[PICTURES] Usama Mir's wicket of Van Der Dussen in ICC World Cup 2023: Can Hawk-Eye be edited?'

[Major]

The Usama mir wicket, well before they should the wicket, they accidentally ran the following video, and than they corrected it immediately on and got it both umpires call.

later it changed to wickets UMPIRE'S Call

 

[pillionrider]

I reckon a lot of people noticed this and have taken snapshots. An explanation will be expected. Very interesting to hear what this could be.

 

[Kianig89]

Don't forget Mohali 2011 Hawkeye is dubious

 

[oatmeal]

The Usama mir wicket, well before they should the wicket, they accidentally ran the following video, and than they corrected it immiedietly on and got it both umpires call.

I took a pic of it.
View attachment 138333

later it changed to wickets UMPIRE'S Call

The system might allow edits of parameters, but there would be logging of any changes done and it would be a huge scandal if it was changed during the review.

Sometimes the models run multiple times to get a high confidence result, and my guess is that the final result was clipping the bails

 

[SportsWarrior]

it looked suspicious. Live i thought, it was missing leg. May be just different parameters. but if the parameters can be tweaked at run time, then this is a disaster waiting to happen. what is stopping them to adjusting it however they want?

 

[Mahee]

normally No but it will be edited in India's KO matches this tournament.

 

[Saj]

When I first saw it, to me it looked like it was missing the leg stump.

 

[Major]

The system might allow edits of parameters, but there would be logging of any changes done and it would be a huge scandal if it was changed during the review.

Sometimes the models run multiple times to get a high confidence result, and my guess is that the final result was clipping the bails

make sense if they do this. Cause they look at predictin and possibility.

 

[Mamoon]

Imagine the reaction if this was done in favor of an Indian bowler. Absolute absurd decision by the umpire and it seems like Hawkeye was tempered with as well.

Shameful. Can Pakistan win without cheating?

 

[gazza619]

This is cheating of the highest order

 

[oatmeal]

Imagine the reaction if this was done in favor of an Indian bowler. Absolute absurd decision by the umpire and it seems like Hawkeye was tempered with as well.

Shameful. Can Pakistan win without cheating?

I dont get this argument, The DRS is not controlled by Pakistanis, its the broadcaster. and its not one person operating the system, there 's a team with access to it.

In modern softwar applications, everything is logged and audited. You cant just make a change and no one will find out

 

[Major]

Imagine the reaction if this was done in favor of an Indian bowler. Absolute absurd decision by the umpire and it seems like Hawkeye was tempered with as well.

Shameful. Can Pakistan win without cheating?

so if you claim that Hawk eye was edited here so maybe it was edited during Tendu's mohali innings?

 

[MenInG]

I dont get this argument, The DRS is not controlled by Pakistanis, its the broadcaster. and its not one person operating the system, there 's a team with access to it.

In modern softwar applications, everything is logged and audited. You cant just make a change and no one will find out

It's like trying desperately to blame Pakistan for anything!

No sense.

 

[Buffet]

I dont get this argument, The DRS is not controlled by Pakistanis, its the broadcaster. and its not one person operating the system, there 's a team with access to it.

In modern softwar applications, everything is logged and audited. You cant just make a change and no one will find out

+1

It's not like one person can do whatever they want no one will find out.

 

[mominsaigol]

The Usama mir wicket, well before they should the wicket, they accidentally ran the following video, and than they corrected it immiedietly on and got it both umpires call.

later it changed to wickets UMPIRE'S Call

No, they don't show the simulations on live television.

I've been to scg and inside the commentary booth as well as the recording systems for where the drs takes place and the machine does its thing.

The run atleast 5 to 8 simulations very very quickly and choose the results via a mean average.

In this case the model is showing umpire call more times then it showed missing wickets so they went for umpire call and showed it on live television.

 

[Extra_Cover]

This is very suspicious. RVD extremely unlucky on it.

 

[mominsaigol]

Imagine the reaction if this was done in favor of an Indian bowler. Absolute absurd decision by the umpire and it seems like Hawkeye was tempered with as well.

Shameful. Can Pakistan win without cheating?

Bro it's not cheating. I've been to stadiums inside tours.

They run the simulation multiple times and choose which to show on TV via mean average. More simulations showed umpire call then missing.

The missing simulation was accidentally shown on tv, but many simulations aren't shown in TV, only the final result.

The technology isn't perfect so they need to run it various times before the 3rd umpire makes a decision

 

[happydavy]

The system might allow edits of parameters, but there would be logging of any changes done and it would be a huge scandal if it was changed during the review.

Sometimes the models run multiple times to get a high confidence result, and my guess is that the final result was clipping the bails

Don't make science up, rather research and understand how drs works. There is no modelling going on here, the only data point that matters and which is in the hands of analysts sitting there is the exact point of impact. From there the software calculates projected path and it gives only one result. Yes DRS can be manouvered as there is an explicit human process involved in marking the exact point of impact. And how that matters, well the ball is round in shape and one partial rotation between 2 frames could change the trajectory projection and it's upto the human sitting there to pinpoint the exact frame when ball hits the pad. I am not saying for one second that someone manouvered in this case, it could very well be that while they were still figuring out the exact impact point on their software, tv screen showed one result whereas they would have decided, in their wisdom, to change the impact point frame while taking the final decision.

 

[Major]

No, they don't show the simulations on live television.

I've been to scg and inside the commentary booth as well as the recording systems for where the drs takes place and the machine does its thing.

The run atleast 5 to 8 simulations very very quickly and choose the results via a mean average.

In this case the model is showing umpire call more times then it showed missing wickets so they went for umpire call and showed it on live television.

So could it be possible they show a model that is missing stumps?

Thanks for the bts info

 

[happydavy]

Bro it's not cheating. I've been to stadiums inside tours.

They run the simulation multiple times and choose which to show on TV via mean average. More simulations showed umpire call then missing.

The missing simulation was accidentally shown on tv, but many simulations aren't shown in TV, only the final result.

The technology isn't perfect so they need to run it various times before the 3rd umpire makes a decision

Wrong wrong and wrong. Simulation, which is projecting trajectory, happens based upon deciding the impact point. If u decide the impact point, post that only results show up, no averaging of simulations take place. They may very well decide to use the preceding or succeeding frame as impact point and that could different results but no averaging of simulations take place.

 

[Major]

Don't make science up, rather research and understand how drs works. There is no modelling going on here, the only data point that matters and which is in the hands of analysts sitting there is the exact point of impact. From there the software calculates projected path and it gives only one result. Yes DRS can be manouvered as there is an explicit human process involved in marking the exact point of impact. And how that matters, well the ball is round in shape and one partial rotation between 2 frames could change the trajectory projection and it's upto the human sitting there to pinpoint the exact frame when ball hits the pad. I am not saying for one second that someone manouvered in this case, it could very well be that while they were still figuring out the exact impact point on their software, tv screen showed one result whereas they would have decided, in their wisdom, to change the impact point frame while taking the final decision.

He isnt making stuff up. Well its not my field. But hawk eye works on prediction, and in prediction you run models multiple times. I think is it because of the bell curve of probability?

 

[Major]

Wrong wrong and wrong. Simulation, which is projecting trajectory, happens based upon deciding the impact point. If u decide the impact point, post that only results show up, no averaging of simulations take place. They may very well decide to use the preceding or succeeding frame as impact point and that could different results but no averaging of simulations take place.

Bro is this your field?

 

[mominsaigol]

Wrong wrong and wrong. Simulation, which is projecting trajectory, happens based upon deciding the impact point. If u decide the impact point, post that only results show up, no averaging of simulations take place. They may very well decide to use the preceding or succeeding frame as impact point and that could different results but no averaging of simulations take place.

Bro they check simulations. They don't run it only once. Because running it only once can lead to inaccurate results. The technology ain't perfect.

 

[HitWicket]

Don't make science up, rather research and understand how drs works. There is no modelling going on here, the only data point that matters and which is in the hands of analysts sitting there is the exact point of impact. From there the software calculates projected path and it gives only one result. Yes DRS can be manouvered as there is an explicit human process involved in marking the exact point of impact. And how that matters, well the ball is round in shape and one partial rotation between 2 frames could change the trajectory projection and it's upto the human sitting there to pinpoint the exact frame when ball hits the pad. I am not saying for one second that someone manouvered in this case, it could very well be that while they were still figuring out the exact impact point on their software, tv screen showed one result whereas they would have decided, in their wisdom, to change the impact point frame while taking the final decision.

The point of impact hasn't had any manual intervention for years. The full trajectory has also always been predicted before the point of impact is identified, changing the point of impact can't therefore impact the trajectory.

 

[happydavy]

The point of impact hasn't had any manual intervention for years. The full trajectory has also always been predicted before the point of impact is identified, changing the point of impact can't therefore impact the trajectory.

Pls see this

Point of impact is marked by the analyst sitting there. He can only see frames and there could be times when they may have to select between 1 of the 2 frames.

 

[oatmeal]

Don't make science up, rather research and understand how drs works. There is no modelling going on here, the only data point that matters and which is in the hands of analysts sitting there is the exact point of impact. From there the software calculates projected path and it gives only one result. Yes DRS can be manouvered as there is an explicit human process involved in marking the exact point of impact. And how that matters, well the ball is round in shape and one partial rotation between 2 frames could change the trajectory projection and it's upto the human sitting there to pinpoint the exact frame when ball hits the pad. I am not saying for one second that someone manouvered in this case, it could very well be that while they were still figuring out the exact impact point on their software, tv screen showed one result whereas they would have decided, in their wisdom, to change the impact point frame while taking the final decision.

The one projected path is a result of the mean of multiple model runs. I admit I have no insight into hawkeye specifically but generally to predict anything you have multiple simulations to achieve high confidence

I have worked with CFD and the system runs simulations thousands of times before giving the final result.

Either way, if a manual intervention was made *after* the first result, it would have been recorded

 

[HitWicket]

He isnt making stuff up. Well its not my field. But hawk eye works on prediction, and in prediction you run models multiple times. I think is it because of the bell curve of probability?

The only benefit to running a prediction multiple times is if something in the input is changing. In this case it's a single delivery with the cameras pinpointing the ball at very regular intervals. There's no benefit to running it multiple times because it should always return the same result.

 

[happydavy]

The point of impact hasn't had any manual intervention for years. The full trajectory has also always been predicted before the point of impact is identified, changing the point of impact can't therefore impact the trajectory.

Pls check the video in my post. However, if you are sure that this is not how drs works anymore, then I will take back my words. Any video link will be helpful.

 

[happydavy]

The one projected path is a result of the mean of multiple model runs. I admit I have no insight into hawkeye specifically but generally to predict anything you have multiple simulations to achieve high confidence

I have worked with CFD and the system runs simulations thousands of times before giving the final result.

Either way, if a manual intervention was made *after* the first result, it would have been recorded

I am not saying in the backend modeling may not be happening, I am saying the software doesn't give you intermediate results flickering between yes and no 1000 times based upon calculations running in the background. Software gives you one result.

 

[HitWicket]

Pls check the video in my post. However, if you are sure that this is not how drs works anymore, then I will take back my words. Any video link will be helpful.

The video shows Virtual Eye who only provide ball tracking services in Australia for CA events (and potentially NZ although I'm unsure with their recent broadcaster changes), the rest of the world and any ICC events use Hawkeye instead.

 

[AAli98]

Indians doing it deliberately they want Pak in the semis for a 1 sided victory

 

[DeadlyVenom]

It wasn't in this case.

It was edited in 2011 by the Indian hosts + broadcasters though.

 

[Anupam]

If it was Indian team that got the decision in its favor, everyone would have been moaning about BCCI fixing the decision. The hypocrisy is amazing.

 

[oatmeal]

I am not saying in the backend modeling may not be happening, I am saying the software doesn't give you intermediate results flickering between yes and no 1000 times based upon calculations running in the background. Software gives you one result.

lol dude of course the software gives you one result. That's what it is designed to do.

But this one result is the result of a complex computer model. you need to combine the ball trajectory of different cameras and do complex calculations to achieve that one result. this isnt a definitive calc, it involves modelling internally. They know how a ball moves and then they use it predict the trajectory of this particular delivery using visual inputs from all the Hawkeye cameras. You absolutely need to run it multiple times.

I dont know if those intermediate runs produce the final trajectory visual that was shown on the screen at first. But It's obvious that in any system being used today, you have multiple people involved . so one person cannot just tweak it in one team's favour and not leave an audit trail. That's terrible for the sport and the company which owns hawkeye

 

[happydavy]

Pls see this

Point of impact is marked by the analyst sitting there. He can only see frames and there could be times when they may have to select between 1 of the 2 frames.

The interviewer asked the analyst 3 or 4 times that it's the human sitting there who decides the exact point of impact and hes saying they do it based upon the 2 frames - one which shows hitting and the one before which doesn't.

 

[topspin]

Bro it's not cheating. I've been to stadiums inside tours.

They run the simulation multiple times and choose which to show on TV via mean average. More simulations showed umpire call then missing.

The missing simulation was accidentally shown on tv, but many simulations aren't shown in TV, only the final result.

The technology isn't perfect so they need to run it various times before the 3rd umpire makes a decision

Brilliant insight.

 

[mominsaigol]

Brilliant insight.

The thing that's usually replayed is the point of impact for wicket. The predictor is always correct where the ball lands usually and can tell if it's pitched outside leg or not.

But after that it's all a guessing game for the machine, as it assumed where the ball would go had the batsmen not obstructed the path. So they run it a couple of times.

 

[topspin]

Wrong wrong and wrong. Simulation, which is projecting trajectory, happens based upon deciding the impact point. If u decide the impact point, post that only results show up, no averaging of simulations take place. They may very well decide to use the preceding or succeeding frame as impact point and that could different results but no averaging of simulations take place.

Tu rehnde

 

[HitWicket]

lol dude of course the software gives you one result. That's what it is designed to do.

But this one result is the result of a complex computer model. you need to combine the ball trajectory of different cameras and do complex calculations to achieve that one result. this isnt a definitive calc, it involves modelling internally. They know how a ball moves and then they use it predict the trajectory of this particular delivery using visual inputs from all the Hawkeye cameras. You absolutely need to run it multiple times.

I dont know if those intermediate runs produce the final trajectory visual that was shown on the screen at first. But It's obvious that in any system being used today, you have multiple people involved . so one person cannot just tweak it in one team's favour and not leave an audit trail. That's terrible for the sport and the company which owns hawkeye

I think you're significantly overestimating the amount of factors considered for how a ball moves after impact. In terms of line it's just extending it straight from the point of impact, in terms of height it's just applying gravity.

 

[Marooned]

Seems fishy as hell.

BCCI tampering Hawkeye. Again

 

[owenhcricket]

The system might allow edits of parameters, but there would be logging of any changes done and it would be a huge scandal if it was changed during the review.

Sometimes the models run multiple times to get a high confidence result, and my guess is that the final result was clipping the bails

Yep, happens with computer models all the time.

Probably ran on tv by accident but everyone has a right to question this and it hohuld be explained for clarity by the ICC.

 

[topspin]

The thing that's usually replayed is the point of impact for wicket. The predictor is always correct where the ball lands usually and can tell if it's pitched outside leg or not.

But after that it's all a guessing game for the machine, as it assumed where the ball would go had the batsmen not obstructed the path. So they run it a couple of times.

Also, that explains why it takes time for ball tracking to load up because they need to generate multiple simulations. The expected outcome i.e. average - is the one seen by the 3rd umpire and the viewers.

 

[SpiritOf1903]

VDD getting out hampered Pakistan...

 

[oatmeal]

I think you're significantly overestimating the amount of factors considered for how a ball moves. In terms of line it's just extending it straight from the point of impact, in terms of height it's just applying gravity.

Its no fluid dynamics but that doesnt mean its very straightforward either. You still need to calculate the trajectory, distance left to traverse, speed to calculate the how high it would go. These are complex calculations.

 

[HitWicket]

Its no fluid dynamics but that doesnt mean its very straightforward either. You still need to calculate the trajectory, distance left to traverse, speed to calculate the how high it would go. These are complex calculations.

Correct, but they're all fixed inputs for each delivery that aren't going to change each time you run the prediction.

 

[MenInG]

So Miller decided to hang around after being out and wait for DRS - think THAT is what I call cheating.

 

[mominsaigol]

Also, that explains why it takes time for ball tracking to load up because they need to generate multiple simulations. The expected outcome i.e. average - is the one seen by the 3rd umpire and the viewers.

It doesn't even make sense for edits.

Like yeah bro, someone took a screenshot, edited it in photoshop and quickly plastered it back together for the viewers all in a span of a few minuted without anyone on the broadcast team noticing because the Indians umpire is a pakistani fan boy

 

[oatmeal]

Correct, but they're all fixed inputs for each delivery that aren't going to change each time you run the prediction.

I think its a bit more complex that being fixed inputs as pretty much all of it is a combination of multiple camera angles. But lets agree to disagree. I do think the broadcaster should provide an explanation of what happened

 

[MenInG]

It doesn't even make sense for edits.

Like yeah bro, someone took a screenshot, edited it in photoshop and quickly plastered it back together for the viewers all in a span of a few minuted without anyone on the broadcast team noticing because the Indians umpire is a pakistani fan boy

For people who do not like Pakistan, yes very possibe.

 

[sweep_shot]

That was very odd indeed.

Why was first result different? Looking forward to explanation.

 

[Saanji]

Don't forget Mohali 2011 Hawkeye is dubious

So BCCI working to favor pakistan now in 2023?

 

[Saanji]

The Usama mir wicket, well before they should the wicket, they accidentally ran the following video, and than they corrected it immediately on and got it both umpires call.

later it changed to wickets UMPIRE'S Call

Ball is pitched in line.. impact is umpire call
Logically the ball should continue to move right.
How at hitting wickets remain umpires call?

Probabilistic models are just yummy for simple minded conspiracy theories

 

[Major]

some very good insights by posters, i actually learned soemething new about cricket.

I wonder if Virtual Eye is more accurate than Hawk Eye

 

[oatmeal]

"In today's match between South Africa and Pakistan, an incomplete graphic was erroneously displayed during the LBW review of Rassie van der Dussen. The completed graphic with the right details was ultimately displayed."

ICC has issued a statement

 

[Major]

From ICC: "In today's match between South Africa and Pakistan, an incomplete graphic was erroneously displayed during the LBW review of Rassie van der Dussen. The completed graphic with the right details was ultimately displayed."

 

[Kaptan]

Even if there are hundreds of simulation running before the actual thing, everything would just be stored as numbers in backend why would anyone make visualisation for simulations.

Only after all the simulation are done and result is know (in form of numbers), the visualisation should me made.

 

[ArabKnight]

If you have ANY ANY ANY sort of intelligence OP, you should know a computer only does what it’s designed to do, it doesn’t have a mind of it’s own.

Both DRS projections are exactly what the system is designed to produce by the people have engineered the DRS system.

 

[Cpt. Rishwat]

so if you claim that Hawk eye was edited here so maybe it was edited during Tendu's mohali innings?

I said this at the time, pretty sure it would be easy enough to change the trajectory a touch. Who's going to ask for the digital footprint once the game is over?

 

[cricketjoshila]

Don't forget Mohali 2011 Hawkeye is dubious

As per who?

 

[cricketjoshila]

No, they don't show the simulations on live television.

I've been to scg and inside the commentary booth as well as the recording systems for where the drs takes place and the machine does its thing.

The run atleast 5 to 8 simulations very very quickly and choose the results via a mean average.

In this case the model is showing umpire call more times then it showed missing wickets so they went for umpire call and showed it on live television.

This is correct. Hawkeye has no manual inputs.

 

[oatmeal]

Even if there are hundreds of simulation running before the actual thing, everything would just be stored as numbers in backend why would anyone make visualisation for simulations.

Only after all the simulation are done and result is know (in form of numbers), the visualisation should me made.

The software likely has a option to generate the visualisation from the projection after combining it with the replay. And someone screwed up there.

hope they have safeguards around it. If this happens in a more crucial knockout game people will lose their mind

 

[Saanji]

Pls see this

Point of impact is marked by the analyst sitting there. He can only see frames and there could be times when they may have to select between 1 of the 2 frames.

That's a very critical point of error .

 

[160KPH]

Even during the first time where it showed it missing the wickets, it actually still played the sound of it hitting the wickets.

 

[Kianig89]

As per who?

Any proof of legitimacy

 

[BouncerGuy]

As per news reports:

Despite agreeing that there was a glitch in the technology, ICC clarified that the right decision was made eventually.

"In today's match between South Africa and Pakistan, an incomplete graphic was erroneously displayed during the LBW review of Rassie van der Dussen. The completed graphic with the right details was ultimately displayed,"