From our readers


August 8, 2013

DRS will survive, warts and all

Russell Marks

Michael Clarke asks for a review, Australia v South Africa, first Test, Brisbane, November 9, 2012
DRS was sold to players and spectators on the sole basis that it would practically eliminate the shockingly bad umpiring calls. It patently does not do that © Getty Images

Since late 2008 the cricket world has hotly debated the Decision Review System, or DRS. Technology had long been used by the commercial television broadcasters - especially Australia's Nine Network - to enhance their viewers' experience of watching the game. Well before DRS, fans had become accustomed to having umpiring decisions scrutinised by Snicko, Hawk-Eye and Hot Spot, all applied in extreme slow-motion.

It was inevitable that this technology would be employed to actually review umpires' decisions. Fans and commentators had always been harsh on umpires when available evidence suggested they had made poor decisions. But umpires were sitting ducks against the tides of criticism that came from living rooms - and journalists' pens - as each of their mistakes was isolated, magnified and replayed over and over when subject to advanced audio-televisual technology. Since 2006 tennis line-calls had been subject to a version of Hawk-Eye which was more or less successful in reviewing close decisions without detracting too much from the flow of the game. And cricket's initial foray into the technological review of umpiring decisions - the "third umpire" review of run-outs and stumpings - had been working well since 1992, in much the same way that controversial tries in rugby had been subject to replay and review since 1996.

But DRS soon proved far more problematic than the first generation of TV-based review. The BCCI continues to oppose it, and the ICC, while endorsing the system, leaves the final decision up to the teams involved in each match. Both teams must agree to use DRS, otherwise it isn't used. India has copped a lot of flak for its recalcitrance. When its team toured Australia during the 2011-2012 summer, the host nation's fans made little effort to hide their frustrations at the visitors' refusal to accept the inevitable.

After a bewildering series of DRS decisions in the current Ashes Tests, Australian fans no doubt have more sympathy with India's position now. Usman Khawaja's dismissal in the first innings of the Third Test had the feel of 'enough is enough'. DRS may be the least of Australia's problems in England, but it does provide an external focus for its fans' frustrations. Their most urgent debates are immediately suspended when DRS throws up yet another howler. Twitterverse explodes with indignation at the injustice of it all. Even the Prime Minister, Kevin Rudd, felt the need to tweet about the Khawaja clanger.

DRS was sold to players and spectators on the sole basis that it would practically eliminate the shockingly bad umpiring calls. It patently does not do that, for two reasons. The first relates to the system within which its use is embedded. The ICC decided to leave the use of DRS reviews up to the players. So as to prevent every single decision being reviewed, teams were necessarily limited to two unsuccessful reviews per innings - just as tennis players are limited to three challenges per set. What this means, of course, is that once a team has spent its reviews, umpiring howlers against it go uncorrected. This system could have worked spectacularly well, because it incorporates an additional level of tactical play. But it turns out that the vast majority of the cricket world appears to value accuracy over tactical play, at least on this issue.

The second reason why DRS doesn't eliminate shockingly bad umpiring calls is to be found at the intersection between the technology and its use by humans. It is rarely accepted that the technology is not and cannot be 100% accurate. Hot Spot relies on a heat signal produced by the ball touching the wood of the bat or the padding of the gloves. Produced originally for military application, it is fallible on a number of grounds, including the fact that the cameras still operate on a frames-per-second basis and so leave open the possibility that the conclusive frame is nonexistent. Hawk-Eye works on principles of geometric triangulation: the vision from six high-performance cameras is triangulated to estimate the future path of the ball based on its existing path before it hits an obstruction. Used for leg-before-wicket reviews, it is far more problematic than Hot Spot, because most of the time the input information - the actual path of the ball before it thuds into the batsman's pad - is less than a metre, given that the ball has just bounced on the pitch. The margin for error for Hawk-Eye in tennis (where the "input" data can be taken all the way from its impact with the racquet of the player at the other end of the court) is 3.6 millimetres, which is extraordinarily accurate until you consider that decisions are often confirmed or overturned on margins well within that margin for error. But in cricket, where there is much less input data, the margin for error is much larger.

Because we live in an age dominated by the ideology of technological progress, and because we tend to believe what we see (especially if it is presented to us with deceptively uncomplicated clarity - think of the crisp, animated images of the Hawk-Eye flight path in tennis and cricket reviews), we prefer to expect a level of accuracy from the technology that's just not possible. The ICC implicitly acknowledges the relatively large margin of error in its application by having the third umpire defer to the field umpire's original call when Hawk-Eye indicates that less than half the ball would have hit the stumps, or pitched in line with them. But fans (and players) just see almost half the ball hitting the stumps and wonder why the third umpire can't overrule the field umpire on that basis.

"Implicit in the claim that DRS would improve umpiring decisions is the expectation that DRS will add value, that it will not detract from the experience of playing and watching cricket."

Added to these expectations are the practical difficulties in operating the review technology as a third umpire. Much of the time, the umpire is still making judgements on the strength of two-dimensional representations of three-dimensional realities, while being seductively deceived by the strength of the images in front of him. And on what basis is he to make his decision? Should the benefit of the doubt continue to go to the batsman, as it does when the field umpire makes a decision? What is the standard of proof that he requires: the balance of probabilities, or being beyond reasonable doubt? Should his starting-point be the decision as made by the field umpire, or should he start from scratch with the evidence before him? From time to time the ICC seeks to clarify these questions. Fans and players are mostly in the dark.

Part of the reason the Khawaja decision by third umpire Kumar Dharmasena was received so badly is that while the technology appeared to show that the ball passed Khawaja's bat without hitting it, Dharmasena seemed to be looking for conclusive, positive evidence to overturn field umpire Marais Erasmus's original decision, which was "out". And almost by definition, the technology cannot produce positive evidence to show that ball did not hit bat. It's the philosophical equivalent of trying to prove that it did not rain on Tuesday, or that Santa Claus doesn't exist: negative proofs are extraordinarily difficult. Which rule was Dharmasena applying as he reviewed the replays? Australia has requested a "please explain" from the ICC, if only to clarify the process used by third umpires in DRS reviews.

The only claim that can be legitimately made about DRS is that it improves the quality of umpiring decisions on average. It is impossible to know to what extent it achieves this, because we have no window into the objective reality of what happens on the field. But advocates of DRS point to this unarguable claim, and for them, that is enough to justify its continued application. DRS advocates also tend to place their faith in the inevitability of technological progress and make the additional claim that the objective accuracy of DRS can only improve in the future.

But these claims ignore a nagging doubt in the minds of DRS sceptics and opponents. Implicit in the claim that DRS would improve umpiring decisions is the expectation that DRS will add value, that it will not detract from the experience of playing and watching cricket. In 2009 this implicit claim was unarguable: better umpiring decisions would mean better cricket. But this Ashes series has seen doubts about this implicit claim crystallise for two reasons. Firstly, it seems that DRS has sharply raised players' and fans' expectations of good umpiring: even if it could be shown that the average quality of umpiring decisions has improved, this average has not risen commensurate with our expectations. And secondly, the long delays associated with each review, together with the frustrations inherent in the two-review limit, have led to an overall increase rather than decrease in the frustrations inherent in watching and playing cricket.

An image captured from the new Hot Spot camera
Hot Spot cameras still operate on a frames-per-second basis; What if the conclusive frame - when ball passes bat - is non-existent? © HotSpot

While nothing is done, the calls for some kind of official response will become deafening. And yet it is now extraordinarily difficult to know what to do from here. At one end of the debate is the call to scrap DRS entirely. If this happens, the situation would revert to the way it was before 2009: extended, slow-motion, close-up replays would show particular umpiring decisions to be shockers, and nothing could be done about them. I for one would be comfortable with that, but I expect that umpires and players whose livelihoods depend in part on the quality of umpiring decisions would not. The only solution to that problem - banning telecasters from using technology to critique field umpires' decisions - is untenable on commercial, practical and commonsense grounds, and perhaps it even violates the way we like to understand our freedoms of innovation, enterprise and even communication. At the other end of the debate is the suggestion that teams be allowed unlimited reviews, or that the power to review a decision be placed with the field umpires themselves. Both would have the same ultimate effect: every single decision would be reviewed. This would quickly prove untenable. Australia trialled this in its domestic one-day competition last summer, and abandoned it part-way through: it made field umpires redundant, and it slowed the momentum of play to such an extent that even fans of a game its detractors compare to watching grass grow were driven to distraction.

I began writing this article expecting to conclude it with a call to abandon DRS. But, as the technological ideologues say, the genie is well and truly out of the bottle. Like representative democracy, DRS may well present the "least worst" of all possible alternatives, given current constraints. Some tinkering may improve it, or not: what seems vital is that the ICC and/or its Elite Panel of Umpires ensures that the cricketing world - including umpires, players and fans - better understand the limitations of the technology and the rules which govern its application. Australian commentator Jim Maxwell has suggested that the third umpire has an inherent conflict of interest in wanting, above all, to confirm the decision of his on-field colleague, and that this conflict requires third umpires to be drawn from a pool different to that of field umpires.

Perhaps. Whatever happens to DRS from here, it will stay, and largely in the same form. The cricket world will need to come to terms with the fact that science has introduced to it a tool which at the same time adds to and detracts markedly from the enjoyment of watching and playing the game. Cricket has survived other such innovations: bodyline; professionalisation; commercialisation; Twenty20. I daresay it will survive DRS too.

If you have a submission for Inbox, send it to us here, with "Inbox" in the subject line

© ESPN Sports Media Ltd.

. Your ESPN name '' will be used to display your comments. Please click here to edit this.
Comments have now been closed for this article

Posted by Gautam on (August 10, 2013, 14:19 GMT)

How people can still claim "DRS is working" befuddles me. Their claim that NEW technology has increased correct decisions is dubious. The real question is what % of those improved decisions could have been resolved by just looking at slow mo action replay. I would bet that more than 90% of improvements can be realized just be allowing 3rd umpire or even field umpires to look at replays. Now if it is indeed proven that slow mo can realize 90% of improvements - so basically going from say 95% to 97.7% - and rest of technology adds only to 0.3% - will DRS supporters be men-enough to admit rest of fancy technology genie can be put back into bottle without further ado?

Posted by Dummy4 on (August 9, 2013, 12:35 GMT)

There are very many welcome efforts to improve the existing technology. But, I don't see any serious effort to train the umpires in using it right and consistently. Many are talking about the need for guidelines / rules / protocol for its application. But, that will not be enough. Training will continue to be the missing link. What Dhoni said a few years ago will still be valid. To paraphrase him: He is not against the technology, but is against the corruption of technology with human (mis)judgements. How accurate he was; and how prescient he turned out to be in this series.

Posted by Dummy4 on (August 9, 2013, 12:21 GMT)

DRS will have to be there and try to make it 100% accurate so that BCCI will accept it with the series against South Africa is very important

Posted by Dummy4 on (August 9, 2013, 10:38 GMT)

Come on guys. Its evolution stage for DRS. Pretty simple!

Posted by Indian on (August 9, 2013, 7:45 GMT)

In the "light" of this matter can we now please refer to hotspot as "Coldspot" and the DRS be officially renamed to "Dodgy Random Simulation" :)

Posted by Anand on (August 9, 2013, 7:44 GMT)

@jmcilhinney on (August 9, 2013, 2:05 GMT), as you say Sniko is not real time as yet - but as far as i am aware it will be an option by the Winter Ashes series so not far away from being an option.

With regards speculative reviews and time wastage, as a spectator i would have to say i dont fined reviews detract from the flow of the game. If anything they are the moments of high drama - particularly for the TV viewer. The issue of time wastage is more relevant to teams slowing down over rates etc.

Posted by Ashish on (August 9, 2013, 5:51 GMT)

Extremely good article. However, it misses another human angle/bias on the part of the on-field umpires. If one team has exhausted its quota of DRS reviews, the on-field umpires will be subconsciously biased against that team as any unfavourable decision against the team with available reviews is likely to be challenged. I feel India are justified in their doubts about DRS.

The other point missed by this article (but mentioned by others in this debate) is the fact that the benefit of doubt is moved from the batsman to the umpire.

Having watched Sky's coverage of Cricket in the past few years I am very much interested in Sir Ian Botham's blatantly partisan views on the Indian position on DRS. He had criticised their position very vociferously. Where is Sir Ian Botham now?

Posted by John on (August 9, 2013, 4:10 GMT)

@rattusprat on (August 9, 2013, 1:12 GMT), agreed. It's said that "absence of evidence is not evidence of absence" but in the case of Khawaja's dismissal, the absence of both a mark on HotSpot and a noise on the stump mic may not be proof the he didn't hit it but it is evidence. It's also evidence that the on-field umpire was mistaken in the first place because what were they then going on? I think that the on-field umpire thought that there was a noise but that was shown to be bat on pad before the ball passed the bat, so the original decision was still wrong, even in the unlikely event that Khawaja did edge it.

Posted by John on (August 9, 2013, 3:03 GMT)

cont... If the third umpire was looking for proof that Khawaja didn't hit the ball then really only one thing would do, i.e. a replay that always showed a gap between bat and ball. That was obviously never going to happen because of the path of the ball and the camera angles. Some people claim that the rear angle showed a definite gap between bat and ball but it did no such thing. At one point the ball was obscured by the bat so just because there was a gap later on does not mean that that is the point at which the ball passed the bat. If you could sync side and rear views then you might be able to confirm a gap as ball passed bat. Regardless, absence of a noise as the ball passed the bat and absence of a mark on HotSpot renders the probability that the ball was edged so low that I would consider that sufficient evidence that the on-field call was wrong. Even if you give the on-field umpire the benefit of the doubt, I just don't see sufficient doubt there to give them the benefit of.

Posted by John on (August 9, 2013, 2:55 GMT)

I think that unrealistic expectation is the biggest issue where DRS is concerned. I hope that I'm not misquoting but I believe that the BCCI have said that they will not accept DRS technology that is not 100% accurate. Certainly their supporters have. To say that, you have to either have no understanding how technology, and particularly this technology works or else be unwilling to accept technology at all and just use that line as excuse to avoid saying so. Noone has ever claimed that DRS technology is 100% accurate or could ever be so. How would we even measure that anyway? They have also said that the system needs to be fool-proof, which is much more realistic. To my mind, that's more about procedures than technology. It literally means that a fool could use the system and get the same result as an expert. That requires clear guidelines on its use. If Khawaja could be given out then the guidelines are either not clear enough, being ignored or are just plain wrong. tbc...

Posted by John on (August 9, 2013, 2:05 GMT)

@Ramski1 on (August 8, 2013, 15:27 GMT), it doesn't make sense for the current incarnation of Snicko to be included in DRS because, as has been said many times in many places, it takes too long. There is a real-time version of Snicko in development but, currently, Snicko as part of DRS is not an option.

With regards to teams keeping their review on a marginal LBW, fewer reviews lost and less chance of a howler being missed, but lots more speculative reviews called for. I think in that situation there is no perfect system because many already complain about the time DRS takes so more reviews will mean more of an issue there. You simply cannot please everyone, as shown by the fact that many who oppose DRS now were complaining just as loudly as anybody else about poor umpiring before its inception.

Posted by Rattus on (August 9, 2013, 1:12 GMT)

Agreed Manso. The current issue is requiring the players to judge what constitues a "howler". If a player nicks the ball to second slip and is given NOT OUT, thats a howler and can be overturned without hot spot. If a player nicks it with no deflection and is given NOT OUT, and hot spot shows nothing, then its not a "howler" and can stay not out.

If, as we have seen this series, a player doesn't hit it and is given OUT, as no hot spot cannot be considered "conclusive" you may as well not bother looking at it. On the standards adopted this series, the 3rd umpire will only overturn this one if there is a clear gap between bat and ball on a slow-mo replay. So, a batsman given out when he hasn't nicked it should only review if (1) he has missed it by at least 6 inches (how does he know?) AND (2) he knows the keepers head isnt in the way of the replay (eg. Kawaja), otherwise HE HAS WASTED A REVIEW. DRS does not function in this case, and must be changed.

Posted by Steven on (August 8, 2013, 23:38 GMT)

You make it sound as if there is an alternative to frames-per-second cameras out there: all video is a sequence of frames, there is no other technology out there. You can use higher frame rates (I think they now use 100fps, for instance), but you'll always have frames.

If there's an issue with poor heat-transfer contact, why not mandate the use of high conductivity tape on bat edges? The ICC has no problem prohibiting the use of tape on bats (remember the carbon taping fiasco a few years back?), so how about this time, they try a different approach, requiring the use of high conductive tape?

Posted by mansoor on (August 8, 2013, 19:04 GMT)

If the 3rd Umpire is not very sure himself, or finds insuffiecient eveidence and so turns down the appeal, he shd be allowed to hand back the "over ruled" DRS.

If LBW DRS and ball is marginally clipping stumps, So NOT OUT . Hand back the DRS.

Also if Field umpire HIMSELF 'not so sure' when making orignal decision, He shd be allowed to CONSULT DRS UMPIRE before making his decsion. Maybe 2 DRS Referalls shd also be allowed to each umpire.

Posted by Kapil on (August 8, 2013, 17:35 GMT)

Good case in favor of drs - it can surely improve. Easiest way is to return the review back to the reviewing team in case drs doesn't have enough data to conclude e.g. Frame that should have recorded the bat ball contact is not available. And what about adding some competition - why do we have to depend only on one company? I think hawk eye or whatever it is, is minting money for free. Lets open the market - and see what is available. Duck worth Lewis is statistical. Why can't we add statistics help predict deliveries? And for sure there ar many more brilliant ideas. Why no give review feed to more than one analyst

Posted by Asheesh on (August 8, 2013, 17:27 GMT)

DRS is a good system. It is well used. There are some exceptions at times.

Think the world needs to read your opinions on cricket? Here's your chance to be published on ESPNcricinfo.



Tales from Thiruvananthapuram

As Thiruvananthapuram gears up to host its first T20I, a flashback to visits ...

Are we entering a golden age of Indian fast bowling?

Could an Indian pace bowling pack all capable of hitting 140kph inspire a ste...

The Run-Out XI

The most dangerous runners between the wickets in cricket history, all in one...