for?

This actually doesn't make any sense what you have written, each frame goes for the same duration. There's also something called timecode which syncs and measures how many frames, seconds, minutes, hours have gone by. Every piece of gear in the OB is aligned to that.

The metronome blinks once every second, it's pulse last for a duration of 0.04 seconds, if any of the cameras were out of sync it will be blinking into a different frame and that could be measured by it happening on a different timecode.

I'd say in this case, there is far greater margin for error from the goal umpire rather than so called differences between frames.