Sri Lanka v Australia, 3rd Test, Colombo, 2nd day September 17, 2011

Not out the right call

The decision review system is not perfect. That's why it was right to reprieve Tharanga Paranavitana on the second afternoon in Colombo

How much can Hawk-Eye's lbw predictions be trusted? One-hundred percent? Not at all? If the answer is somewhere in between, then it makes sense for the ICC to apply the technology with a margin of error. And if there is a margin of error, there must be a line at which doubt disappears. And if there is a line, decisions will occasionally come down to millimetres.

Enter Tharanga Paranavitana and Trent Copeland. On the second afternoon at the SSC, Paranavitana was saved by an imaginary vertical line down the middle of the off stump. Had Hawk-Eye predicted that the centre of the ball was hitting that line or inside it, Aleem Dar's on-field not-out decision would have been overturned.

Instead, a fraction less than half the ball was within that 'out' zone and Copeland was denied despite viewers at home seeing the graphic of a ball smashing into off stump. "That's ridiculous. That can't be the umpire's call," Tony Greig said on the TV commentary. Twitter lit up with denunciations of the DRS. It was an infinitesimal margin.

But millimetres have always mattered in cricket. Why should this line be any different?

Pitch the ball a whisker outside leg stump and lbw is out of the question. Fail to plant a sliver of your heel behind the crease and you've bowled a no-ball. Slide in the outfield and kiss the boundary rope with body or clothing and you give away a four. These lines provide players, umpires and viewers with certainty.

So does the imaginary thread that bisects each of the outside stumps. A player knows that when he makes the 'T' signal that unless there is overwhelming evidence that the on-field umpire got it wrong, his review will be wasted. Ball-tracking is not intended as an omnipotent umpire-killer. The review system is based on the tenet that the on-field umpire is correct unless comprehensively proven otherwise.

"It's not necessarily about benefit of the doubt to the batsman," one of the world's leading umpires, Simon Taufel said earlier in this series, "but that the benefit of doubt goes with the original decision, and that's sometimes hard for people, and sometimes umpires, to get their head around."

To remove that premise is to turn umpires into no more than middlemen between computers and players. Retaining some of the human element in a game that has relied on it for 150 years cannot be a bad thing.

The answer is not to throw out the predictions made by ball-tracking entirely. A bowler should be able to have a not-out decision overturned if the ball is clearly hitting the middle of middle stump. But neither should ball-tracking be trusted implicitly on something so tenuous as clipping the top outside corner of the bail.

Significant faith in the technology would also be required if the imaginary line was to be moved to the outside of the stump instead of the middle. Is the science that good? Why not err on the side of caution?

Nor should the TV official tell his colleague what percentage of the ball is hitting and allow the on-field umpire to make the call. Some umpires would overturn if 1% of the ball was hitting and others would refuse to do so even if 49% was striking the stumps. Players deserve to know where they stand before requesting a review.

Under the current system, the rules are rigid and transparent. Perhaps a team could be allowed to keep its review if an lbw is shown to be an "umpire's call" according to the system. If it's that close, why penalise players for asking the question?

But unless ball-tracking is thrown out entirely or trusted completely, there must be a line. Wherever that imaginary border is placed, there will be line-ball decisions. And there's nothing wrong with that.

Brydon Coverdale is an assistant editor at ESPNcricinfo

Comments have now been closed for this article

  • David on September 18, 2011, 8:09 GMT

    The problem I have always had with HawkEye is calibration. Ideally you'd set it up in the morning and someone would throw down 100 balls or so - the system would process them and predict a hit or miss at the crease or just a tad further out. You could then compare that to the actual results. However it is clearly unacceptable to throw a hard cricket ball 100 times into a pitch about to be used! Could you do it on the pitch next to the live one? Unfortunately not as then all the angles for the cameras have changed and you have calibrated nothing.

    Another interesting thing is that the people who push hawk eye have never offered this sort of demo even in a controlled environment. (What a great doco that would be - "The Proving of HawkEye!". I reckon that a demo like this would make me a lot more comfortable with it.

  • alan on September 18, 2011, 7:44 GMT

    I was not a fan of DRS initially but am being won over to the benefits of removing the howler from the game. That said this decision was a howler and i cant understand why it wasn't out. DRS said that 3/4 of the ball would be hitting off stump. If that had of hit the stumps the stump would have been cartwheeling away! That is a howler in my book! I understand that there is a margin for error but this is not a ball just clipping the stumps and knocking 1 bail off where you can give the benefit to the umpires decision! If you are giving the benefit it should be EITHER that it half the ball has to be hitting OR that part of the ball must be hitting the middle of the stump NOT both cos if the DRS requires a margin of error of 3/4 of the ball then it is not reliable and should just not be used at all!

  • Dummy4 on September 18, 2011, 7:15 GMT

    The purpose of introducing DRS was to eliminate howlers. In the case of marginal decisions, the benefit of doubt must stay with the batsman. Players all around the world would not grudge if the above two principles are upheld. But the DRS in the current form is failing to achieve these objectives. We saw in the English summer how marginal decisions went against the batsmen (Rahul Dravid on a number of occasions), and how howlers were not overturned because of how DRS has been currently adopted (Harbhajan's LBW to Broad immediately comes to mind). Yesterday's decision was also another case in point. To the naked eye, it looked quite plumb. And hence it became another howler which was not overturned by DRS. Therefore, to justify the DRS decision of y'day by making claims about millimeters is like trying to split hairs. Administrators should attempt to achieve the fundamental objectives of DRS. We need to look at the big picture, not the millimeters as suggested in this article.

  • Dummy4 on September 18, 2011, 6:56 GMT

    Ever since India has stood out its neck protesting the unreliability of DRS, two things are conspicuously happening. the umpires are constantly going out of their way to victimise India by giving all marginal decisions and some even not that marginal against the team. And the technology available itself vindicating india's opposition to its use. Technology itself seems about fifty percent sure. So what is the point of using it if substantial doubts still remain about the decision.

  • The on September 18, 2011, 6:41 GMT

    I agree with @Dashgar that (after the series, not in the middle of one as happens when the BCCI whinges about something) they could look at the consistency of moving the line to the outside of the stump, not the middle of the stump.

    After all, this is the rectangle used to define "pitching in line" and contact for Hawkeye. Now we have half the ball used as the margin of doubt for impact but half the ball PLUS half the stump for hitting.

    I don't agree that an "ump call" should restore the appeal - they get two appeals for CLEAR errors, and a ball clipping the outside of leg is not what UDRS is there for. But I do believe the appeal should be reinstated where the "ump call" is due to technology failure rather than measured tolerances. There is precedent from that match in Wellington played in a Cat 5 cyclone when no UDRS was possible due to camera movement. If the 3U believes that UDRS was "unavailable or unable to assist with forming an opinion" no appeal is debited.

  • Dummy4 on September 18, 2011, 6:26 GMT

    Another evidence of how much precise Aleem Dar's umpiring is. He can spot even the millimeters or within 2% of that imaginary line :P

  • Mark on September 18, 2011, 3:52 GMT

    We now need DRS for all decisions when playing sub continent teams as they have shown that they are not to be trusted when claiming catches etc.

  • Chris on September 18, 2011, 3:25 GMT

    @cyniket - I get your point, but we can't trust it completely (I say that as an engineer who understands the physics involved), and decisions will always be better if a bit of judgement from the on-field umpire is involved. Meanwhile, whilst it might not be written down in the rule book, there's always been an implied benefit of the doubt given to the batsman. As for " it creates the potential injustice of two identical deliveries having different results, one team might get an lbw whilst the other team doesn't because the umpire happened to give one of them out." well that's nothing new either - DRS does make things better in that regard though.

  • Chris on September 18, 2011, 3:20 GMT

    As an engineer who understands the physics behind Hawkeye, I trust the technology more than most. However it's not 100% accurate and I agree they've got it spot on - the point of DRS isn't to make the decisions, but to get rid of the howlers, at which it does a superb job. Arguably the edge of the stumps is a better line, but everybody knows what the rules are (or if not - MSD - they should). Where I do disagree is with the team getting the review back if it's umpire's call - if the decision is that close, that's the chance they take, and allowing multiple failed DRS appeals in this way could open the floodgates to lots more appealing to DRS.

  • Tissa on September 18, 2011, 2:46 GMT

    The issue is that ball tracking lines shown are giving a wrong message to the Users showing it as a single line with same size as the ball. A SOLUTION to this is, to show the tracking line in a cone shape...... meaning after the impact, the imaginary line to be shown as a line increasing in diameter (in the shape of a cone) considering the margin of error. This will give user a real perspective and show the ball could have gone anywhere whining that line spreading wider and wider as it goes alone. Anyone?

  • No featured comments at the moment.