Hawkeyes Accuracy.

[thevortex]

And to those really interested here is an article on Hawkeye at Wikipedia. Please refer specifically to the doubts section. It will make plenty clear as to how perfect the technology really is. By the way the claimed error margin for Hawkeye is 3.6 mm. And there are already critics who say that the margin is way over that - proved in specific incidents. http://en.wikipedia.org/wiki/Hawk-Eye

[thevortex]

"On no occasion have we said that this technology is perfect," said Stuart Miller, head of science and technical issues at the ITF. Collins and Evans said that while what Hawk-Eye achieves is remarkable, its use in tennis needs to be refined. "It should be used like a spell-checker on your computer," Collins said. "It's not right all the time, but it's a useful adviser." http://sports.espn.go.com/sports/tennis/wimbledon08/news/story?id=3452293

[Gunner_Mania]

http://www.hawkeyeinnovations.co.uk/UserFiles/File/Hawk-Eye%20accuracy%20and%20believability2.pdf This is a link to one of the documents that BossBhai posted above. At the ned you can see that Mickey Arthur has become a convert:) Thevortex, I think you should write to Paul Hawkins, may be you will get more responses and you might be converted:)

[thevortex]

Here you go: A former employee of Hawkeye Inc. speaks his mind out on exactly why Hawkeye cannot be as accurate as Bossbhai thinks. This is with specific reference to tennis - as far as the ball pitching goes. But cricket offers a different set of problems. ----------------------------------------------------- Hawkeye can't decide line calls I cannot believe I am actually doing this. In the fourth set of the final, a Nadal 'winner' was decided to be that. Because, Rafa challenged the call otherwise. A review by the Hawkeye System predictably left Federer agitated. So was I. For twin reasons. As a Fed fan first and secondly, as someone who worked for Hawkeye a while, I cannot believe it is being used to judge such line calls. Two things, 1. The spatial analysis does not result in an absolutely error free result for the x1, y1 of the ball at time time t1. This is just for camera 1. There are at least 6 cameras and the error simply gets added with each camera. 2. Secondly, the system's output gives a graphic. This simulates the ball at a pre-determined size. Which, at the time I was working, looked to be the most logical thing to do. Especially since I was working on the cricket software and the game was played with a hard ball. Now, when one determines the size of the ball and uses it to determine line calls in a Tennis match, it's plain ridiculous. Well, I did not realize it was when I was at it -- since the approximation appeared natural. But watching the game reminds you -- the tennis ball does not keep its full size on impact. It is a soft ball. To extrapolate it to its full size, on top of error 1 and then decide line calls, in Wimbledon finals, is stupid. I have no other word for it. 3. All this would have still worked fine had the laws of the game not been so idiotic to say -- even a nanometer of the line being caught by ball is good enough. That's a margin the system can't handle. In my experience, 15% of the ball size would be possible. But that is for a hard ball. For a soft ball, leave it to the line umpire. 4. Sorry Paul. http://themaanga.blogspot.com/2007/07/hawkeye-cant-decide-line-calls.html

[thevortex]

http://www.hawkeyeinnovations.co.uk/UserFiles/File/Hawk-Eye%20accuracy%20and%20believability2.pdf This is a link to one of the documents that BossBhai posted above. At the ned you can see that Mickey Arthur has become a convert:) Thevortex, I think you should write to Paul Hawkins, may be you will get more responses and you might be converted:)

See, I dont blame Hawkins. His is a commercial enterprise. And as such he is trying hard to get Hawkeye to be the problem-solver. Its upto the governing bodies to take the call. And I squarely place the blame on them for not evaluating sufficiently and for implementing haphazardly this whole Hawkeye technology in the game. Both tennis and cricket have been affected. I also think certain individuals and teams have become good at certain aspects of gamesmanship. Think Nadal challenging calls at opportune (or inopportune if you are Federer) times. Or think of Sri Lanka making wonderfully accurate referral decisions where India were found woefully wanting. It is not about accuracy. It is about playing the game that the technology plays. Adjust your appeals such that you know when Hawkeye shows you it is out. Probably not when it is actually out :).

[f2000560]

And to those really interested here is an article on Hawkeye at Wikipedia. Please refer specifically to the doubts section. It will make plenty clear as to how perfect the technology really is. By the way the claimed error margin for Hawkeye is 3.6 mm. And there are already critics who say that the margin is way over that - proved in specific incidents. http://en.wikipedia.org/wiki/Hawk-Eye

worst case scenario is hardly the way to look at reliability of the system. and in that case the worst umpiring decison is worse than the worst hawkeye decisons. on an average the hawkeye is prone to lesser error than the umpire. the error rate and the variance (with the real path which is hypothetical) will only grow lesser with refinement and improvement of technology. the biggest plus is its unbiased and fair.

[thevortex]

worst case scenario is hardly the way to look at reliability of the system. and in that case the worst umpiring decison is worse than the worst hawkeye decisons. on an average the hawkeye is prone to lesser error than the umpire. the error rate and the variance (with the real path which is hypothetical) will only grow lesser with refinement and improvement of technology. the biggest plus is its unbiased and fair.

See - my question is, why the burning hurry to implement this? I would love for the principles of natural justice to prevail here. I would much rather Hawkeye not be used if it were to give a not out as out. That is important to any sportsman. Any fair-minded sportsman or fan. Especially given that the umpires in all probability will not have given the batsman out. I dont like the margins imposed by Hawkeye. In ball games, the ball has an uncanny knack of hitting lines and it is not as if these marginal decisions are going to be too few as to be disregarded. Tennis is slightly better off than cricket given that the predictive element is thrown out. However, in some respects tennis is suffering even more given that the rule that even a smidgen of the ball touching the line is an 'in'. The error margins for Hawkeye dont help in such cases. Dont you think?

[Gunner_Mania]

See - my question is, why the burning hurry to implement this? I would love for the principles of natural justice to prevail here. I would much rather Hawkeye not be used if it were to give a not out as out. That is important to any sportsman. Any fair-minded sportsman or fan. Especially given that the umpires in all probability will not have given the batsman out. I dont like the margins imposed by Hawkeye. In ball games, the ball has an uncanny knack of hitting lines and it is not as if these marginal decisions are going to be too few as to be disregarded. Tennis is slightly better off than cricket given that the predictive element is thrown out. However, in some respects tennis is suffering even more given that the rule that even a smidgen of the ball touching the line is an 'in'. The error margins for Hawkeye dont help in such cases. Dont you think?

How is that a given? If it were a given, we probably wouldn't have even turned to technology for help. It is not like Umpires are perfect and there is no margin for error. You accept any margins of error from a human but say I cant accept one from technology.

[CSK Fan]

http://sports.espn.go.com/sports/tennis/wimbledon08/news/story?id=3452293

Which confirms Bossbhai's point. For those who use them, spellcheckers on computers have reduced by a large margin the spelling errors made by human. the margin of error a spellchecker has and the impact of an error made by it is negligible compared to human errors.It has definitely improved documents and e-mails

[thevortex]

How is that a given? If it were a given' date=' we probably wouldn't have even turned to technology for help. It is not like Umpires are perfect and there is no margin for error. You accept any margins of error from a human but say I cant accept one from technology.[/quote'] That is why I said 'in all probability'. So its not a given :). I have never said umpires are perfect. We have gone over this before - in the other UDRS thread :).

Which confirms Bossbhai's point. For those who use them' date=' spellcheckers on computers have reduced by a large margin the spelling errors made by human. the margin of error a spellchecker has and the impact of an error made by it is negligible compared to human errors.It has definitely improved documents and e-mails[/quote'] Actually no. Hawkeye is a great parlour-tool. It can be entertaining to the TV audience. It can even make commentators talk some cricket once in a while rather than prattle on about MRF and DLF maximums and the like. But to base actual decisions on a predictive algorithm - dont you think you want more accuracy and more independent verification before you commit to that? Dont you think these administrators owe to the game even that much?

[Roshanrocks]

the hawkeye must have already been tested and proven to be scientifically correct by third party inspection agencies before being brought into play. SO I would rather believe the hawkeye than the umpire

[thevortex]

the hawkeye must have already been tested and proven to be scientifically correct by third party inspection agencies before being brought into play. SO I would rather believe the hawkeye than the umpire

Ask Bossbhai - the strongest supporter of Hawkeye - if that is so. Whether there is some independent third party assessment of Hawkeye? Fact is - there is nothing out there.

[Guest BossBhai]

-- Removed on request of the user --

[f2000560]

See - my question is, why the burning hurry to implement this? I would love for the principles of natural justice to prevail here. I would much rather Hawkeye not be used if it were to give a not out as out. That is important to any sportsman. Any fair-minded sportsman or fan. Especially given that the umpires in all probability will not have given the batsman out. I dont like the margins imposed by Hawkeye. In ball games, the ball has an uncanny knack of hitting lines and it is not as if these marginal decisions are going to be too few as to be disregarded. Tennis is slightly better off than cricket given that the predictive element is thrown out. However, in some respects tennis is suffering even more given that the rule that even a smidgen of the ball touching the line is an 'in'. The error margins for Hawkeye dont help in such cases. Dont you think?

vortex i think the only probable sample where the predictive element can come under serious doubt is if it hits him on the full. then the question arises would it have spun/ seamed after it bounces, will there be variable bounce beyond the most LBW's the ball is not going to bounce do something and then suddenly do something else this is a cricket ball here not a golden snitch. As long as HE is able to predict the coordinates, velocity and the the angle of incidence how wrong can the predictive path be ( i would add rotation on the ball and wind velocity) and then you have most things needed to plot a path. :dontknow:

[CSK Fan]

. But to base actual decisions on a predictive algorithm - dont you think you want more accuracy and more independent verification before you commit to that? Dont you think these administrators owe to the game even that much?

Heck rocket ships and space-craft containing people are designed and guided on the basis of predictive algorightm programs and run on computers. This is only a game

[CSK Fan]

vortex i think the only probable sample where the predictive element can come under serious doubt is if it hits him on the full. then the question arises would it have spun/ seamed after it bounces' date=' will there be variable bounce beyond the most LBW's the ball is not going to bounce do something and then suddenly do something else :[/quote'] Actually in all LBW descisons taken by human umpires, if a ball hits a batsman on the full, it is considered the ball will go straight on

[f2000560]

Actually in all LBW descisons taken by human umpires' date=' if a ball hits a batsman on the full, it is considered the ball will go straight on[/quote'] true ! i was just mentioning that to say that HE predictive part will be doubtful only then when you don't need it to be pin point. other wise it should be spot on or nearly nearly spot on.

[thevortex]

vortex i think the only probable sample where the predictive element can come under serious doubt is if it hits him on the full. then the question arises would it have spun/ seamed after it bounces' date=' will there be variable bounce beyond the most LBW's the ball is not going to bounce do something and then suddenly do something else this is a cricket ball here not a golden snitch. As long as HE is able to predict the coordinates, velocity and the the angle of incidence how wrong can the predictive path be ( i would add rotation on the ball and wind velocity) and then you have most things needed to plot a path. :dontknow:[/quote'] There are other probable samples, f200560 (you have got to give me a different handle to write your name:)). We just need to look at footage. When the ball is spinning, Hawkeye is so very lost. The same with prodigious swing. There are other problems as well - as in the tendency to depict almost every ball as grazing the leg stump even when it is speared into the batsman and would go down the leg side harmlessly.

Heck rocket ships and space-craft containing people are designed and guided on the basis of predictive algorightm programs and run on computers. This is only a game

We are not designing or guiding spacecraft are we? Pardon the pun - but cricket is a simple game. I ask a simple question. Has Hawkeye's accuracy been measured? If so, give me the numbers. If not, just say so that you love technology for its own sake and we can move on. I am in the IT field, My two cents. And to believe that computers can trump human intuition and/or intelligence is just plain hogwash.

Actually in all LBW descisons taken by human umpires' date=' if a ball hits a batsman on the full, it is considered the ball will go straight on[/quote'] Says who?!

[yoda]

Has Hawkeye's accuracy been measured? If so' date=' give me the numbers.[/quote'] Can we see some numbers for Bucknor? :hehe:

[Manny_Pacquiao]

The issue isn't the degree of hawkeye's accuracy. the issue is how accurate hawkeye would be relative to the dumb **** cocksucking umpires we have today. i feel confident in saying that it is far more accurate than your average international umpire. at the very least, using hawkeye on referred appeals would eliminate the howlers made so often by the incompetent chumpires. that alone, makes hawkeye worth having.