DRS through a Microscope. Eagle Eye or Hawk Eye, but not the Bull's Eye!!

Is DRS a Decision Review System or Dead Revived System? Looked like it was laid to rest after India's tour to the Old Blighty, when ICC declared it no longer mandatory. However, no sooner did the Indian team arrive down under, the issue has been raised thanks mainly to the Eagle Eye technology owner or supporter, Channel Nine, who seem to be convinced that their word on everything in world cricket has to be final. Sky TV commentary team in England did the same and gleefully, trumpeted poetic justice, every time a decision went against Indians...till VVS Laxman was reprieved by the so-called infallible "Hot spot" technology (use of Infrared camera to detect contact between bat and ball). Michael Vaughn's infamous tweet about Laxman's cheating by applying Vaseline on the edges of the bat to "hide" the hot spot, proved at least one thing, irrevocably; EVERY ONE is still suspicious about the capability of ANY technology to be 100% correct. Something that Indian captain Mahendra Singh Dhoni has been saying for ages. That shut up the Sky TV channel commentary team for sure. Poms and Aussies have this endearing quality of coming around to what is obvious only after they have suffered!!!  

However, looks like the inventors of the technology and investors of big bucks in them, are not going to die so easily. Especially when you have the best possible medium of TV and assured contract to beam free-to-air live cricket to the masses, you want to use that to push a technology even if it is still in its infancy. That is, what DRS actually is. Not surprising then, why after the first day of the Boxing day test in Melbourne, Crime master Gogo was seen asking all and sundry, "Yeh DRS DRS Kya Hai, Yeh DRS DRS Kya Hai?"

I want to show you DRS through my microscope. Before we do that, however, understand this; DRS is not merely applying all available technology to arrive at correct decision most of the times, as it is touted to be. Let me clarify. DRS or Decision Review System is a "PROCESS" set up by ICC to give SOME opportunity to players to question on-field umpire's decision. IT IS NOT AN ATTEMPT AT GETTING ALL DECISIONS 100% CORRECT!!! This is where a cricket fan is being taken for a ride. 

The reason for my above argument? In every innings, both teams are given ONLY TWO opportunities to question umpire's decision. Once you loose both opportunities, you have no right to question Umpire's decision. In other words, after you have lost both of your appeals, an umpire is ALLOWED to make mistakes!!! Let us presume this system was used in the infamous Sydney test match of 2008, of monkey-gate fame. Even before Andrew Symonds was given not out to a blatant thick edge of Ishant Sharma, Ricky Ponting was reprieved by Mark Benson, off Saurav Ganguly; a possible faint edge. Let us say that Indians reviewed and had lost that appeal since hot spot is not necessarily a full-proof technology. Subsequent to that, on that first day itself, Andrew Symonds was reprieved twice on stumping appeals, once by the third umpire who had the benefit of the technology of a slow-mo replay, even when it was obvious to a blind that Symonds was out. Indeed, even the Channel Nine commentators had given the verdict as out. This means Indians would have lost both of their appeals on the first day itself, possibly even before tea!! What good would this so-called high tech DRS would have done for the good of cricket in this case? The results could still have been the same; jaundiced decision making from one-eyed and motivated umpires.

There is also another example of MS Dhoni given out in spite of the bowler having over-stepped in West Indies this year, simply because the technician brought up WRONG REPLAY, or better put, replay of legal delivery stride from another delivery!!! All that talk about technology being savior of the world cricket makes no sense when the control of decision making goes into the hands of a total outsider who may be motivated by commercial interest of the technology provider.

Last week I heard Channel Nine commentator Tony Greig lamenting about the lack of use of the Eagle Eye that they use to judge the umpiring performance. He went on and on like a broken record about how Indians through their power of money are in fact acting against the interest of the game!! He also quoted other sports people from USA who, according to him, were astonished that Cricket refuses to use technology and has still remained an18th century sport. Ah, the beauty and benefit of having the platform to dish out drivel!! The only choice poor audience has is to shut off the TV, isn't it?? There were references made to another ball tracking system called Hawk Eye, which according to the geniuses at Channel Nine is a very inferior technology, merely due to the number of frames captured per second, and perhaps the reason why Indian cricketers hate DRS.

I brought out this issue in one of my previous blogs, and received a very interesting comment from one reader who claims to be the inventor of Eagle Eye tracking program; one that is touted as Bee's knees by Channel Nine and the whole of Australian fourth estate. Paddy's comments are  presented below, verbatim:

_____________________________________________________________________________

-->"I wrote the Eagle Eye tracking system. It works like this... All the measured points over all frames triangulate to 3D ball positions of the flight. Given there is a factor of noise, pixel error, measurement error etc, these form a probability distribution. The system then attempts to solve a multivariate equation of flight that most closely matches this distribution. This involves deriving obvious things like initial position and velocity, flight characteristics such as lift and drag, and more importantly fluid dynamics tailored to the flight of cricket balls. There are a few tricks in here that compensate for numerous things that cannot be measured. The result is a large equation that can be used to regenerate the 'actual' trajectory of the ball and extend it past the point where observations cease (ie impact on the pad). Having more frames helps this process. As you pointed out, the physics after the bounce is completely separate from that leading up to the bounce. The more data that can be collected post-bounce, the better the chance of calculating the more obscure components of flight. Operating at 230fps means every frame captures about 10cm of travel. So most of the time, enough data is captured for an informed prediction. When there is not enough data, there is a roll-off built into the system whereby factors such as swing, lift, drag and even bounce deviation cannot be sufficiently derived. Attempting to do so generally leads to solutions that look wrong. The best answer in extreme cases is found by dropping the higher-order components of drag, ignoring lift, assuming the ball does not deviate after bounce, and presume no post-bounce swing is in effect. The system is not bullet-proof, and I disagree with the assumption that "Eagle Eye is twice as good as Hawk-Eye". The two systems are totally different and each have their strengths and weaknesses. However, it is a fact that the more frames you capture, the better your chances of predicting in extreme situations. I think it's important to remember that the UDRS was supposed to prevent gross misjudgments. There are limits, and they should be scientifically explored. The ICC 'testing' done on either system is a joke."
_______________________________________________________________________ 

There you go....I have no way to confirm that the writer is the originator of the Eagle Eye program or that he is not. However, the comments very well describe the logic behind defining algorithms for tracking the flight of the ball. It is obvious from the above comments that there are sufficient doubts about the validity of either tracking system; and most importantly, none of the system is "scientifically explored" sufficiently. Exactly the point I made in my blog. Also, interestingly, the originator accepts that the algorithms can not predict accurately, complex factors such as "swing, lift, drag and even bounce". This means, prediction of the flight of the ball post impact (after it has been intercepted by the batsman) is very difficult. Perhaps that is why we saw the drama of DRS not applied during the World Cup, when the distance between the pad and stumps was more than 2.5 meters. Indeed, increased data in the form of more picture frames per second gives a better opportunity to predict the flight of the ball post impact, and to that effect Eagle Eye may be better than other technology. However, even high frames per second will not be able to predict changes in bounce, pace and turn off the wicket, that eventuate as the pitch changes its nature during the course of the test match. Forget about the difference in every ball that a bowler is able to employ, e.g. cocked wrist versus loose wrist while delivering a ball, that can generate different pace off the wicket. So when one can not be sure about the validity of this tracking system to predict the flight of the ball from day one to day five on the same wicket, how can one be confident that it can perform with acceptable accuracy, across all types of wickets; from a low slow subcontinent to a spongy English to a Perth or a Centurion bouncing one. One thing is certain: the originator of the Eagle Eye is saying that ICC has not done sufficient testing on either tracking system!!! Does it not then, sound a bit rich, Channel Nine's constant tirade against Mahendra Sigh Dhoni, or his team or indeed BCCI in not accepting DRS? Why this uni-directional tirade using questionable evidence, called BCCI bashing? Ihas now become the favorite past time of some Cricket playing countries ......Tony Greig, take off that mask of being a savior of world cricket please. You are nothing but a mouth-piece of a commercial venture, looking to pass off its yet incompetent technology. Cash for comments, of course, what else??

Let us not forget that DRS does not consist of technological solution to get only LBWs correct.It also includes other technologies like "Hot Spot" to detect a nick; something that looked to be the only infallible component of the system, till that fateful Laxman-Vaughn episode. Now that technology looks as much dicey as the tracking system. Perhaps the only technology that is close to being 100% correct is the pitch map; to confirm if the ball has pitched within the statutory area as per the LBW law. I full heartedly agree that this technology  must be used to provide on-field umpire the required support. As for other decisions like stumping and run outs, TV slow-mo replays have been accepted long back, even though they have produced wrong decisions, for a variety of reasons. Andrew Symonds' stumping, is a cases in point.

http://www.youtube.com/watch?v=92zqspN3W2c&feature=related

So it is obvious that there is sufficient doubt about the capability of the tracker system to predict the path after impact. Let us leave that aside for a moment and accept that even if not 100% certain, the technology will give a prediction of the flight of the ball, somewhere close to an umpire's eye and brain. I am ok, if the technology comes somewhere close to human capability, since it has one thing which is better than a human mind. Technology can not be "pressurized", like an umpire can be, under extreme conditions. So one can accept the fallibility of the system to the level of a human mind; with a caveat..."as long as ICC officials manage the technology". This is to prevent commercial interests from influencing the decision making. Or tomorrow's headlines will read something like.."TV technicians being investigated for Match fixing"..

Apart from the uncertainty of technology, my grouse against the DRS system, is the process of giving right to the players to challenge umpire's decision. That is eroding the very basic tenet of this game; a player shall accept umpire's verdict. Yes, on-field umpires can do with the support from technology, no doubt about that. But instead of players being given TWO appeals to get their decision right (!!), I would like to see that process completely in the hands of on-field AND third umpires. Both should work in unison. The TV umpire should have a say in every contentious decision and he should have a right to communicate or indeed over-turn on-field umpire's decision if he sees evidence to do so. In this system on-field umpire's role does become more of a constant partnership with the third umpire. There is a potential for slowing down of the game, but by now it seems every one has accepted that extension of the game time is not important if right decision could be made. Don't know though, how this will allow ICC to maintain minimum 90 overs a day requirement. Let the players be players rather than trying to be umpires also..

As the system has "evolved", there is a slight change to the name. UDRS has become now DRS. Looks like ICC has decided to take the U (Umpire) out of the name. Yes, literally that is what the system is meant to do. While the supporters of this so-called bee's knees technology claim that it is to support umpires, in actual effect, its practical purpose is to take an umpire out of the decision making.....slowly slowly, completely; well, if not today, tomorrow.