Why doesn't cricket trust the DRS?

The DRS system was bought in because cricket followers got sick of the famous "home town decisions" by umpires. The problem now is that the umpires are still controlling the decision. A "not out" decision when reviewed remains not out yet if the Umpire had said "out" then the same ball would remain out when reviewed by the batting side. Bmac reviewed decisions due to the match situation. His team needed decisions in their favour. Smith did the same thing when it was his team that needed something to change (his own LBW ) so I believe the best way to solve this is to make each decision either wrong or right, not dependant on the decision already made.

A $2 coin is not worth $2, it's about perception and all parties accepting the "reality" that it's worth $2. If all parties accept the reality that if the tracking shows the ball hitting then it's out, then the perception will become reality and we can all stop worrying about it.

I thought it was pretty funny, actually.

I guess those in charge of the technology believe that is genuinely the required margin for error, which doesn't fill me with confidence.

I have seen margins like that mentioned in relation to tennis, which is a lot less complicated than estimating the path of a cricket ball after it hits the pads. If that is what they claim for cricket, then I'm surprised at their precision. It appears that the ICC may be a little sceptical as well, given the current arrangement for umpire's call.

Company that sells hawk eye claims its plus or minus 1mm from 7 meters away.

The current system is good. Players and teams are slowly getting better and resisting the temptation to review everything and really only reviewing the calls that they are more sure of. McCullum has made a few mistakes with it this series and Smith is certainly getting better with it. They are saving it more often for the case when they are closer to 100% sure of the outcome or when it's worth taking the risk. E.g a key batsmen like Williamson or Smith or big moments when the game is on the line. Guaranteed if there were any really close calls in the last hour of the game last night either side would review. Or last wicket of the innings etc. The system for the most part is good.

I don’t know the margins for error either, but I have always assumed they are relatively large, given the latitude implied in “umpires call”. You raise a good point in your second paragraph. I take it you are saying, for example, if a batsman is given out LBW and ball tracker shows the ball missing the leg stump by a hairs breadth, then why is the third umpire’s decision not “umpires call”. To be consistent, I think it should be. I have always assumed that if a ball is shown to be grazing a stump, you could assign a reasonable probability that it would have hit the stump. You suggest that if the margins for error are this large, then why not scrap it. I think this is the basis of the Indian Cricket Board’s position. Wikipedia refers to a peer reviewed article on this topic in 2008, in which “the authors ….noted that it was probably fallible to some extent, and that its failure to depict a margin of error gave a spurious depiction of events. The authors also argued that the probable limits to its accuracy were not acknowledged by players, officials, commentators or spectators. They hypothesised that Hawk-Eye may struggle with predicting the trajectory of a cricket ball after bouncing: the time between a ball bouncing and striking the batsman may be too short to generate the three frames (at least) needed to plot a curve accurately”. Certainly, the Channel 9 team mislead their viewers in terms of the accuracy of the technology. To be fair, I doubt that they understand the concept of margin for error.

Just because something is not used for it's correct purpose, it doesn't mean we should change it's purpose to suit the faulty interpretation of it. As everyone's getting used to the DRS, the cleverer captains are gradually working out how to use it effectively - by returning to the original purpose. Smithy, for example, is turning down a lot of decisions the Aussie team would have wasted their reviews on even a year ago. To be honest, I think McCullum is still at where Clarke was two years ago - anything that looks vaguely promising, he thinks about not reviewing for a second, then says, "Nah, what the hell" and goes for it. What you need to be doing with an LBW, for example, is asking yourself, "Is it going to be Umpire's Call?" It's all very well to say, "I reckon it must be pretty close to leg stump," but it can't just be clipping, it's got to be smashing into it. You've got to have incontrovertible evidence that it was a bad call before you can overturn anything, and if it was only clipping leg stump, or grazing the top of the bails, really the decision was fair enough.

Well, I reckon I've seen times where the Hawkeye has shown the innacuracy the rules refer to. For instance, I've definitely seen it happen once that the ball has looked from slo-mo replays to hit the top of the knee roll and then roll slightly up the pad before bouncing off entirely, but Hawkeye has marked the point of impact at the place it rolled up to, a few centimetres higher than the place it hit initially, and based it's calculation on that. I've also seen it swing or turn more in the predicted path than seemed right compared with the ball's previous trajectory. I could be wrong, of course, but people with far more expertise than me have decided the technology is not infallible - and let's face it, is technology ever infallible? - and if that's the case then allowing some sort of margin for error is fair enough. Besides, when a decision comes back Umpire's Call because less than half of the ball is hitting the stumps, that shows that it was a 50/50 decision on the field in the first place - and therefore not a howler, which is the actual purpose of the DRS.

But thats the point we havnt lost that point, it has never been used in that way. It has always been used for borderline calls or just because it gets refreshed after 81 overs. If we were serious about using DRS for only howlers we would only give each team 1 review per game or take it out of their hands completely and give it all the third umpire. Even 1 review a game is alot if we are only going to use it for howlers.

Agree. DRS is there to allow teams to review the howlers, it is not there to review every decision. I think people have lost this point along the way.

No, I don't know the margin for error, but Andy's call above about almost half the ball being shown to hit the stumps isn't even what they are doing at the moment, it's the centre of the ball and the centre of the stump. That means that you have balls crashing straight into the middle of the stump where the graphic shows part of the ball actually on the inside of the stump (because the ball is wider than the stump) and yet it's given as umpires call because the centre of the ball is 1mm outside the centre of the stump. I don't know the official specifics of ball-tracker, but if that amount of margin for error is required then we probably should just skip using it at all because it's too inaccurate. Alternately, if the ball slamming into the stumps falls within the margin for error of being not out, why isn't one that is shown to be missing the stumps by 1mm also in the margin for error that means umpires call? If they are saying that a ball that is shown by ball tracker to be hitting almost middle of the stump could potentially be missing because of the error margin, then surely a ball that is shown to be missing by 2mm could potentially have actually been hitting and should also be within the margin for error and be umpires call.

This article and line of argument makes my head hurt. So many contradictions. If you go down this path there’s no need to even have umpires in international cricket anymore. If all decisions are going upstairs to be subject to DRS you don’t even need someone standing out in the middle who is an expert on making decisions themselves. Also, this statement is totally flawed. “…another option which could be explored could be the removal of player input into the DRS altogether. Perhaps, the umpire could be in charge of when a review is necessary and refer the decision directly to Hawkeye. This would, for all intents and purposes, remove all forms of human error.” Except when an umpire doesn’t request a review and the ball turns out to be hitting the stumps. Or the batsman has nicked it behind. Or whatever. If you have every decision subject to technology you’re going to wind up with a horrible stop start affair where every decision or non-decision is pored over endlessly, sometimes multiple times in a single over. Not exactly conducive to speeding up over rates. I get the distinct impression most people demanding perfection from technology in cricket haven’t played much of it, and certainly haven’t done any officiating.

So then why have it? Its not to eliminate howlers, its never been used that way and we still have mistakes or unlucky calls even with drs in place. Either trust it or dont have it. With a little more luck this series could have been anyones, McCullum out on 80 given because he wasted reviews and both Warner and Kwarja out before they got to 100 or 50 at the WACA.

If all previous decisions over the last fifty years before DRS could theoretically be reenacted and reviewed, it might well prove that up to 50 % of decisions were incorrect. Test matches and the series, including of course The Ashes, were decided by well meaning best guesses. Consider, close LBW's, run outs, and caught behinds. It' s just not cricket !.

Because it is not foolproof

Chris – what is the margin for error with ball tracker? I am assuming that you know this given you believe the umpires call margin is way too big. I have always assumed that the ICC issued guidelines on using “umpires call” based on their knowledge of the actual margin for error for ball tracker. I believe that a lot of dissatisfaction with DRS stems for the fact that the graphical presentation on TV appears very precise. People seem to believe that if DRS shows that the ball was clipping the top of the bail, then the ball would have clipped the top of the bail. Of course DRS cannot be this precise. DRS is essentially a graph in three dimensional space of the estimated path of the ball, and whenever a researcher produces estimates in graphical form, they should also produce error bars. In other words, the DRS is actually estimating the path of the ball, plus or minus, say, 2cm up or down, right or left, with, say, 95% confidence. If the TV illustrated the range of possible paths of the ball, within whatever confidence interval, then viewers would understand that DRS is not sufficiently accurate to overturn close decisions. But getting back to my question, do you know the margin for error for DRS?

DRS was originally there for the howler, not the maybe. Every LBW shout would be reviewed if just clipping was all that was required to get a wicket