Cricket/Science question about the Hot Spot Technology

[Amie]

So, I don't know if many of you watch cricket and are aware of the Hot Spot Technology but here's some background if you don't: This technology uses infrared cameras and senses the heat generated from friction (that can come about either as a result of the ball on the bat or the ball on a batsman's pads). It uses some subtraction mechanism which forms "a series of black-and-white negative frames is generated into a computer, precisely localising the ball's point of contact". [Source: Wikipedia]

This technology was deemed to be infallible and costs a lot ($6000 bucks a day for two cameras). However, its accuracy was questioned after a series of incidents wherein the umpires in the match and the bowlers heard a definite crack but it was impossible to see on the hot spot. It is impossible to see these cuts or nicks in normal replays when it's a really faint edge. A Hot Spot replay usually shows a white spot on the edge of the bat (or the pad) if the ball has hit it, thus allowing the viewer to distinguish clearly if the offense has been made. After India toured England in 2011 July 2011, a decision was ruled in favour of an Indian batsman - VVS Laxman after looking at the hot spot replay of a shot. The replay showed no edge on the bat even though the players were convinced that there was a definite edge and the umpire was hesitant to go against his not-out decision because he too, believed that he had heard a definite sound of the ball hitting the bat.

There have been several other instances of this happening but I won't list them all. I don't have much/any knowledge of science in these situations and a google search didn't give me anything. I need to know if this technology can have such huge error margins (because this has happened twice in one tournament and a few other times in the last year) and if it can, what the science behind it is. I'd really appreciate any opinion on this because it has been gnawing at my head for a while now.

[ElWanderer]

I don't think anyone considers the Hot Spot, Snickometer or Hawk-eye technology to be infallible, as such. Or rather, I hope not! The problem with Hot Spot seems to be that the thinnest of edges result in the smallest amounts of friction between bat and ball, meaning any temperature gain will be so small it could be lost in the noise.

In more detail, but guessing somewhat: I presume the system works by setting a threshold temperature above which things are coloured bright white to highlight nicks, whilst everything else is left alone - that threshold will have to be a certain level above the average temperature of the bat, or else there will be lots of speckles of white (and therefore apparent nicks) from natural temperature variations, both in the bat and in the sensor of the IR camera. Alternatively or in addition, there may be a threshold for a number of adjacent pixels that have to show an increase for a spot to be shown as white. However it works, there will be steps to rule out as many false positives as possible, but those steps allow some false negatives, where the faintest of nicks to go undetected by the system.

Leaving behind the guesswork, I think this is one of the situations Wikipedia refers to:

However, Broad's aggressive approach paid off when Rahane miscued a pull shot to Samit Patel at fine leg and in his next over, he took the prized wicket of Rahul Dravid to leave India 86-2 although the wicket was not without controversy.

On-field umpire Billy Doctrove adjudged that Dravid had not made contact as the ball went through to wicketkeeper Craig Kieswetter but, although the "Hotspot" thermal imaging camera concurred, TV umpire Marais Erasmus said there was a sound and that he could see a deflection - a view later backed up by the "Snickometer" which highlighted a clear noise as the ball passed the bat.

BBC Test Match Special summariser Mark Butcher said: "Hot Spot is not something you can completely rely on.

"It's a bad piece of kit for those fine nicks but Dravid's body language told me he was out because the millisecond the ball passed the bat, he jerked his head to see behind."

And after the game, Parthiv Patel admitted: "I heard a noise."

(From http://www.bbc.co.uk/sport/0/cricket/14774834 )


And the BBC live text from the Laxman non-dismissal:

UMPIRE REVIEW
Anderson... Laxman... Prior celebrates, and England think they have their man here, even if Umpire Rauf says no...

BBC Test Match Special's Phil Tufnell
"I think he nicked that. Matt Prior has shaken Jimmy Anderson's hand and said well bowled. I don't see any point in hotspot - it's absolutely worthless, unless you get a fat edge it doesn't show. You know as a cricketer when someone has got a fine scratchy outside edge."

NOT OUT
Let's have a look at the replays - bat nowhere near the pad, and there looks to be a clear - if tiny - deviation as ball passes bat. Nothing on Hot Spot, and third umpire Billy Bowden just will not give that. English fielders hang heads - they were convinced by that, and whispers re the use of Vaseline on bat edges to defeat heat-seeking technology will only get louder. Other petroleum-based tinctures also available.

1136 India 48-1
This from Bowden (I'm paraphrasing): "I heard a sound, couldn't be 100% certain it came from the bat because of nothing on Hot Spot." So Bowden heard a nick. All the ex-pros around me are saying that was out. Laxman's head even flinched round to follow the ball. Hmm. Maiden from Broad to Dravid; India trail by 173.

(From http://news.bbc.co.uk/sport1/hi/cricket/9551196.stm )


Basically, if there is a noise (that can't have come from anywhere else) on Snickometer and a visible deflection of the ball's flight, it shouldn't actually matter if Hot Spot doesn't show anything. Whether all the 3rd umpires look at it that way or not is another matter. It depends on whether they see "no hot spot" on Hot Spot as absence of evidence (i.e. inconclusive in of itself), or evidence of absence (i.e. means there was no nick)...

[Adam H]

I need to know if this technology can have such huge error margins (because this has happened twice in one tournament and a few other times in the last year) and if it can, what the science behind it is.

The real interesting science is the statistics. How many times was hot spot used last year? How often did hot spot get it wrong? And are you SURE hot spot got it wrong?

"Huge error margins" in this case would be less than about 90% accuracy, I think. I'll bet you that of all the calls that hot spot overturns (so the most difficult ones) hot spot gets it right more or less 9 times out of 10. It's just that you remember the ones where it is most debatable.

Also, I would probably put a little less faith in the spectators hearing a nick. Your ears subconsciously brace themselves for loud sounds if you know they're coming. So if it's not picked up on video/snickometer, it's pretty weak evidence IMO.

[Sean Quixote]

I thought this thread was going to be about wifi.

That is all.

[Amie]

Ahhh, I knew I could count on the fora!

The real interesting science is the statistics. How many times was hot spot used last year? How often did hot spot get it wrong? And are you SURE hot spot got it wrong?

"Huge error margins" in this case would be less than about 90% accuracy, I think. I'll bet you that of all the calls that hot spot overturns (so the most difficult ones) hot spot gets it right more or less 9 times out of 10. It's just that you remember the ones where it is most debatable.

Also, I would probably put a little less faith in the spectators hearing a nick. Your ears subconsciously brace themselves for loud sounds if you know they're coming. So if it's not picked up on video/snickometer, it's pretty weak evidence IMO.

Okay. I can see what you mean. However, like I said, is it likely to occur twice over, in the same tournament?

I don't think anyone considers the Hot Spot, Snickometer or Hawk-eye technology to be infallible, as such. Or rather, I hope not! The problem with Hot Spot seems to be that the thinnest of edges result in the smallest amounts of friction between bat and ball, meaning any temperature gain will be so small it could be lost in the noise.

In more detail, but guessing somewhat: I presume the system works by setting a threshold temperature above which things are coloured bright white to highlight nicks, whilst everything else is left alone - that threshold will have to be a certain level above the average temperature of the bat, or else there will be lots of speckles of white (and therefore apparent nicks) from natural temperature variations, both in the bat and in the sensor of the IR camera. Alternatively or in addition, there may be a threshold for a number of adjacent pixels that have to show an increase for a spot to be shown as white. However it works, there will be steps to rule out as many false positives as possible, but those steps allow some false negatives, where the faintest of nicks to go undetected by the system...

..Basically, if there is a noise (that can't have come from anywhere else) on Snickometer and a visible deflection of the ball's flight, it shouldn't actually matter if Hot Spot doesn't show anything. Whether all the 3rd umpires look at it that way or not is another matter. It depends on whether they see "no hot spot" on Hot Spot as absence of evidence (i.e. inconclusive in of itself), or evidence of absence (i.e. means there was no nick)...

How can we confirm if the noise can't have come from anything else? It can be a faint crack of the wind amplified by our imagination as Adam H suggests that our brains can do that, right? No? I mean, I remember one of the instances where the ball touched a batsman's shoelace's aglet and there was a sound! The decision was against the batsman. Judging whether there's a nick or not through normal replays is really hard because it's sometimes hard to see with the naked eye.

Again, really clueless on the science bits, so just asking if there are possibilities of determining with certainty where the sound came from. Thanks a bunch, guys.

[Adam H]

Okay. I can see what you mean. However, like I said, is it likely to occur twice over, in the same tournament?

Well, if it's 90% accurate (completely made up number), and it's used 5 times a tournament (again, completely made up), then at least two errors will occur in about 1 out of every 12 tournaments.

Don't get me wrong, I'm not defending hot spot. How often do you see a call overturned because the hot spot picked up a nick that the ump didn't catch? If the answer is "never", then there's a problem.

[Amie]

If the stats really stand at less than 90% as you suggest, then that's really terrible for a technology this expensive. An umpire will be right more number of times than that!
The only question that remains to be answered is if you can really determine with accuracy where the sound comes from.

[Zamfir]

The (made up) 90% would only be for cases where the umpire already has doubts and therefore relies on the machine. These are often going to be the hard cases, also for the machine. It's very possible that the umpires are much less than 90% correct for these cases.

And we don;t know that the machine had made two mistakes. For that, you would need an even better technology to verify it. That might be done in a laboratory setting, but then you have to be very sure that the relaibility results transfer to the field.

[Amie]

It's very possible that the umpires are much less than 90% correct for these cases.

I doubt that very much. At least when it comes to experienced and well seasoned umpires. The technology, in my view, is more often than not, used for validating an already made decision in the umpire's head. When you've been standing in countless matches, you develop that kind of a keenness.

[ElWanderer]

How can we confirm if the noise can't have come from anything else?

This is what the Snickometer is for - the sound recording from the microphone (in the stumps I presume) is lined up with the video from the cameras. The tricky part of the technology is precisely synchronising the two (the sound & video on my telly at home often wander in and out of sync - it used to be really noticeable when we had TiVo recordings of NTL cable television), but I don't know how that is done. Then the 3rd umpire scrolls through the frames until they find ones where there is noise or where the ball appears to be passing the bat. If there's a noise, they look at the position of the ball, and possibly also the shape of the soundwave.

If there is a large noise on the recording, but the ball is visibly nowhere near the bat at that timestamp, then clearly the source of the noise is not a nick. If there happens to be a noise with a sound wave similar to that of bat on ball at the same timestamp as the ball is passing the bat, then that's most probably a nick. Yes, it could conceivably be something else that has happened to have made a similar noise at that precise moment, but it is very unlikely. It is impossible to have a decision-making system that is 100% accurate for this kind of thing.

The technology, in my view, is more often than not, used for validating an already made decision in the umpire's head.

Ummm, I doubt that. The 3rd umpire's view on an lbw or caught behind decision has to be requested by the players when they don't agree with the decision of the on-field umpire. It's not called for by the umpire themselves. The umpire will often ask the 3rd umpire for help in determining if a ball carried, whether someone was run-out or stumped, as those are very hard to judge on the pitch, but easy for someone with a TV replay - those decisions don't need Hot Spot, Snickometer or Hawk-eye, though.

[Zamfir]

I doubt that very much. At least when it comes to experienced and well seasoned umpires. The technology, in my view, is more often than not, used for validating an already made decision in the umpire's head. When you've been standing in countless matches, you develop that kind of a keenness.

How can you know? It's surely possible that umpires are correct in more cases than a particular technology. But the only way to be sure of that is to have a third method that is more reliable than either of them, and use that verify them both.

[Amie]

This is what the Snickometer is for - the sound recording from the microphone (in the stumps I presume) is lined up with the video from the cameras. The tricky part of the technology is precisely synchronising the two (the sound & video on my telly at home often wander in and out of sync - it used to be really noticeable when we had TiVo recordings of NTL cable television), but I don't know how that is done. Then the 3rd umpire scrolls through the frames until they find ones where there is noise or where the ball appears to be passing the bat. If there's a noise, they look at the position of the ball, and possibly also the shape of the soundwave.

If there is a large noise on the recording, but the ball is visibly nowhere near the bat at that timestamp, then clearly the source of the noise is not a nick. If there happens to be a noise with a sound wave similar to that of bat on ball at the same timestamp as the ball is passing the bat, then that's most probably a nick. Yes, it could conceivably be something else that has happened to have made a similar noise at that precise moment, but it is very unlikely. It is impossible to have a decision-making system that is 100% accurate for this kind of thing.

That makes, sense. Thanks.

Ummm, I doubt that. The 3rd umpire's view on an lbw or caught behind decision has to be requested by the players when they don't agree with the decision of the on-field umpire. It's not called for by the umpire themselves. The umpire will often ask the 3rd umpire for help in determining if a ball carried, whether someone was run-out or stumped, as those are very hard to judge on the pitch, but easy for someone with a TV replay - those decisions don't need Hot Spot, Snickometer or Hawk-eye, though.

Right, but that's what I am saying. The player can apparently only make two unsuccessful challenges in a match (I think it's just one challenge in ODIs - a colleague told me as much). However, this is used for confirming if the decision was right or not. If the umpire is proven wrong, then it's fine but if the player gets it wrong twice then he's lost his chance of more appeals. So the favourable position here is given to the umpire.

How can you know? It's surely possible that umpires are correct in more cases than a particular technology. But the only way to be sure of that is to have a third method that is more reliable than either of them, and use that verify them both.

Right, sorry. I meant that if in hard cases the technology being used is only reliable less than 90% of the time, then the umpire is believed to be right. Not IS right. Poor phrasing on my end.

The point is that if the player can only make two unsuccessful appeals, and if hot spot makes an error during both of those, then it is pretty much pointless.

A colleague told me that he had read the accuracy to be over 95%. Thanks for all the responses.

[Zamfir]

95% is very bad, actually. It would be great if the system was over 95% correct in difficult cases, and 99,99% correct in all other cases.

But if it was 95% correct overall, it would mostly be correct in the easy cases that the umpire can do without aid, and suck in the cases where you actually need the system.

95% would only be useful if it was completely uncorrelated with the umpire. So cases that the system finds difficult are mostly different cases from those that the umpire finds difficult. Then the combination umpire+system is better than either apart. But I doubt that this is the case.

[ElWanderer]

Right, but that's what I am saying. The player can apparently only make two unsuccessful challenges in a match (I think it's just one challenge in ODIs - a colleague told me as much). However, this is used for confirming if the decision was right or not. If the umpire is proven wrong, then it's fine but if the player gets it wrong twice then he's lost his chance of more appeals.

Yes, it's definitely two unsuccessful reviews (where unsuccessful means the out/not out decision wasn't overturned by the review) per side per innings in test match cricket. It may well be one per side per innings in the shorter forms - I don't know off the top of my head - though it also wouldn't surprise me if one day matches are different to Twenty20 matches in this respect. The purpose of the limit is to prevent a side from reviewing every single decision that goes against them; it means they only review decisions they really think have a good chance of being overturned.

So the favourable position here is given to the umpire.

I'm still not really sure what you mean by this.

[New User]

I don't watch sports much, and I definitely don't watch cricket since I live in the USA (if cricket is even televised here, it must be very obscure. Many Americans don't even know what cricket is). But when I was growing up, I remember watching sports on television with my father and the referee or umpire made some sort of foul call. I don't even remember the sport, but for the sake of example let's say it was Gridiron Football and the referee called a player as out of bounds, which would cause a stoppage of play. After the call, the camera replay showed that the referee's call was incorrect and that the play was not a foul (in this example, the player in question was not out of bounds). I remember asking my father why they allowed the game to continue, and he said that because the referee made the call and that was final. What was shown on camera was not of consequence to the game. The cameras were there to show the game to audiences at home, but the players, referees, and other game officials did not make use of them to determine if the rules of the game were being followed. Ever since that time in my life, I have always taken it to be a matter of course that the presence of video cameras at professional sports venues are not used to determine legality, and that only the referee or umpire can make such a determination. If this is no longer the case, I don't know because I don't watch sports.

So can anyone tell me? I'd like to know, how often in professional sports are camera replays used instead of a human referee to determine legality? The only examples I know of are races, for which I understand a "photo finish" has been used for years to determine which participant crossed the finish line first.

[Adam H]

Different sports have different rules. Gridiron Football (both NFL and college) now uses instant replay video evidence to overturn or confirm ref's calls. It's way overused IMO, most of the time it's not indisputable evidence so the call "stands" after a 3 minute delay.

I know they stop basketball games a lot to make sure they have exactly the right time on the clock, but I don't think they check other calls.

I'm pretty sure they don't ever question a call in baseball.

The europeans here can tell you whether football/soccer refs can be overruled (it's gotta be a no, since they never stop the game clock, right?)

[Amie]

Right, but that's what I am saying. The player can apparently only make two unsuccessful challenges in a match (I think it's just one challenge in ODIs - a colleague told me as much). However, this is used for confirming if the decision was right or not. If the umpire is proven wrong, then it's fine but if the player gets it wrong twice then he's lost his chance of more appeals.

Yes, it's definitely two unsuccessful reviews (where unsuccessful means the out/not out decision wasn't overturned by the review) per side per innings in test match cricket. It may well be one per side per innings in the shorter forms - I don't know off the top of my head - though it also wouldn't surprise me if one day matches are different to Twenty20 matches in this respect. The purpose of the limit is to prevent a side from reviewing every single decision that goes against them; it means they only review decisions they really think have a good chance of being overturned.

So the favourable position here is given to the umpire.

I'm still not really sure what you mean by this.

Yes I know it's normally two unsuccessful appeals. I was just wondering if its lesser for one days and it is.

About the umpire being favoured -- I mean to say that no matter what technology you bring in to make the decisions fairer and more accurate, if you're not going to use it enough and leave the ultimate decision to the umpire, then it's pretty pointless.

[ElWanderer]

I'd like to know, how often in professional sports are camera replays used instead of a human referee to determine legality? The only examples I know of are races, for which I understand a "photo finish" has been used for years to determine which participant crossed the finish line first.

Well, for the most part, video replays still have to be looked at and analysed by another human referee, but the ones I'm aware of (aside from cricket):

Football (soccer) has been slow to look at using technology and television replays, so pretty much the only thing replays are used for is punishing players after a match for serious foul play if the ref didn't spot it during the match. If a ref did see something (and mentions it in their report), no action can be taken later by the authorities - that way the authority of the on-field ref is maintained (even if they've made a massive mistake, so not everyone agrees with this approach). Quite a few people are in favour of bringing in goal-line technology to detect when the ball has crossed the line and/or some form of video ref for big decisions - i.e. if a goal has been scored or not, awards of penalties, sendings off.

Rugby union (don't know about league) uses a television match official (tmo) to determine if a try has been scored or not where the on-field officials think there is any doubt. I don't think there are any other situations where they can call on the tmo. The tmo looks at replays from all available camera angles, taking as long as they want (the match clock is stopped). If replays are inconclusive (as they sometimes are - when you see five replays from five different camera angles and still can't work out what happened, you feel for refs who have to make snap decisions based on their view alone), it depends which question the ref asked the tmo.

• "Was it a try, yes or no?" means a try can only be awarded if it is clearly shown on the replay - basically the on-field ref has enough doubt that they wouldn't award the try themselves if there wasn't a tmo to ask.
• "Is there any reason why I can't award a try?" means a try is awarded unless there is a clear reason why it can't be from looking at the replay - the on-field ref is pretty sure a try has been scored, but wants confirmation there weren't any infringements he didn't spot.

In tennis, I think a player is allowed to challenge up to three line calls per set. If a line call was incorrect, the point is replayed. It's similar to the cricket decision review system for lbw decisions, in that the players request this rather than the umpire (who already has the power to over-rule line calls they believe to be incorrect) and the decision is made by the ball-tracking camera technology, rather than a human.

With something like Formula 1 racing, the stewards' decisions are pretty much all based on video replays as the sport is (physically) too big for one person on the sidelines to make a decision based only on their own sight. They also have access to telemetry data from the cars, which helps to check for various technical infringements.

About the umpire being favoured -- I mean to say that no matter what technology you bring in to make the decisions fairer and more accurate, if you're not going to use it enough and leave the ultimate decision to the umpire, then it's pretty pointless.

Italics are my emphasis - define "enough"

How often to use the technology in some sports is one of those big debates that may never be resolved. It's pretty pointless if you use the technology so much that it drastically slows down matches and undermines the authority of the on-field umpire, so there have to be some limits. But where those limits should be is another matter. To me, 1 unsuccessful review per innings sounds appropriate for Twenty20, but possibly 1 too few for a one day game.

Everyone wants decisions to be made correctly and consistently, but they also want them swiftly. Who'd be a ref, eh?

[Amie]

How often to use the technology in some sports is one of those big debates that may never be resolved. It's pretty pointless if you use the technology so much that it drastically slows down matches and undermines the authority of the on-field umpire, so there have to be some limits. But where those limits should be is another matter. To me, 1 unsuccessful review per innings sounds appropriate for Twenty20, but possibly 1 too few for a one day game.

Everyone wants decisions to be made correctly and consistently, but they also want them swiftly. Who'd be a ref, eh?

I agree with you. It *is* a huge debate. Also, I know all those things I am not a cricket "n00b" so to speak I raised these questions here because I was really curious about the "science" behind all these technologies and because it came up at work. I have to think about it because I might want to do an article (I work for Wisden India) on these limits and the authority of the umpire and the duration of the game -- all coming into question because of these new technologies. When I say not "enough" I mean the price that they're paying for it is so huge that there have been lots of debates on who should incur the costs. The ICC wants the broadcasters to pay for it and vice-versa. The meat of the matter is, as you have mentioned, "where those limits should be" and nobody is talking about it anymore. The DRS (Dravid Removal System xD) deserves a more serious look anyway. If the same thing were to happen to, say, Sachin Tendulkar, there'd be an uproar and studies and researches and whatnot.

[ElWanderer]

I agree with you. It *is* a huge debate. Also, I know all those things I am not a cricket "n00b" so to speak I raised these questions here because I was really curious about the "science" behind all these technologies and because it came up at work. I have to think about it because I might want to do an article (I work for Wisden India) on these limits and the authority of the umpire and the duration of the game -- all coming into question because of these new technologies. When I say not "enough" I mean the price that they're paying for it is so huge that there have been lots of debates on who should incur the costs. The ICC wants the broadcasters to pay for it and vice-versa. The meat of the matter is, as you have mentioned, "where those limits should be" and nobody is talking about it anymore. The DRS (Dravid Removal System xD) deserves a more serious look anyway. If the same thing were to happen to, say, Sachin Tendulkar, there'd be an uproar and studies and researches and whatnot.

Ah I did wonder if you were approaching this from an Indian point of view. The BCCI's insistence that Hawk-Eye (or equivalent) couldn't be used for the England v India test series was interesting, especially as it could have saved one of the Indian batsmen from being given out LBW very early in the series (I forget who this was)! There seems to be a lot more enthusiasm about using the technology in the UK, possibly because we have long memories for when umpires make bad decisions, but easily forget when they get things right. Oh and the broadcaster that shows live cricket here is loaded with money! Heck, I rarely watch live cricket because it is on Sky Sports, which I don't have.

Charging the broadcasters does make some sense to me as they seem to use the systems more often than the umpires (or is it that they use for random bits of analysis because they've paid for them...?) and the costs seem relatively small for a large broadcaster, but then I don't know how much all the other cameras, broadcast links, technicians etc. cost.

Is it not still the case that which bits of the review system will be used for a series have to be agreed beforehand by the two teams? I thought the ICC had backed off from mandating their use.

As to the science - if we're talking Hot Spot, then I guess the question is how "fine" a nick can the system detect, compared to on-field umpires; that's assuming fine nicks are harder to detect on playback than determining whether a ball that definitely hit something hit the pad, bat or glove etc.

If someone mentions a % accuracy figure, there are various things that could be measured - it could be the % of the time the system returns the "correct" result, but does that mean from every single delivery including all the dot balls and those that hit the middle of the bat, that would never be referred, or just the deliveries that are interesting? And who works out if it is correct or not (maybe run tests with a sensor on the bat or in the ball?) If we're giving the bastman the benefit of any doubt, we might be more interested in reducing the rate of false positives than the rate of false negatives when considering times the system returns the "wrong" answer. It also depends if that figure includes the analysis by a human or just the output of the technology... I guess I'm mostly just repeating what Zamfir has already said here, I'm afraid.

I still come at this with the view that umpires will make mistakes, that can't be avoided, and that having the technology as a way to reduce the number of mistakes is a good thing, at least at the higher levels of play where it can be afforded and where it doesn't slow down the game too much. But I remember series where players seemed to be appealing with every delivery (eventually resulting in fines after the match for doing it too aggressively) and I would hate for them to be able to review every delivery as well. I guess your point is that if you lose both reviews to situations where there was a faint nick that the system couldn't detect, you'd feel quite annoyed, but I feel such situations are pretty rare. Most reviews I've seen (in my limited experience) are LBW calls, and a lot of those end up "umpire's call", because the error margins are relatively large and well-understood.

[Amie]

Ah I did wonder if you were approaching this from an Indian point of view. The BCCI's insistence that Hawk-Eye (or equivalent) couldn't be used for the England v India test series was interesting, especially as it could have saved one of the Indian batsmen from being given out LBW very early in the series (I forget who this was)! There seems to be a lot more enthusiasm about using the technology in the UK, possibly because we have long memories for when umpires make bad decisions, but easily forget when they get things right. Oh and the broadcaster that shows live cricket here is loaded with money! Heck, I rarely watch live cricket because it is on Sky Sports, which I don't have.

Charging the broadcasters does make some sense to me as they seem to use the systems more often than the umpires (or is it that they use for random bits of analysis because they've paid for them...?) and the costs seem relatively small for a large broadcaster, but then I don't know how much all the other cameras, broadcast links, technicians etc. cost.

Is it not still the case that which bits of the review system will be used for a series have to be agreed beforehand by the two teams? I thought the ICC had backed off from mandating their use.

As to the science - if we're talking Hot Spot, then I guess the question is how "fine" a nick can the system detect, compared to on-field umpires; that's assuming fine nicks are harder to detect on playback than determining whether a ball that definitely hit something hit the pad, bat or glove etc.

If someone mentions a % accuracy figure, there are various things that could be measured - it could be the % of the time the system returns the "correct" result, but does that mean from every single delivery including all the dot balls and those that hit the middle of the bat, that would never be referred, or just the deliveries that are interesting? And who works out if it is correct or not (maybe run tests with a sensor on the bat or in the ball?) If we're giving the bastman the benefit of any doubt, we might be more interested in reducing the rate of false positives than the rate of false negatives when considering times the system returns the "wrong" answer. It also depends if that figure includes the analysis by a human or just the output of the technology... I guess I'm mostly just repeating what Zamfir has already said here, I'm afraid.

I still come at this with the view that umpires will make mistakes, that can't be avoided, and that having the technology as a way to reduce the number of mistakes is a good thing, at least at the higher levels of play where it can be afforded and where it doesn't slow down the game too much. But I remember series where players seemed to be appealing with every delivery (eventually resulting in fines after the match for doing it too aggressively) and I would hate for them to be able to review every delivery as well. I guess your point is that if you lose both reviews to situations where there was a faint nick that the system couldn't detect, you'd feel quite annoyed, but I feel such situations are pretty rare. Most reviews I've seen (in my limited experience) are LBW calls, and a lot of those end up "umpire's call", because the error margins are relatively large and well-understood.

Oh I don't disagree about the fact that we need technologies to reduce the risk of wrong decisions! I am not looking at it from an Indian point of view, believe me. If I were I'd be mad at all things DRS! You're right, most decisions end up as finally being the umpire's call because of the giant error margins. That's my point -- should we be using that technology at all then, if we have to pay such a high price for it, only to finally favour the umpire over everything else and know that the device can be erroneous? Maybe if it were cheaper and/or if its use can be extended to more than just a couple of unsuccessful appeals (at best) in the one day matches. I thought the BCCI people were characteristically being jackasses when they refused it in bilateral series but at the same time, I need to know if it's really worth all the money and the hoopla surrounding it. I know there are constraints with the match duration and stuff but that's exactly why we should reconsider these new, horribly expensive inventions. I am not saying it because I am being favourable towards Indian sentiments -- that's quite far from the actual truth.

[ElWanderer]

Oooops, sorry if I'm mischaracterising your views or putting words in your mouth.

I guess I'd need to see how the costs stack up compared to all the other things involved in a match before deciding if they're too expensive or not. Hopefully the costs (and error margins) can only go down as the technology matures.

Maybe it would help if the umpires could call on the technology more often themselves if they're unsure, rather than having to make a decision and wait to see if the players choose to challenge it (assuming they have any reviews left over).

Meanwhile, I see Dravid has retired from internationals. I can vaguely remember his test match debut - cricinfo tells me that was Dickie Bird's last match as umpire - didn't we (England) drop Dravid in single figures then he went on to make (almost) a century? That could describe various players that made their debut against England around that time, mind.

[Amie]

Meanwhile, I see Dravid has retired from internationals. I can vaguely remember his test match debut - cricinfo tells me that was Dickie Bird's last match as umpire - didn't we (England) drop Dravid in single figures then he went on to make (almost) a century? That could describe various players that made their debut against England around that time, mind.

Yes. I am shattered. Absolutely devastated. He was one of the few players that I loved to watch and it was doubly special for me because where I grew up, Sachin Tendulkar was the god. Dravid was more of a sidekick. To me, he is the embodiment of a perfect cricketer. I remember seeing his various faces in victory and in anguish. He was always the antithesis of what my birth city's population thought a "perfect" batsman to be. But he's the reason I watched cricket even though it was only in small doses (I have always loved football more) and he's the only reason why I went back to watching cricket. I can remember, as I'm sure many Indians can, his disbelief and pain when he dropped a catch as a slip... he sat motionless on the ground for a while. I also remember how he used to stay and bat for long minutes without scoring, buckets of sweat trickling from his helmet, only to score the monstrous unexpected six or four.

Every sport has its heroes and Dravid is the perfect ode to the spirit of Cricket - the gentleman's game.

[lutzj]

HULK pretty sure they don't ever question a call in baseball.

They've actually just implemented video reviews in the MLB in a very small, precise set of situations, such as when a ball bounces off the high part of the fence in the outfield and it needs to be determined it hit the fence (not a home run) or something above the fence (home run). Other than that sort of thing, the MLB has a tradition of steadfastly preventing computers from coming in and taking umpires' jobs, with arguments like "mistakes by umpires are okay because they add a 'human element' to the game."

[Nath]

Meanwhile, I see Dravid has retired from internationals. I can vaguely remember his test match debut - cricinfo tells me that was Dickie Bird's last match as umpire - didn't we (England) drop Dravid in single figures then he went on to make (almost) a century? That could describe various players that made their debut against England around that time, mind.

Yes. I am shattered. Absolutely devastated. He was one of the few players that I loved to watch and it was doubly special for me because where I grew up, Sachin Tendulkar was the god. Dravid was more of a sidekick. To me, he is the embodiment of a perfect cricketer. I remember seeing his various faces in victory and in anguish. He was always the antithesis of what my birth city's population thought a "perfect" batsman to be. But he's the reason I watched cricket even though it was only in small doses (I have always loved football more) and he's the only reason why I went back to watching cricket. I can remember, as I'm sure many Indians can, his disbelief and pain when he dropped a catch as a slip... he sat motionless on the ground for a while. I also remember how he used to stay and bat for long minutes without scoring, buckets of sweat trickling from his helmet, only to score the monstrous unexpected six or four.

Every sport has its heroes and Dravid is the perfect ode to the spirit of Cricket - the gentleman's game.

I remember Dravid and Ganguly joining the squad at about the same time. Dravid was basically the Indian cricket team's Spock. On the other hand, I can't really imagine Spock doing a 'jam-jam-jammy' commercial for Kissan.

[Amie]

He may surprise you

I'm still not over it.

[K-R]

I know they stop basketball games a lot to make sure they have exactly the right time on the clock, but I don't think they check other calls.

In the NBA, the referees can check:

• Whether a made shot was for two or for three (checked during the next timeout so as not to hold up play)
• Who touched the ball last before it went out (final minute only)
• Whether a made field goal occurred before or after the buzzer sounded to signal the end of the quarter (may also apply to the shot clock, I'm not sure)
• Any situation in which a Flagrant 2 (automatic ejection) foul is called, to determine the extent of the contact
• Any fighting situation, to determine who should be charged with what

Under FIBA rules, the only thing that can be checked is whether a made field goal at the end of the final quarter was released from the hand before or after the buzzer sounded to signal the end of the quarter, and whether that shot should be worth two or three points.