Monday, December 26, 2011

To DRS Or Not To DRS, That Is Not The Question

Two Australian batsmen - Michael Hussey and the impressive debutant Eddie Cowan were given out, apparently incorrectly by Umpires Erasmus and Gould respectively on Boxing Day 2011 at the MCG. Both decisions were not howlers in my view, Cowan's decision even less so, even though in both cases, Hotspot showed nothing. Later in the day, Zaheer Khan was denied a very close LBW decision.

Hotspot has its problems. Warren Brennan, chief executive of BBG Sports (who provide the technology), recently said that "BBG Sports is disappointed at how the Hot Spot performed in the UK during the England vs India series." In a report from November 2011 after the Cape Town Test between South Africa and Australia (96 all out, 47 all out and all that), Brennan also said that Hotspot has had problems with blurring. This problem, as Brennan explains, arises when the player swings quickly. The appeal against Eddie Cowan falls in this category. In July this year, Brennan told Cricinfo that Hotspot's accuracy (about 90-95%) was also affected by whether the impact was on the bat or the glove.
"Things off the glove can be a bit hit and miss... There is padding on the gloves, which is obviously quite soft and other parts of the glove probably have some metal, usually on the side of the fingers. We get different heat impressions from the glove. While the metal can heat up, it'll be quite hot and when the ball strikes the soft part of the glove, it doesn't leave a great heat signature."

Technology has its limits. It is not surprising that within the last 18 months, the use of DRS has generated plenty of controversy. In the first Ashes Test at Brisbane, a clear Umpiring error could not be corrected because the side which suffered from that decision had no reviews left. Earlier in 2010, the Pakistan batsman Azhar Ali failed to refer, what, according to the technological evidence was an umpiring error against him, and was panned for not doing so. In November 2010, in a series where DRS was not used, VVS Laxman was robbed of a great second innings century when he was given out LBW against Daniel Vettori inspite of a palpable inside edge. The Australian captain Ricky Ponting was censured for dissent after he got into an argument with an Umpire about a review (which was available to everyone on the big screen) against Kevin Pietersen. In each of these cases, the evidence was available, but could not be used for various reasons - reasons which DRS did not, or would not necessarily have overcome.

When India toured South Africa, DRS was not in use. This became an issue when a bad decision was made against Zaheer Khan (who was batting) in one of the Tests, and other more marginal decisions went against AB deVilliers and Mark Boucher. A commentator was quick to use the decision against Zaheer, and add the decisions against deVilliers and Bouchers, to build a case for DRS. The commentator was South African. The evidence from the technology that would normally be in play with DRS, has been used by partisan commentators to criticize Umpires, even in marginal cases. This, you would say is merely a case of incompetent commentator, but the evidence does arm the incompetent commentator.

The 2011 World Cup saw the 2.5m rule become an issue. I later learnt that the 2.5m rule had nothing to do with the limitations of ball tracking technology and everything to do with the ICC's desire to preserve accepted and currently prevalent interpretations of the LBW law.

In the same tournament, the DRS also produced a peculiar, but largely ignored situation. M S Dhoni asked for a review after the Umpire had declared leg byes.

In the Semi Final of the World Cup, Sachin Tendulkar got a crucial reprieve when a decidedly marginal decision against him was reversed upon review.

In another (non-DRS) instance of the use of technology, the wrong replay was used to check for a no ball in a Test Match at Barbados.

When Australia were in Sri Lanka in August and September of 2011, a DRS-based LBW decision against Philip Hughes invited the attention of the ICC's premier Umpire because the ball track seemed to be palpably wrong. Umpire Taufel's comments on the matter also exposed the limitations of his understanding of how ball tracking works, which, given his status as the ICC's premier Umpire, is troubling. In the same series, there was also the problem of different pitch mats.

Finally, in Australia's recent loss at Hobart against New Zealand, ball tracking and hotspot appeared to provide contradictory evidence about the location of the pitching point of the ball.

Technology complicates things. The DRS is a complicated issue. It is not merely about the ball-track or hotspot and their limitations, even though these are significant and interesting. It is also about the many crucial decisions that are made by the ICC without which the DRS wouldn't come into being. Here is short list of the aspects of the decision review system that lie outside the limitations of cameras, microphones and computers.

1. The number of reviews per innings.
2. The definition of margin of error at the stumps in the case of the ball track.
3. The 2.5 metre rule (which has been modified multiple times)
4. The communications protocol - the inability of the TV Umpire to initiate a review.
5. The choice of applicable evidence from each technology - Hotspot can provide a more reliable pitching point than ball tracking, but this is not used.
6. The choice of which technologies will be used (in ICC negotiations, compromises have involved discarding ball tracking, and making hotspot mandatory)

The DRS has also had consequences.

It has gotten spinners to believe that it is producing more front foot LBW decisions. Graeme Swann thinks this is the case, even though only 2 of his 40 odd LBW decisions have been properly overturned in his favor upon review. In contrast, LBW decisions against batsmen off his bowling have been unsuccessfully reviewed 16 times. Shane Warne and Anil Kumble have had far more to do with the change in umpires attitudes to front foot LBWs than DRS or Graeme Swann.

It has affected the nature of the marginal decision in cricket, and as Sachin Tendulkar's exemplary case in the World Cup Semi Final shows, defines a truly marginal decision far too narrowly.

By allowing players to dispute umpire's decisions, it has completed the transition from the age of walking to an age of utter ethical agnosticism. It has also changed the nature of dissent in cricket.

The question is not whether or not DRS should be used. It is fairly obvious that some evidentiary assistance from technology is welcome in cricket. The DRS, as it is currently designed and the form in which it is currently advocated every time a decision goes wrong, goes much further - it adopts a deeply unsubtle attitude to the limits of evidence.

Some have asked why BCCI opposes the use of DRS despite India's experience in the Sydney Test of 2008. This question misunderstands what happened at Sydney. This was a Test in which the current Australian captain Michael Clarke stood after edging a ball to first slip! Steve Bucknor (and more than Bucknor, Mark Benson) and his umpiring was, but a small part of that story. There was monkeygate, the catching agreement, Australian behavior. After all, India's captain did not complain about the Umpiring after that game, he said that "Only one team was playing with the spirit of the game. Thats all I can say".


Hotspot and/or ball-tracking both have limits. They will, from time to time provide wrong evidence. At other times, they will be unable to provide conclusive evidence, or will even provide contradictory evidence. Anybody who is seriously interested in technology must surely understand this. Further, in the last 18 months every link in the DRS, has been shown at one time or the other, to be a point at which wrong or at least debatable decisions have been made. Marginal decisions have been reversed, ball track has been obviously wrong, hotspot has provided both false positives and false negatives (batsmen have walked in cases where hotspot showed nothing).

There is a fundamental problem. Cricket is trying to eliminate controversial decisions and Umpiring errors without asking players to take any responsibility. If batsmen walked as a rule (i.e. if it was an accepted and rigorously followed social convention that a batsman who knows he's out should walk), there would, by definition, be fewer Umpiring errors. Using technology is never going to overcome this basic deficit.

Using DRS is not the answer if cricket's problem is the elimination of obvious umpiring errors. The response to this is often that DRS does help address obvious umpiring errors. The point though, is that it does so at great cost. The cost is in the creation of other controversies, in a change in the interpretation of LBW decisions, in the nature of dissent, in a cricketer's understanding of his ethical responsibilities on the field. DRS also does this very inefficiently. Even Tests where it has been used have not been immune to so-called "howlers". But Tests where it has been used have been affected greatly by the reversal of arbitrary decisions, often to dramatic effect (such as Australia being 47 all out).

The obsession with DRS is preventing, or at least interrupting much needed attention that needs to be directed at addressing the basic problem. The problem can be stated as follows. The questions become successively more specific :

How can available technological evidence be used to assist Umpires in decision making? In which areas do Umpires need this help? How should this help be provided to the Umpires? What is the role of the TV Umpire in this process? How can such a process take into account the limitations of available technology? How can this be made affordable?

This questions cannot be answered in sequence, but must be taken up in dialogue. It goes without saying that answers to these questions will be limited until technologies and protocols are tested. But should these tests happen at the highest level of cricket?

Technology in Cricket is a complicated issue. If nothing else, I hope you are willing to accept this. It is not merely a choice between using DRS or not using DRS.

13 comments:

  1. This is what I don't understand (and I am an Aussie, but don't assume that this means I am suddenly all for DRS because it has adversely effected my team - I'm not, I assure you) - who decides on whether DRS will be used in a series? I was under the impression that the host nation made the decision and paid for it if it was used. If that is the case, then why has the BCCI been allowed to prevent its use in this series against Australia?

    ReplyDelete
  2. As usual, a very fine piece highlighting many issues in need of resolution. However, surely it is something much simpler that is causing so much frustration among followers of the game (and even international players) around the world? And that is the inconsistency in the application of review technology. It seems absurd that UDRS can be used in one series, but not another and that different versions of the review system are also used. In most major sports a competent governing body would have initiated trials, produced reports and recommendations, had a vote and then settled on something that would be applied universally (but with the possibility for modifications as experience was acquired and technology developed). Sadly, the ICC can't do that, and the result is a barrage of complaints and expressions of baffled frustration every time a series provides another exception to general practice.

    ReplyDelete
  3. The two teams in the series have to agree to use it. If they can agree to use it, then the host nation can include it in their contract with the broadcasters, which means that the extra requirements for equipment, technicians etc. kicks in.

    ReplyDelete
  4. Exactly, Kartikeya, that's the problem. UDRS should not be a matter for negotiation between two teams; it should not be a part of separately-agreed playing conditions. The ICC should (but won't) behave like the governing body of a major sport and mandate standard practice for every international series, everywhere. Your remarks on problems to be resolved would still hold force, and the process of finding the best practices would continue, but in a much less strained environment.

    ReplyDelete
  5. The ICC is not that kind of organization, with good reason. It is merely constituted by the 10 Test playing nations. It has no leverage over other agencies. The ICC does not negotiate TV contracts, it does not pay the players, it does not organize bilateral Test series.

    This is not an arrangement without merit.

    Cricket is unique as a major sport in that it is actually an international sport, as opposed to being an inter-franchise sport. Its boards are typically genuinely bottom-up organizations (like BCCI) rather than being constituted by one rich person or one rich corporation. It is an aspect of cricket which people always forget.

    Ultimately, I think the ICC will have to rely on the member boards to try out different things at first class level, maybe even pay member boards to do so.

    Cricket did not become interested in technology because they felt decision making was poor. They became interested because broadcasters bought tracking technology and heat signature technology and it gave partisan commentators a lot of ammunition to argue about umpiring decisions. DRS was the Cricket's way of putting out a fire started by broadcasters.

    A lot of criticism is directed at a lot of people, but serious attention has to be paid to broadcasters, who are accountable to nobody, and commentators, who are even less accountable, if this possible.

    ReplyDelete
  6. http://www.youtube.com/watch?v=b86yWY-HDYo&feature=related

    love your writing.. Wanted to point this out to you..Tony Greig's comments are just classic!

    ReplyDelete
  7. This comment has been removed by the author.

    ReplyDelete
  8. It has shortcomings. No doubt about that. But I can't figure out why the ICC is allowing this rule to be put to use in some series & not in another one. It has come to a stage where the DRS is not made use of in every series featuring India. Its perfectly okay to voice your opinion against it. BCCI & most of the Indian players don't approve of it. Fine. But you need to have uniformity when it comes to rules in this modern era.Its high time the ICC took a stand. Either they should make it mandatory or revoke it once and for all...

    ReplyDelete
  9. Excellent overview, thanks. Can I add a couple of points?
    Firstly lets separate the technologies involved. Regarding Hot spot and snicko, both excellent aids that the umpires should have access too. Neither should have changed the decisions given against Hussey or Cowan as they only didn't demonstrate evidence of a snick, and not evidence that there was no snick. I, however, am certain that neither edged, only, and ONLY because of their their denial after the events (one of which was petulant and unpunished; it appears that batsmen want to cheat by not walking when they snick, and then petulantly complain when the situation is reversed).
    Finally, the issue of Hawkeye: an excellent TV gimmick, period. There is no proven predictive accuracy anywhere. What appears to be missing from discussion is that there is a technician at a TV screen placing a spot on a virtual pitch determining where the computer traces the ball from bowler to pitch, and then there is a second spot placed on where (and when/which frame) the ball strikes the batsman. Both of these are not placed by an umpire but by a technician and their accuracy determines the computer's track (bear in mind the computer can't determine these positions because at balls travelling at 80mph the ball travels 12 inches between frames, and hence requires a technician to best guess these positions). I find this issue of most concern having seen numerous dubious placements of ball strikes on pitch (in and off stump line) and also against pads (once again in and off line. At a recent KFC big bash match a hawkeye replay inadvertently showed both spots before the bowler released, only for the ball's path to actually pitch leg side of the spot (obvious technician error, unquestioned and uncorrected), and wide of where the ball struck the left hand opening batsman's (can't recall name) pad. The batsman was given out by the umpire (wrong on replay) and also by hawkeye, even though the unadulterated replay showed the ball pitching outside the leg stump line. Video replays, snick, hotspot should be available to the umpire, to assist in making decisions, hawkeye however is pseudoscience, and should remain in the TV domain. Apologies for the rant, it started as a comment!

    ReplyDelete
  10. Now watch the replay of Sydney 2008 and tell me how long it took Sharma to walk after nicking to the RIGHT hand of first slip? (And while you're at it, talk to us about the Ganguly appeal in the World Cup final.) Are you seriously telling us that the debacle after the Test was about someone not walking? (BTW, Clarke gloved it and probably thought he got less on it than he did - he likely thought it was a keeper deflection off a faint glove rather than a ball ballooning. He also apologised for not walking after the match. Did Sharma?)

    ReplyDelete
  11. Here's the deal, simplified right down:
    - The umpires are said to get 95% right, but most appeals are clear cut, so let's say they get 80% of the close ones
    - UDRS picks up, let's say, 80% of the remaining errors.
    - A trained 3U (and there's only about 20 of them) has the skill to decide when the technology has clearly failed, and is under no obligation.

    The only objections to UDRS have been of the form "It's not perfect / see, here's a rare error it made under unusual conditions [Hughes]" or "Here's some things it doesn't do well [faint edges]." None of this is fatal if the 3U has suitable discretion, and it remains a massive improvement on unaided field umpire rulings.

    ReplyDelete
  12. Re Anonymous Jan 14 3:38am. Garbage. Exactly what is the 95% or 80% compared to. i.e. what is the gold standard (the 100%) that your claims of 80% and 95% are against. We have no gold standard to compare with, so your attempt to quantify accuracy is no more than a subjective (and hence qualitative) assessment. What exactly measures hawk eye or virtual eye accuracy....... thats what the ICC are purporting to be doing now. Interesting that the current SA/NZ series is highlighting the issue without the distraction of the BCCI stance particularly as the maker of the technology used has now admitted that they don't back the accuracy of the predicted path in low light conditions (cloud/dusk and sometimes night lights).

    ReplyDelete
  13. hi..Im student from Informatics engineering, this article is very informative, thanks for sharing :)

    ReplyDelete