The DRS should Eliminate Howlers. And Those Alone

By Vineet Goyal

Much has been written and debated about the effectiveness of DRS and it is fair to say that the jury is still out. Recent comments from Jacques Kallis are a good indication that the majority of players are not entirely comfortable with the system. However, most agree it is quite effective in eliminating blatant errors and the game is better off with some sort of DRS in place. Eliminating howlers was in fact the primary intended purpose of the system. So why is there still opposition to the system if it serves its primary intended purpose so well.

The answer lies in not what it does but in what it does not do effectively. It is completely ineffective in marginal decisions – the current technology used in DRS especially for predicting the trajectory of the ball in an lbw decision is not accurate enough for players and administrators to feel confident. But marginal decisions! No one was worried about those anyway – at least not the players. All we wanted was to eliminate the howlers for which DRS seems to work quite well. The problem however is that once the system is in place, we cannot ask the players to use it only in cases of a howler. The players realize that the technological shortcomings of the system can be exploited and therefore, they can use the DRS strategically to get a decision in their favor. This is very discomforting from everyone’s perspective.

The objective of DRS should instead be to eliminate howlers and just that. It would be futile to formally define a howler but in principle from the perspective of the batsman, we can think of a howler as a decision where the batsman feels (in his mind) that grave injustice has been done to him in giving him out. Now, an lbw decision where the ball was perhaps just hitting the top of middle stump is not a howler (assuming there was no inside edge and impact was inside the line of stumps) – in this case, a batsman would be disappointed but in his heart knows that he was beaten and perhaps out. Clearly, we cannot have a system where by rule the players can ask for a review only in case of a howler. This should happen by design. Here is a proposal that I feel should achieve this:

1. The batting team is allowed only ONE unsuccessful review for both the innings.
2. The bowling team is allowed two unsuccessful reviews per inning as in the current system.

The asymmetry in batting and bowling reviews is because of the asymmetry in knowledge between batsman and bowler. The batsman knows whether he nicked the ball almost surely but the bowler may not. 

So lets see how this system would play out. A batsman will review only if he is absolutely sure that he is not out. Except lbw decisions, the batsman knows whether he is out or not. Anyone who has played cricket would know that you feel the vibrations from the faintest of nicks and you know it. If the batsman didn’t edge the ball and is given out, the review will certainly reverse the decision – none of the technologies including hot-spot and snicko give a false positive. So non-lbw decisions will be ruled correctly. 

It is a bit more trickier for the lbw decisions because the batsman does not know whether the ball was going to hit the stumps or even whether the impact was in line or not. It’s amazing to see the performance of umpires on the elite panel currently – even the players would testify that the umpires are extremely accurate about things like estimated path of the ball and line of impact. Of course, they are humans and making real-time decisions, so they are bound to make marginal errors of a millimeter here and there. And I think the players are ok with such errors. If Kallis is beaten to an incoming delivery that is crashing onto the middle stump and hits him a millimeter outside the off-stump, I don’t think he will feel cheated if he is given out. The howlers in lbw decisions happen (at least these days) when batsman are given out after an inside edge which is easy for batsman to detect and then review. But since batsmen have only one unsuccessful review for two innings, they will not review marginal decisions.

What about the bowling team? Can they use it for marginal decisions? Since there are two unsuccessful reviews allowed, they might like it happens in the current system. However, this should not be a problem. This system does not produce false positives, i.e., if the batsman did not edge, there will be no hot-spot. So if the batsman is indeed not out, the system would rarely reverse the on-field not-out call. However, in case of a blatant error, the decision is reversed. Therefore, the DRS essentially comes into play only in blatant errors and has no impact in marginal decisions.

Alcohol and the Jesse Ryder Assault Case

My article at The Cordon last week on the Jesse Ryder assault casesparked some strong critical responses (mainly on Twitter). These roughly amounted to the following:

1. That I had blamed Jesse Ryder for the attack, thus blaming the victim of a seemingly almost-fatal assault.
2. That I was wrong to speculate or conjecture about what had happened that night, wrong to offer an opinion on what might have happened. This criticism is related to the one that I was wrong to be offering any sort of commentary on a man who had been the victim of a beating.
3. That I am a puritan about alcohol, who thinks alcohol causes violence, that it was the sole cause of the incident that night.

First, I reject the canard that I was blaming Jesse Ryder for the attack. Nowhere in my post do I suggest that Ryder’s drinking in particular caused the fight, that he started the fight, that he was drunk, that he was to blame. Describing the incident as alcohol-related does not mean blaming Jesse Ryder. Indeed, in my post, I suggested Ryder might even have been picked on by his attackers. That is, he might have been harassed by his attackers, who knowing Ryder’s past, made offensive remarks pertaining to that. My critics should acknowledge though, that even if Ryder had been stone cold sober and not drinking—though we have been told he was—and had been attacked by a pair of drunks, this would have been an alcohol-related incident. More importantly, you do not have to be drunk to get into an alcohol-related fight; you simply need to have your tongue loosened up a bit. That is all that is needed to implicate alcohol in this incident.

I will admit the use of the word ‘everything’—in the title of my post and then again later in the post itself--was unfortunate. I meant it as a rhetorical flourish in response to the NZCPA and police statement, along the lines of ‘Whaddya mean this had nothing to do with alcohol?! This had everything to do with alcohol!’ By using ‘everything’ I made it seem as if alcohol was the sole cause of the attack, but all I meant to say was the news made it seem very likely alcohol was causally implicated in the attack on Ryder, that it was a contributing factor. (Note my use of ‘likely’; I will return to this soon enough.) But nowhere in my post did I say alcoholcauses violence; I merely pointed out a correlation between alcohol consumption and violence. This correlation is visible to most: bar brawls, fights in sports stadiums late in the afternoon (there is a reason why, beyond worries about drunk driving, beer sales cease in the late afternoon at many cricket grounds), domestic violence cases etc. A simple googling of ‘alcohol related violence’ throws up a wealth of links which note the correlation between alcohol and violence, such as this one, which links to many studies conducted by the National Institute on Alcohol Abuse and Alcoholism. Social workers, policemen, women’s shelter workers know about this correlation; I took this to be a cultural commonplace, one that warranted my claim that the Ryder assault—especially because of its location and timing--was alcohol-related, but many assumed I was saying one drink turns you into a maniac, that alcohol invariably leads to violence. But that claim was not made by me and neither was it needed to make the case that I did. (Incidentally, those suggesting I am a puritan about alcohol are wildly off the mark. I enjoy wine and beer and whiskey; I go on wine-tasting vacations; I make road-trips to visit my favorite craft breweries; like any middle-class aspirant to the good life, I try to talk knowledgeably about the single-malt whiskies I enjoy: ‘Laphroaig? Pshaw! Too peaty!’).

On a related note, though I did not explicitly and extensively indict masculinity in my original post, I did note in my invocation of the Martian anthropologist that it is the ‘males of the species’ who might brawl after drinking. It is the combination of a misguided masculinity and the inhibitions loosened by alcohol that often contribute to post-alcohol consumption violence. Both these factors make it possible for me to make the following claim: if you hear two men have been in a fight in a bar, it is reasonable to infer/speculate/conjecture/assume/hypothesize that alcohol was involved. I do not need, at this stage, to rely on the additional datum—as in this case--that one of them has a prior history of similar incidents. The mere reporting of the first is enough to warrant the drawing of that conclusion.

This brings me to the criticism directed at my ‘speculating’ or ‘conjecturing’ about what happened that fateful night. Many commentators wrote as if speculating or conjecturing about unseen, but reported on, and incompletely known, events was probably the worst sin a blogger could commit. To that, there are two responses. First, my piece was not reportage. It was an opinion piece. I was not writing as a beat reporter but as a blogger responding to a news item; this sort of opinion piece, consisting of speculative commentary, the plausibility of which can be judged by readers and responded to critically, is perfectly appropriate. I was writing to express my considered belief—my opinion--on what happened that night. People reading my opinions are free to critique the process by which I reached those conclusions or to point me to additional evidence that might affect my beliefs, and thus, perhaps, to show why conclusions were not warranted. It is no argument, however, to command an opinion writer, ‘Thou shalt not speculate!’ A blanket ban on speculative commentary by opinion writers – like, for instance, on those who speculate on what went on behind closed doors at a deliberation of Supreme Court judges before they delivered their ruling - thus ruling out the inverse of a species of commentary called ‘reading the tea leaves’, would be far too restrictive. If it’s OK for opinion writers to speculate about what might happen in the future, then why not about the past? I formed the beliefs that formed the basis of my post on the basis of the evidence I had. I have been accused of not being diligent enough in my evidence collection, but even after reading additional news reports my hypothesis about what happened that night remains the same as before. I do not think Ryder was the victim of a premeditated, conspiratorial assault; the facts still seem to point to an edgy, alcohol-infused encounter having gone wrong. It is far more plausible to infer this and indeed, to continue to not blame Ryder, than to assume that there was a conspiracy to attack Ryder, that the assault was planned.

Furthermore, I hate to break the news to those who accused me of ‘speculation’ and ‘conjecture’, who, possessed by a spirit of epistemic rectitude, have suggested that we should sit on our hands and not form any beliefs till ‘all the facts are in’: much of our daily reasoning consists precisely of speculation, conjecture, guessing, hypothesizing, and so on – and thus, offer opinions about all sorts of matters based on this. We constantly form beliefs on the basis of incomplete evidence; if we were to continue to wait until all the facts were in we would never form any beliefs. There are, of course, good variants of that activity and there are bad variants of that activity. This is why it is silly to assume Martians have landed when you see a broken window in your apartment but far more reasonable to conjecture that the neighbor’s kids have been playing pranks again. I happened to think my conjecture was a good one, which is why I wrote the post I did. Here is why: I heard a fight had taken place in a bar, that there were men involved. I reacted on the basis of prior beliefs that bars have often served as venues for brawls, that the intersection of masculinity and alcohol—as bar bouncers and visitors to Yankees games and the SCG in the old days will testify—lies quite often in violence. That was the basis of my reaction. 

Third, many accused me of poor taste in writing about a man who had been attacked and was near death. But Ryder is a public figure; what happens to him makes news; responding to that news is perfectly appropriate for bloggers and journalists. If Ryder had been a completely obscure person, living his life out of the public eye, presumably the news of his assault would not have made the news and there would have been no press conferences or newspaper articles on his health (and indeed, neither would anyone care he had ever gotten into a brawl in a bar previously). And of course, once news of the attack broke, there was plenty of commentary—about Ryder, his past, his career, his character--in any case. So apparently, it is acceptable to write about a near-dead man so long as you write the ‘appropriate’ things, that all one should do when confronted with news about a public figure is offer bromides and wallow in agnosticism, scrupulously refusing to draw any inferences from the news presented to us. When Christopher Hitchens died last year, many commentators,including myself, wrote articles where we detailed our disagreements with that rhetorical pugilist. Predictably, there were those who suggested that these writings were in poor taste, that one should not speak ill of the dead. This fastidiousness about the dead or the injured public figure is curious; it seems to find its grounding in a misunderstanding about the nature of public discourse, which is not circumscribed in the ways my interlocutors might want. (What Glenn Greenwald wrote in response to the commentary on Hitchens’ death applies to those who would critique my writing about an injured public figure as well.) 

Importantly, in Ryder’s case, it was only because of his public past indiscretions that the need was felt for the clarification offered by the NZCPA and the police (one made, we should note, without all the facts being in, for after all, they did not have Ryder’s testimony at that point in time, and indeed, as seems likely now, will never have). So, again, why the restrictions on commentary on that statement? Why is it not acceptable for an opinion writer to respond to a public statement about a public figure?

In closing let me repeat something I have already stated above: the next time I hear about a fight between two men in a bar, I will speculate, assume, conjecture, hypothesize, abduce that alcohol was involved. It is a reasonable inference to draw and unless the ground facts about masculinity and alcohol change I will continue to draw that inference in the future. It is not an unreasonable inference to make. If you think it is, please state your argument, one that does not rely on arbitrarily circumscribing public commentary or a misplaced fastidiousness about public discourse on public figures. I stated mine in my original post and have reiterated it above. My readers are, of course, welcome to simply remain agnostic and suspend belief till the facts are in. They might be surprised at how difficult genuine agnosticism actually is.