DRS, Hot Spot and Billy Bowden

[Old guy]

Are you sure that hawkeye account for all the variations of 1. Pitch conditions ( Different countries, different grounds and different days of the match), 2. atmospheric conditions (cloudy, sunny etc), 3. height of the bowler (the height from which the ball is delivered) 4. Condition of the ball (new, 40 overs old, 70 overs old etc) 5. Speed of delivery. A slight variation in any of the above factors can make change the trajectory and direction of the ball after pitching. Regarding your example of GPS calculations, it is not apt as there are hardly any variables. Once u know the coordinates the position will not change.

on top of that in england we have seen ball swing so much all of sudden in air thats why wicket keepers hate to keep in england how can hawk eye predict that? i never understood :P

[zep1706]

:facepalm: do u understand this: s = u*t + 0.5*g*t^2 ? and Hawkeye calculations are much more than that... calculating the trajectory/direction/speed of ball before impact is much more difficult than what happens after impact...and is reasonably well understood.. btw, do you know we have been able to launch satellites into orbits, based on some 'rudimentary' maths and physics equations.. and those satellites with few more equations/maths, are good enough to help us calculate GPS positions, within few meters... and yet, people think Hawkeye is not even as good as slot machine from Mahabharat yuga.

Just because we have been able to launch satellites and missiles doesn't make Hawk Eye perfect on its own.

[MCcricket]

At this point of time DRS is just a weapon used by loosers to gain advantage over asian sides,it is a very costly, inaccurate technology, I dont like the way it is used, should be used by umpires to help decide not by 3 umpire to overule decisions.

[akshayxyz]

Are you sure that hawkeye account for all the variations of 1. Pitch conditions ( Different countries, different grounds and different days of the match), 2. atmospheric conditions (cloudy, sunny etc), 3. height of the bowler (the height from which the ball is delivered) 4. Condition of the ball (new, 40 overs old, 70 overs old etc) 5. Speed of delivery. A slight variation in any of the above factors can make change the trajectory and direction of the ball after pitching. Regarding your example of GPS calculations, it is not apt as there are hardly any variables. Once u know the coordinates the position will not change.

The equation was in response to your comment - 'Hawkeye just calculates straight line'...anyone with basic maths/physics education would know that trajectory of projectile motion under gravity is anything but straight line ( except when horizontal velocity is zero). On rest of your points..unless you are trolling - NO - Hawkeye does not need to take care of these. It is not a bowling machine or virtual batsman or something...it just needs to track the ball, calculate the point of impact and velocity just before the impact and then extrapolate for rest of the 1-2 meters of distance. as for GPS example.. well its not like once satellite sits in the geostationary position, it is perfectly still. Besides, my point was about the process of putting it up, and that is much more complex than 40-50 variables you can come up with. If you can come up with glaring problems or technological challenges in 'tracking the ball', then we can talk.

on top of that in england we have seen ball swing so much all of sudden in air thats why wicket keepers hate to keep in england how can hawk eye predict that? i never understood :P

As explained above, all Hawkeye needs to do is estimate the trajectory for the couple of meters, from impact to stumps, based on real data collected from tracking the ball - which would already take care of all the pitch/atmosphere conditions. And I think Hawkeye team mentions somewhere that, if the impact is within 2.5 meters of the stumps, the error is not that much. I guess, cant be more than an inch, so except for the cases where it shows the ball is barely clipping/missing the stumps.. it is fairly accurate.

Just because we have been able to launch satellites and missiles doesn't make Hawk Eye perfect on its own.

That was an example to show that much more complex problems are being solved using similar maths/physics. Problem Hawkeye has to solve is pretty simple - the most difficult part is tracking the ball, and this also has been solved and being used in various fields within acceptable limits. Also - on the argument - Hawkeye is not 100% correct. Well - it is not required to be. It just needs to be reasonably better than naked eyes - which it certainly is... Or someone can go ahead and enlighten us how/why naked eye is better equipped to predict the path (post impact).

[akshayxyz]

To give you an idea of - how advanced technology has become in solving ball/object-tracking problem... here are some demos. [ame=http://www.youtube.com/watch?v=t_qN3dgYGqE]Robot plays table tennis (vs Robot, vs Human) - YouTube[/ame] [ame=http://www.youtube.com/watch?v=flyEY7P48D0]Robot catching and Juggling a Ping Pong Ball - Demo - YouTube[/ame] [ame=http://www.youtube.com/watch?v=3CR5y8qZf0Y]Quadrocopter Ball Juggling, ETH Zurich - YouTube[/ame] And before you reject above demos, bringing in pitch conditions and stuff - let me re-iterate, this is about calculating the motion of an object from the video source - that too in real time!!.

[Needforspeed]

akshayxyz...No I am not trolling, these are genuine doubts. You seem to be quite Knowledgeable about such stuff..but I still have some doubts. Now take an example where say Ishant Sharma bowls 2 balls - one in Perth and another in Chennai. Let us say everything else is the same and ball hits the batsman's pad while he is on the front foot. The bounce of the ball in Perth will be different from Chennai. Even if the impact point on the pad is same, the extra bounce in Perth would have taken in over the stumps in Perth while it would have maybe taken the bails in Chennai.

[Needforspeed]

Now in the above experiments (let us take pingpong) Suppose we take introduce some variables. Suppose I put some chalk or dust at random points on the table (mimicking the different pitch behaviour) , and at times the ball skids thru will the ping pong match continue? Again it is not my intention to contradict anyone, just thinking aloud.

[Old guy]

akshayxyz...No I am not trolling, these are genuine doubts. You seem to be quite Knowledgeable about such stuff..but I still have some doubts. Now take an example where say Ishant Sharma bowls 2 balls - one in Perth and another in Chennai. Let us say everything else is the same and ball hits the batsman's pad while he is on the front foot. The bounce of the ball in Perth will be different from Chennai. Even if the impact point on the pad is same, the extra bounce in Perth would have taken in over the stumps in Perth while it would have maybe taken the bails in Chennai.

actually this one is quite easy. ball will bounce at diff angle on those 2 pitches.. so it can be easily estimated . to me swingy conditions are doubtful cos ball swings after passing bat..so many times all of sudden and itsn ot little swing its huge.. seen wk's going in wrong direction i mean u cant predict that ball going straight and all of sudden swinging in air i get it 2.5 metre rule but in countries like england it swings enough to hit or miss how many times have u seen batsmen leave ball.. and then ball first going straight then all of sudden swing in marginally ? to get so close to stumps u really cannot predict it bounce is easy

[akshayxyz]

akshayxyz...No I am not trolling, these are genuine doubts. You seem to be quite Knowledgeable about such stuff..but I still have some doubts. Now take an example where say Ishant Sharma bowls 2 balls - one in Perth and another in Chennai. Let us say everything else is the same and ball hits the batsman's pad while he is on the front foot. The bounce of the ball in Perth will be different from Chennai. Even if the impact point on the pad is same, the extra bounce in Perth would have taken in over the stumps in Perth while it would have maybe taken the bails in Chennai.

as Sabby already explained - bounce is very easy to account for - given the ball has traveled decent distance before hitting the pads, to give enough data points for calculating velocity ( direction + speed) in 3D. As for swing - it should not be much different in that case either - if the ball has traveled sufficiently after pitching and before hitting the pad - Hawkeye would have enough data points to figure out the curvature of path. It is reasonable to assume that - ball would traverse a smooth curve after pitching and that curve can be figured out given that the impact is somewhere around at half way ( between pitching point and stumps). Regardless, even with limited data points, estimation of Hawkeye is going to be more accurate than that from naked eye. Human eye ( +brain) is just not good enough for precision calculations, they are good at something else - pattern matching and stuff. So, if Hawkeye can't predict certain cases, human umpire would do even worse job - more often than not. e.g. lets say you throw a ball - Hawkeye and a trained human is allowed to view the trajectory until half way, and both are asked to pinpoint where the ball would land - who do you think is expected to be more accurate? On the other hand, human eye is still better (so far) at stuff like - identify the objects in a given picture/scene/video.

[Old guy]

as Sabby already explained - bounce is very easy to account for - given the ball has traveled decent distance before hitting the pads, to give enough data points for calculating velocity ( direction + speed) in 3D. As for swing - it should not be much different in that case either - if the ball has traveled sufficiently after pitching and before hitting the pad - Hawkeye would have enough data points to figure out the curvature of path. It is reasonable to assume that - ball would traverse a smooth curve after pitching and that can be figured out given that the impact is somewhere around at half way ( between pitching point and stumps). Regardless, even with limited data points, estimation of Hawkeye is going to be more accurate than that from naked eye. Human eye ( +brain) is just not good enough for precision calculations, they are good at something else - pattern matching and stuff. e.g. lets say you throw a ball - and Hawkeye and an expert human is allowed to view the trajectory until half way, and both are asked to pinpoint where the ball would land - who do you think is expected to be more accurate? On the other hand, human eye is still better (so far) at stuff like - identify the objects in a given picture/scene/video.

but it swings in air..some times it does .. some times it goes straight how can that be predicted ? i know wht ur saying but i have sen balls traveling reasonably in straight line and then coming in or out.. thats why wicket keepers struggle in england its diff from ball which is swinging form start it can be easily predicted same with deviation from pitch ball swingsi n air cos one side is rough and other is smooth its bcos how airmovies around ball i cant explain it properly sorry but i hope u understand wht im trying to say

[Needforspeed]

From the above posts it seems that the technology is fairly accurate. Now ICC will have to sit down and do something about uniform implementation by the umpires. If a decision has been referred to the DRS, then umpires should not have any further discretion in the decision. If technology says it is out, irrespective of the original decision, it should be given out. Otherwise DRS will continue to confuse everyone.

[akshayxyz]

Now in the above experiments (let us take pingpong) Suppose we take introduce some variables. Suppose I put some chalk or dust at random points on the table (mimicking the different pitch behaviour) ' date= and at times the ball skids thru will the ping pong match continue? Again it is not my intention to contradict anyone, just thinking aloud.

Specifically to this case - it depends how these robots are estimating the ball trajectory. If they are using pre-fed parameters of table (friction, bounce ...), then it would fail. And I highly doubt this is the case here .. as that is too primitive, and would fail even for far simpler cases. Most likely, here also they are using some ball tracking technology and the robot keeps adjusting the position and angle of bat as it tracks the ball.. so as the ball skids off due to random dust or something, robotic bat should be able to adjust - as long as there is sufficient distance traveled by the ball - after the bounce/skid.

[akshayxyz]

From the above posts it seems that the technology is fairly accurate. Now ICC will have to sit down and do something about uniform implementation by the umpires. If a decision has been referred to the DRS' date= then umpires should not have any further discretion in the decision. If technology says it is out, irrespective of the original decision, it should be given out. Otherwise DRS will continue to confuse everyone.

especially as compared to the human umpire. So, it should be looked upon as one step further to better umpiring, rather than a leap to perfect umpiring. Besides, no one stops them to use other techs, replays etc. to make a decision in DRS. DRS is a process, and I do not think it should be enforcing use of a particular technology.

[akshayxyz]

but it swings in air..some times it does .. some times it goes straight how can that be predicted ? i know wht ur saying but i have sen balls traveling reasonably in straight line and then coming in or out.. thats why wicket keepers struggle in england its diff from ball which is swinging form start it can be easily predicted same with deviation from pitch ball swingsi n air cos one side is rough and other is smooth its bcos how airmovies around ball i cant explain it properly sorry but i hope u understand wht im trying to say

Lets say the ball pitched at about 4-5 meter (that is about full length zone I guess) and typical of LBW cases. Lets say the ball hit the pad at about 2-3 meter after pitching, and would have further traveled another 1.5-2 meter until it passes stumps. If there is some swing factor, it should kick in those 2-3 meter before impact too, and ball tracking should be able to estimate the curvature of trajectory and extrapolate for rest of the 2 meters or so... (it is immaterial what happens after it passes stumps). It has to be extremely rare, when the ball does not move at all for 2-3 meters before impact and suddenly takes U-turn after impact.... and again - I do not see how "would have" guess of a human umpire can be better than Hawkeye - even in such rare cases. So, I am aware of limitations you are suggesting, but I do not see why and how current umpires can deal with it better.

[Old guy]

Lets say the ball pitched at about 4-5 meter (that is about full length zone I guess) and typical of LBW cases. Lets say the ball hit the pad at about 2-3 meter after pitching, and would have further traveled another 1.5-2 meter until it passes stumps. If there is some swing factor, it should kick in those 2-3 meter before impact too, and ball tracking should be able to estimate the curvature of trajectory and extrapolate for rest of the 2 meters or so... (it is immaterial what happens after it passes stumps). It has to be extremely rare, when the ball does not move at all for 2-3 meters before impact and suddenly takes U-turn after impact.... and again - I do not see how "would have" guess of a human umpire can be better than Hawkeye - even in such rare cases. So, I am aware of limitations you are suggesting, but I do not see why and how current umpires can deal with it better.

ya its still better then human eye im all for it tbh... but even if its not used its same for both teams..(unless ur saying umpires are bought or its fixed) but since most of world wants it there is no real reason to oppose

[zep1706]

That was an example to show that much more complex problems are being solved using similar maths/physics. Problem Hawkeye has to solve is pretty simple - the most difficult part is tracking the ball, and this also has been solved and being used in various fields within acceptable limits.

I am pretty aware that much more complex problems have been solved and I have never questioned the feasibility of a tracking system. But your logic of supporting the Hawk Eye system is based solely on assumptions. Either that or you know and understand the details of the workings of Hawk Eye. There has not been any independent evaluation of the system as far as I know, so it's still just your words vs mine.

Also - on the argument - Hawkeye is not 100% correct. Well - it is not required to be. It just needs to be reasonably better than naked eyes - which it certainly is... Or someone can go ahead and enlighten us how/why naked eye is better equipped to predict the path (post impact).

No, Hawk Eye doesn't need to be 100% correct, but do you know how correct is it? Please do tell me its margin of error. A 5% error at the point of impact would be more than enough to get a wrong prediction in the case of those close ones. My point is that---do an evaluation of the system, if it's good enough, implement it universally and let ICC officials control it, not the broadcasters.

[akshayxyz]

I am pretty aware that much more complex problems have been solved and I have never questioned the feasibility of a tracking system. But your logic of supporting the Hawk Eye system is based solely on assumptions. Either that or you know and understand the details of the workings of Hawk Eye. There has not been any independent evaluation of the system as far as I know, so it's still just your words vs mine. No, Hawk Eye doesn't need to be 100% correct, but do you know how correct is it? Please do tell me its margin of error. A 5% error at the point of impact would be more than enough to get a wrong prediction in the case of those close ones. My point is that---do an evaluation of the system, if it's good enough, implement it universally and let ICC officials control it, not the broadcasters.

Of course I had to assume stuff about internal workings of Hawkeye, as they would definitely not release the details in public domain. But, I do not think my 'assumptions' are too far from reality, for: a) http://en.wikipedia.org/wiki/Hawk-Eye#Method_of_operation ... decent hint about how it works. It clearly mentions they use some sort of ball tracking technology, whether its triangulation or some other form of object segmentation - I would imagine, they are using best possible available tech. otherwise the errors would be too glaring to not notice in every use. b) this is how I would go about implementing a Hawkeye kind of system, and is quite feasible as per my knowledge of latest in related/required technologies - well supported by other research projects I pointed out. As I said - in its current state it definitely has to be better than what a naked eye can predict. P.S. I accept that it is quite possible that Hawkeye is not making best use of currently available algorithms/technologies and may be over-advertising actual accuracy of their product. But that would be hard to imagine, otherwise we would have had competing products, if there was some such glaring gap between available tech/algo and what they are using. P.P.S - I am not against any evaluation, and I hope Hawkeye is also not, it will only help improve the technology and system.

[akshayxyz]

To summarize - currently available tech is good enough to outdo naked eye, and it is highly unlikely that Hawkeye is using some tech, which can be outdone by human umpires - in matters of ball tracking and predicting path after impact.

[Raghav_12]

In both of today's cases, if DRS was not used batsmen were still have to go... so did DRS done anything worse than on-field umpires? NO... But there were few decisions been reversed in match because clear evidence were found against them, so did DRS proved helpful there? YES Moral of the story: DRS is there ONLY to IMPROVE on-field umpiring.

Well. This could have been very good argument, if umpires keep giving decisions in conventional manner, independent of knowing where there is DRS or not and let DRS correct if there is anything wrong. But problem arises when umpire starts thinking how DRS will look at that appeal and starts incorporating that effect within his on-field decisions. For example, now umpires have started giving a lot more front foot LBW's then they would do earlier, because they know if they don't give that decision then DRS will give. Honestly speaking, I won't blame umpires for that line of thinking because no umpire would like to keep getting egg on his face, brought by continuos reversal of his decisions by DRS. I think if DRS is implemented in current format, very soon we'll see hawk-eye being part of coaching manuals for batsmen.

[cybervickey]

In both of today's cases, if DRS was not used batsmen were still have to go... so did DRS done anything worse than on-field umpires? NO... But there were few decisions been reversed in match because clear evidence were found against them, so did DRS proved helpful there? YES Moral of the story: DRS is there ONLY to IMPROVE on-field umpiring.

^THIS RIGHT HERE.I think that its a welcome edition to cricket. Its not perfect but nothing really is.