INBOX

From our readers

Decision Review System

August 10, 2013

A systematic approach to bettering the DRS

Samir Srivastava


Umpire Tony Hill signals for a review of Shivnarine Chanderpaul's lbw decision, West Indies v England, 1st Test, Kingston, February 6, 2009
Instead of aligning its components, DRS pits the on-field umpires against technology and the third umpire © Getty Images
Enlarge
Related Links

The chances of the Decision Review System (DRS) creating another major controversy during the current Ashes series seem greater than another Watson lbw dismissal. And going by the ICC's reaction, the chances of the DRS getting fixed any time soon look even more remote than Watson's waft getting a technical fix. The ICC got it right when it recognised that it was creating a decision-making "system". But it has been ignoring some critical aspects about systems when implementing the DRS.

Systems are about different components working in tandem towards a goal. Instead of aligning its components, the DRS pits the on-field umpires against technology and the third umpire. But first, we need to ask if the DRS is working towards a sensible goal. Should the DRS really be about eliminating howlers? At the elite level, an umpiring howler ought to be an exception. If it isn't, the problem lies elsewhere and throwing a multi-million dollar technological solution at it is not the answer. Surely any decision-making system should try to get as many decisions correct as possible but the desire to eliminate only a howler or two amounts to the DRS seeking very low hanging fruits.

Can the DRS aim for those elusive fruits at greater heights? If yes, at what cost? This brings us to another aspect about systems. It can be counterproductive to try to optimise all the components of a system. Usually, individual components within a system must work within themselves to optimise the whole system. Everyone seems to agree that we cannot get 100% of the decisions right all the time. The challenge therefore is to use DRS in a manner that minimises errors.

The errors are of only two types: a batsman is either declared out wrongly (the false positive decision) or wrongly declared not out (the false negative decision). The DRS can be optimised to almost entirely eliminate false positive decisions. An automatic review of all the out decisions, if implemented sensibly, would ensure this. Such a system would also eliminate the need to grant reviews to the batting team. For sanity to prevail, inconclusive replays should, at all times, result in the benefit of doubt going to the batsman (and not to the on-field umpires as happens currently). So no Hot Spot evidence would mean not out. Similarly, unless there was conclusive proof of a clean catch, the batsman would continue batting.

But what about borderline leg-before decisions? For instance, should a ball that is predicted to merely shave the bails win a positive verdict? The answer is yes, provided we were confident in the accuracy of the predictive paths. If the current technology leads to accurate predictive paths 93% of the time, the solution is to alter the dimensions of the stumps during TV replays by 7%. If a ball is then predicted to kiss the stumps, it would be a kiss of death.

As far as false negative decisions are concerned, the aim should be to reward the bowling team. They could therefore be granted two reviews for every hour of bowling on a non-cumulative basis - if not used within the set time, the two reviews would lapse and a new cycle would commence. Each unsuccessful review would still result in forfeiture, but only for the the hour, after which both reviews are restored. The on-field umpires should be at liberty to seek help from technology at any stage for any decision, however, the bowling team should continue to enjoy the right to seek a review, if only to feel reassured that they are not left completely at the mercy of the umpires' judgment. The suggested changes will lower error rates to an extent that they become a non-issue.

If the current error rates are in single digits, then is all this fuss over fixing the DRS justified? Some argue that the unfair false positive decisions received during the course of a batsman's career get balanced by the lucky false negative decisions they enjoy. But this logic completely ignores spectator sentiments. The spectators judge the DRS by its ability to produce a correct decision in the heat of a contest and that is exactly as it should be. The nature of the game is such that a lot rides on a moment. We should therefore try to eliminate the DRS errors to the extent the human-technology interactions currently allow us to.

That said, the question whether greater certainty in decision making can come only at the cost of slowing down the game remains. Could the proposed changes take the joy out of the spectacle? They potentially could, but not if the ICC really think through how the different components of the DRS interact with each other. In fact, it is possible to not only improve the accuracy of decisions, but also simultaneously save time, add drama, and enhance the balance between bat and ball. Five components of the DRS appear to be particularly critical in this regard: umpires, no-ball decisions, line decisions, Hot Spot and player behaviour.

 
If the current technology leads to accurate predictive paths 93% of the time, the solution is to alter the dimensions of the stumps during TV replays by 7%. If a ball is then predicted to kiss the stumps, it would be a kiss of death.
 

Umpires: The simple fact is that human faculties are not equipped to accurately sense the sorts of things that umpires are expected to adjudicate on. Umpires get a vast majority of their decisions right primarily because of an intuitive feel for the game acquired through hundreds of hours spent making decisions. But concentration lapses can interfere in a big way. To overcome this problem, the four ICC-nominated umpires for each game could rotate among themselves and remain fresh throughout the five days. Ideally, an umpire should be given an hour's break after every two or three hours of officiating. The on-field umpires should also be given pocket monitors to view the replays during the review process. To involve spectators and for greater transparency, the ICC should permit TV studios to telecast audio feed of the umpires conversing with each other during the review process.

No-ball decisions: The need to monitor no-balls means that the umpires must, for a fraction of a second, focus away from the main area of impending action. To overcome this, cricket should consider adopting the electronic line judge used in tennis, which may require a second crease line about 12 inches further away and parallel to the popping crease. No part of a bowler's foot would be allowed to touch the front line (i.e. the line closer to the batsman). A touch would set off a 'no-ball beep' and give more time to the batsmen to react than is normally the case. Though this rule would result in the taller bowlers (those with shoes larger than 12 inches) bowling a bit behind from where they currently bowl, and the shorter bowlers gaining a couple of inches, which should not make any material difference. The back foot no-ball law could be done away with because it is not clear what advantage the bowlers gain by cutting the return crease. Further, waist-high and head-high no-ball calls could be delegated to the third umpire. This would involve factoring in each batsman's height while using the ball-tracking software. And since most of the dismissals require a no-ball re-check, automating no-ball calls would eliminate a large number of replays.

Line decisions: Altering other line decisions could also save a lot of time and facilitate decision making. The ball should remain in play until it touches the boundary line, the coordinates of a fielder's body should not really matter. To prevent a six however, fielders should be expected to remain within the boundary line while trying to keep the ball in play. This would mean making it essential to have a static boundary line. A movable rope would be unacceptable. The current law on catches on the boundary line is a sensible one and should be left untouched. The line laws pertaining to stumpings and run-outs too could be made more television-friendly. The line demarcating the crease should belong to the batsman. It is so much easier to see whether the foot or the bat is on the line than it is to ascertain whether some part of it is inside the crease. Once a batsman has dived to complete a run by touching the line or some part of the crease, the run should be deemed complete. Bat being in the air thereafter should not matter (provided the batsman is in the act of completing the dive and not turning for a second run). Some balance could be restored by ruling that direct hits would not result in overthrows.

Hot Spot: Hot Spot seems to be highly error prone and is known to be susceptible to weather conditions. One wonders whether it might be possible to wrap or spray the bat with materials that make the thermal signatures more visible. Such an approach could prove very cost effective and make Hot Spot decisions less controversial. Generally speaking, sound should not supersede Hot Spot's verdict. Audio signals can be misleading. The woody click from an errant shoelace once prevented Rahul Dravid from asking for a review. It took an observant Shane Warne in the commentary box to spot the culprit.

Player behaviour: In a recent article, Sambit Bal makes a case for demanding honesty from the players. Some of his ideas could be formally captured in an honour code that should, among other things, expect: (i) batsmen to walk if they were certain about having nicked the ball and if the umpire (not the fielder) confirmed that the catch was clean, and (ii) fielders to immediately inform the umpire if they were unsure about the legality of a catch. The players could also be encouraged to take an oath at the start of play on each day pledging against involvement with bookies and use of performance enhancing drugs. Indeed, efforts to improve the DRS would not amount to much if the players were to lack integrity or if the technology was not administered by well-trained umpires and technicians.

While TV studios could continue owning and operating the DRS technologies, decisions about which pictures to use during the review process ought to be made by neutral match officials. Furthermore, an ICC-appointed technical committee could be tasked to develop minimum acceptable parameters for the various DRS technologies.

Admittedly, there is no such beast as a perfectly administered DRS. But as Simon Taufel said in his MCC Spirit of Cricket lecture, "The technology genie has been let out of the bottle and it's not going to go back in." We should now aim to tame the genie and ask it to grab for us those elusive fruits in the upper echelons. Trialling some of the proposed ideas in first-class cricket would be a good start.

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

© ESPN Sports Media Ltd.

TopTop
Email Feedback Print
Posted by   on (August 16, 2013, 10:19 GMT)

I think it's time for icc to sit and improve technology and players play the game according to the spirit of cricket

Posted by   on (August 13, 2013, 22:54 GMT)

@Harmske 2/2 More generally, we can choose to treat players, umpires, rules of the game, technologies as separate entities or as one entity that collectively produce a spectacle for us. Technology should be seen as part of the game and not an imposition. Of course, things need to be rigorously tested at lower-levels. While every sport's reality is different, the ICC could take a leaf out of folks who run field hockey. They have aggressively changed the rules to make the game look very attractive on TV. The ICC should at the very least debate and examine any rule that would make the game more watchable without adversely affecting the balance between bat and ball.

Posted by   on (August 13, 2013, 22:50 GMT)

@harmske 1/2 While it was sensible to switch to the front foot no ball rule, I think the impact it has had on the quality of on-field decision making has been underestimated. All my recommendation does is ask bowlers to bowl from "behind" a given point whereas the current law wants them to bowl from "not ahead" of a given point. The release point will not change. The main point is that we should consider automating these calls. There might be a better way. No-ball check replays not only cause delays but they also detract from the drama of the moment.

Posted by India_ANY_track_bully on (August 11, 2013, 7:31 GMT)

I cannot believe that people are still saying there is nothing wrong with DRS. All right keep your heads in the sand!. A very basic test anyone would do for a "hotspot" system like this is with the various tapes people use on bats to increase longevity of it (nothing sinister). Hotspot has been in place for quite a while and it's only now we are coming out with this revelation that such a basic test has not been done?! I wonder when someone will put their hands up and say ball tracking is really just based on linear extapolation.. not really tested for anything more complicated.. sorry.. oh but all this is only 15000 USD to use for the day.. thank you very much :)

Posted by   on (August 11, 2013, 5:03 GMT)

@Rahulbose -- if you think about it, we should be able to eliminate false positive decisions (i.e., no batsman should be wrongly given out). But it would be a shame if we were to reach such a state at the expense of readily reprieving batsmen in general. I have watched the decisions during this Ashes series very closely, hence the angst. Of course, you have the right to trash my opinion :-)

Posted by   on (August 11, 2013, 3:39 GMT)

@McGorium -- That cricket has been able to retain its old world charm is one of its strengths but the flip side is that the game seems to be having a hard time trying to absorb technology. After all, rules and technologies are meant for humans and not the other way round. A crease line need not be sacrosanct. My sense is that once the bowlers get used to the idea, they will find it easier to comply bowling from behind a line. It should be possible to bury sensors under the line. The batting-end sensors would naturally need to be de-activated.

Posted by   on (August 11, 2013, 3:30 GMT)

@MCGorium -- You seem to have misunderstood. Altering by 7% would mean reducing the dimensions during the replay by 7%. If a ball is predicted to hit a smaller target, the chances are that it would certainly hit the larger target in reality. You are right about the accuracy and precision of Hawk Eye. We certainly could do with more evidence and studies conducted by independent agencies.

Posted by Rahulbose on (August 11, 2013, 1:35 GMT)

"An automatic review of all the out decisions, if implemented sensibly, would ensure this. "

"So no Hot Spot evidence would mean not out. Similarly, unless there was conclusive proof of a clean catch, the batsman would continue batting."

Another technology "expert" who has not bothered to watch the games and reviews in action. Who thinks its a system engineering maths problem. Its thanks to such misguided thinking that we have the DRS and DL systems in effect.

Posted by ActionJacksonMan on (August 10, 2013, 21:28 GMT)

Or, here's a far simpler solution: we return to the days of the on-field unpires' decisions being treated as final, thereby doing away with all the delays, controversies, arguments and complaints that the DRS has brought into the game. Players should simply respect the decisions of the on-field umpires (right or wrong), and accept that good and bad decisions will always balance out over time. Our sport would certainly be the better for it.

Posted by McGorium on (August 10, 2013, 19:44 GMT)

The proposal on No Ball decisions also make little sense: The law, as worded, is that some part of the foot has to *land behind* the crease at delivery-point. The foot may then slide forwards, and this is legal. That's the law that needs to be enforced, and not some arbitrary line drawn that bowlers now have to have their front-foot land behind. We don't alter the laws to make it easier for the human umpire, so why do it for a machine? The move away from the backfoot no-ball rule made things harder for human umps, despite it giving barely any advantage to even the tallest of bowlers. If you want to simplify laws, we can start with the LBW law: come up with some law that allows an easy decision, and problem solved; no need for DRS at all. Also, where exactly are you going to place the line detector? Unlike in tennis, the crease is surrounded by= a playing field 70m in each direction. The current method is to use video cameras+3rd ump, which works just fine.

Posted by McGorium on (August 10, 2013, 19:16 GMT)

I don't understand how increasing the size of the stumps by 7% makes Hawkeye (HE) more accurate. If HE is indeed 93% accurate (we don't know, we only have the manufacturer's word for it), in means that on *average*, 7% of the decisions are wrong. It says nothing how wrong they were, so increasing the size of the stumps by 7% is nonsensical. You could be off by 10cm when you are wrong, and the accuracy would still be 93%. I.E., we don't know if hawkeye is precise (not the same as accuracy), and what the precision is: Would hawkeye produce the same result for the same sample if measured at a different time, or under different conditions? Due to a reduction in frame rates in low-light conditions, HE actually has fewer frames/sec to work with, and hence the errors go up. HE has trouble tracking the ball in and out of shadows, such as late evening. The viewer gets presented a curated view of HE: if the errors are high, they choose not to show it, so we don't really know how bad it can get.

Comments have now been closed for this article

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

FAQ ►

MORE IN INBOX

A lying father

Indian fathers can watch all the matches they want to only if their sons grow...

Low ebb? What low ebb?

If you take the raw emotion of being a supporter out of it, this England side...

It's the build-up, silly

The entire cricket-watching public will have their eyes peeled when MCC takes...

DOPPELGANGERS

GAMES

  • CROSSWORD
    crossword