New Zealand v South Africa, 2nd Test, Hamilton March 17, 2012

Umpires should overrule Virtual Eye errors - inventor


Ian Taylor, the founder of the ball-tracking system, Virtual Eye, has called upon technicians and umpires to "fess up" and overrule DRS technology if they conclude that the system might have become unreliable because of unavailable data.

Taylor's call for umpires to use their discretionary powers to overrule technology followed a malfunction of the ball-tracking technology during the Hamilton Test between New Zealand and South Africa when data was collected from only two of the four cameras used to determine DRS decisions.

The failure by technicians and umpires to admit the shortcomings immediately will further embarrass the ICC, which continues to champion the system against resistance, particularly from India, and which has arranged for Cambridge University to run independent tests into its accuracy.

The latest debate broke out following the dismissal of the New Zealand batsman, Ross Taylor, on the second day of the second Test.

Taylor was given out lbw by umpire Billy Doctrove off the bowling of Dale Steyn late on the second day. The delivery was reverse-swinging and struck Taylor on the full. Ball-tracking predicted it would have hit middle and leg, which would have required a significant and unnatural angle change. In fact, after Virtual Eye was recalibrated, the ball would have only shaved leg stump. Doctrove's decision would still have been upheld by the television umpire, Aleem Dar, but the difference was still striking.

Ian Taylor, shaken by criticism of the system for the second successive Test, called a media conference in Hamilton at which he urged umpires to use the discretionary powers given to them to overrule DRS technology if they knew that complete data was not available or even if they felt strongly about a particular decision.

Virtual Eye's inventor said that the system had erred because after 6pm the ball-tracking cameras were not able to pick as much data up as they were earlier in the day when the light was brighter. In the case of Ross Taylor's dismissal, of the four cameras at the ground, one was facing direct sunlight, so it did not get the image, and another lost the ball, meaning that the data calculated for the predictive path was only obtained from two cameras.

In such a situation, Taylor simply wants his team to press the button that says "insufficient data available" and leave the umpire to take the final call on the dismissal. The implication was that they had not done this. "I said I would fess up if we got it wrong and we've got it wrong this time," Taylor said.

Ball-tracking technology came under the microscope during the first Test in Dunedin when both Jacques Kallis and Doug Bracewell said their respective camps questioned the legitimacy of the system. Kallis comment that "99% of cricketers" will probably agree that they don't believe ball-tracking is as accurate as the makers say it stung Taylor particularly hard.

He was close to pulling out of the series but was convinced to stay on after crisis talks with Sky Television, New Zealand Cricket and the ICC's general manager of cricket, Dave Richardson. "I really wish we had pulled it from the series then the players would have to argue with their boards whether they want to have it or not," Taylor said in Hamilton.

Meanwhile, Ross Taylor said he had no qualms about the system at all. "It's there for both teams and the umpire gave me out and the DRS gave me out as well," he said. When asked if it made a difference that Ian Taylor revealed that ball-tracking got it wrong, Taylor was unfussed.

"I'd rather you not have told me really," he said. "It's there. Umpires make the right decisions and umpires make a bad decision, I guess it's the same as the DRS."

Edited by David Hopps

Firdose Moonda is ESPNcricinfo's South Africa correspondent

Comments have now been closed for this article

  • Venkatesh on March 20, 2012, 10:30 GMT

    Basic question - Why does no one object to using replays for Run out / No ball / Bump ball catch etc decisions? Because these are simply replays of what ACTUALLY happened, and do not require the Technology to make predictive judgements on what MAY HAVE happened! The same logic can apply to using Hawkeye / Hotspot etc to detect thin edges, bat-pad decisions, ball pitching outside leg etc decisions, and so all the above can be subject to DRS without any controversy...

    The problem comes in when we ask technology to make predictive judgements, which will always be prone to errors and cannot be proven accurate in a real match situation. So it makes good sense to leave these to the On-field Umpires' judgement - let them decide how much the ball was swinging, whether it would have hit the stumps, how high it may have bounced etc... the moment we let technology step in here we have dramas galore!!

  • Dummy4 on March 19, 2012, 17:20 GMT

    @mrmonty. The classic question, I agree with you, remains: Who will DRS the DRS ad infinitum. :-))

  • James on March 19, 2012, 16:11 GMT

    Taylor gives a good suggestion here. The truth is that UDRS improves the number of correct decisions, period. That's good for the game. Football is being ruined by not adapting technology, cricket doesn't want to fall foul from stopping embracing it.

  • Ranojoy on March 19, 2012, 13:15 GMT

    For the general information of readers, please list any sports that use predictive technology in refereeing/umpiring (this does not include ball-tracking).

  • Dummy4 on March 19, 2012, 13:15 GMT

    There was no way that ball was hitting middle and leg leave alone hitting leg... looks like this leaves a very big margin for error. Not good for the game

  • Satish on March 19, 2012, 11:47 GMT

    DRS only for eliminating howlers.. Can't howler eliminated without these expensive technological stuffs? 80-90% accuracy can be obtained simply by using normal replays with pitch map.. Pure howler can easily be eliminated by these ways.. For LBW, hitting outside line, Pitched outside leg and inside edge are the howlers.. Can't it be eliminated by replays? Unless you see good deflection or hear concrete sound through replays, don't overrule any decision.. Any howler can be easily found out like this.. Do we need to necessarily spend more than 5000$ for eliminating howlers when we can get it for free?

  • Phil on March 19, 2012, 10:00 GMT

    @SamRoy: In theory it may be good to use it only for "howlers", but how would this work? The on-field umpire gave the decision, so he doesn't think it's a howler. So the third umpire would have to check for himself. He could do this easily if the decision is out, as there would be a natural break in the game, but what about for not outs? The game have to stop after every appeal, in case it's a howler. If not, then the third umpire would have to try and get the decision made before the next ball was bowled, in which case the fielding side would slow the game down anyway.

  • Kiran on March 19, 2012, 9:34 GMT

    @ Brittop -- I have given enough reasons why the UDRS is a flop in my post. A technology that was supposed to eliminate errors by the umpires is now asking the umpires to eliminate its errors. Can anything get more hilarious than this? Have they even tested this thing thoroughly? And it is not me who started assigning spurious motives to the people backing UDRS. It is these very people who were accusing the BCCI of underhand motives when they opposed the UDRS in the series against England. The Nasser Hussains and David Lloyds of the world who were backing the UDRS as if it is the greatest technology ever now have egg on their face. Today they are saying two of the cameras were not working. Tomorrow they will say the cameraman was not working... :)

  • Phil on March 19, 2012, 9:25 GMT

    @samincolumbia: Your logic is absurd!! DRS was brought in to increase the accuracy of decision making. If it improves that accuracy from 95% to 99% why wouldn't you use it?

  • Dummy4 on March 19, 2012, 8:10 GMT

    One fundamental problem that I have always been bugged with is that the curve of swinging ball is never factored in. It is always taken as a straight line.

  • No featured comments at the moment.