Hawkeye founder on 'umpires call' debate

[trundler]

The impact of this is negligible tbh.

Should run Hawkeye on it imo

 

[Athlai]

Cameras can be affected by the wind. The objects they are mounted to (e.g. buildings) can also vibrate or sway due to wind or people moving around on/in them.

I really can't see this playing a large factor outside of the most extreme situations with the 6 camera setup all providing the data to the 3d space.

 

[TheJediBrah]

This is from Hawkins himself:



So every frame, 6 cameras from 6 angles are capturing and placing the ball within a 3-dimensional space to predict how it will move in the air and eventually land. It's just physics. But us armchair ****wits watching from a slightly off centre view think we know better because we're morons.

Are you seriously telling me you've never seen it when this happens? Usually when 2 impacts are close together, when it the tracking is clearly wrong and everyone out there knows it and has a laugh about it?

I thought this was clear but I'm not talking about tracking being a tiny bit different to what someone sitting at home thinks it should be and whinges, which happens too. I'm talking about when the tracking fails completely to pick up movement after pitching (or somehow predicts turn that didn't even happen, ie Zampa one last year) and the end result puts the ball feet away from where it actually would have gone

Because otherwise I don't get what you're arguing

 

[Athlai]

Yeah you're ****ing talking about what the broadcaster puts on the screen and don't know what you're talking about.

 

[Athlai]

Movement after pitching it only tracks the ball after it bounces ffs

 

[Spark]

Are you seriously telling me you've never seen it when this happens? Usually when 2 impacts are close together, when it the tracking is clearly wrong and everyone out there knows it and has a laugh about it?

I thought this was clear but I'm not talking about tracking being a tiny bit different to what someone sitting at home thinks it should be and whinges, which happens too. I'm talking about when the tracking fails completely to pick up movement after pitching (or somehow predicts turn that didn't even happen, ie Zampa one last year) and the end result puts the ball feet away from where it actually would have gone

Because otherwise I don't get what you're arguing

He already acknowledged that this was a problem though. It's a sufficiently rare edge case that IMO it should simply be accounted for automatically by the system.

 

[Athlai]

Parabolas how do they work

 

[TheJediBrah]

He already acknowledged that this was a problem though.

Then what is he talking about and why is he responding to me

It's a sufficiently rare edge case that IMO it should simply be accounted for automatically by the system.

What do you mean by this? Accounted for how?

 

[Dick Rockett]

The impact of this is negligible tbh.

Sure, most of the time. But it is definitely a thing. Just as the software is not, and can never be, completely accurate. Nor the hardware itself, even if somehow mounted to an immovable object, in a vacuum.

 

[TheJediBrah]

Yeah you're ****ing talking about what the broadcaster puts on the screen and don't know what you're talking about.

I have no idea what you're talking about at all or how it relates to what I'm saying

Everything I've said is objectively true, and happens. At no point did I claim to have insight into how the system works or why it happens. Literally just said "this is a problem, should be accounted for" but you keep arguing with me for some reason

 

[Spark]

Then what is he talking about and why is he responding to me

What do you mean by this? Accounted for how?

Warning when the ball and the impact are close together, say within a metre. What you do after that should depend on the nature of the bowler IMO.

 

[Spark]

Sure, most of the time. But it is definitely a thing. Just as the software is not, and can never be, completely accurate. Nor the hardware itself, even if somehow mounted to an immovable object, in a vacuum.

Okay but like my answer to this is the same as my answer to "what about the complex non-linear interactions between the ball and the pitch when the ball pitches" argument brought up earlier in the thread: it simply is too small to matter. Part of good system design and good engineering/physics analysis is being able to distinguish between sources of error which are significant and sources of error that really aren't.

 

[Dick Rockett]

Okay but like my answer to this is the same as my answer to "what about the complex non-linear interactions between the ball and the pitch when the ball pitches" argument brought up earlier in the thread: it simply is too small to matter. Part of good system design and good engineering/physics analysis is being able to distinguish between sources of error which are significant and sources of error that really aren't.

Yet we've all seen occasions when it was obviously incorrect. Why not just accept that Hawkeye isn't perfect? There will never be a system that gets every decision right.

 

[Spark]

Yet we've all seen occasions when it was obviously incorrect. Why not just accept that Hawkeye isn't perfect? There will never be a system that gets every decision right.

Because arbitrarily deciding that Hawkeye getting specific sorts of decisions wrong in a specific way that can be accounted for with specific measures represents all Hawkeye decisions being inaccurate and flawed to the degree that no Hawkeye decision is to be really trusted is a form of anti-thinking, which I think needs to be eradicated wherever it appears.

Upwards of 95% of Hawkeye decisions are accurate to within degrees of error which would be physically impossible for humans to achieve by several orders of magnitude. I'm very confident that's the case. There are many systems in the world which are extremely complicated and rely on very complex, difficult-to-model physical processes with substantial error bras on them. Hawkeye is not one of them.

 

[Daemon]

There's nothing to argue because my skepticism is less to do with the ball tracking tech and more to do with the manual element of setting the impact point. Saying mpire's call is overriding the tech isn't entirely accurate because a key part of the process which affects the predictive path of the ball is in the hands of another potentially useless human.

I listened to Simon Hughes' entire hour long podcast episode there and this wasn't properly discussed. I've seen enough lbws where impact point doesn't seem right (mayank agarwal's lbw in south africa a few years ago being the main example) for me to not completely be on board with it. If there's some proof out there about how accurately they actually set impact point and how they build the track off it, I'll concede.

That’s why I mentioned manual intervention in my post that you cropped.

Even accounting for the human element surely you don’t disagree that it is, a vast majority of the time, more accurate that an umpire judging something in real time? This surely cannot be a debate?

The person deciding the impact point has (at the very least) lots of slow motion video angles to rely on. The umpires have their useless eyes.

 

[Dick Rockett]

Because arbitrarily deciding that Hawkeye getting specific sorts of decisions wrong in a specific way that can be accounted for with specific measures represents all Hawkeye decisions being inaccurate and flawed to the degree that no Hawkeye decision is to be really trusted is a form of anti-thinking, which I think needs to be eradicated wherever it appears.

Oh come on, man. We're talking about decisions made in a game. A game!

I respect you and all but you're being dramatic.

 

[Spark]

Oh come on, man. We're talking about decisions made in a game. A game!

I respect you and all but you're being dramatic.

But I'm also correct. This really isn't a difficult problem and so much time is wasted on these debates because people can't accept that "where will object in free flight go" has been a problem solved to extraordinary accuracy since like the 1960s.

 

[Dick Rockett]

Oh my god.

If you say so. Enjoy being right, my friend.

 

[Spark]

That’s why I mentioned manual intervention in my post that you cropped.

Even accounting for the human element surely you don’t disagree that it is, a vast majority of the time, more accurate that an umpire judging something in real time? This surely cannot be a debate?

The person deciding the impact point has (at the very least) lots of slow motion video angles to rely on. The umpires have their useless eyes.

I mean I get where the confusion comes from, because seeing an LBW upheld because the ball is shaving the top corner of leg stump just looks **** and doesn't look like it should be a legitimate dismissal. But the problem is that the law itself sucks! Not that the technology is inaccurate or whatever.

 

[Daemon]

This is like when @cnerd123 claimed it is beyond science and technology in the year 2020 to automate front foot no balls