Now that the dust has settled on the Olympics and the heartbreak that they brought for Indian athletes and sports fans alike, perhaps it is time to take our Piers Morgan bashing glasses off and look at why every Games we are put through the wringer of the same optimism –> hope –> despair –> disappointment cycle. There was a lot of discussion during the Games about how Indian athletes missed out on medals or qualification for medal rounds because of cruelly tiny margins. Dipa Karmakar, the gymnast was just 0.150 points out of a medal place in the Vault apparatus final. Mairaj Khan the skeet shooter missed out on the semifinal in a shootout. Jitu Rai and Gagan Narang had similar fates in their shooting events, and so did the women archers, losing a real close battle to Russia in the team competition quarter finals. But perhaps the most heartbreaking of all was Abhinav Bindra in the 10m Air Rifle where he finished fourth by pretty much a literal whisker. He was a genuine medal hope [after all, he was the gold medallist in Beijing] and it was news that was hard to take for him and his fans alike.
But while many explanations will be bandied about, we have to wrap our heads around one crucial idea about most Olympic sports – luck plays a significant role in who wins medals [especially second and third places, the kinds that are realistic targets for now for Indian athletes].
Let me elaborate.
I made a visual representation of how closely some Indians over the years and at Rio have missed out on either medals or medal round qualifications using the data on the relative margin [in percentage] between the nearest competitor [or qualifying mark] and what the Indian achieved. The tinier the dot, the closer the margin. In almost all cases the margin was less than a fraction of a percent, a variation so small it could have been caused by anything – a small misalignment on PT Usha’s hurdles in 1984, a strong wind in Milkha Singh’s race at Rome in 1960, or the tiniest of imperfections in the groove of the muzzle of a shooter’s gun – that is really outside the control of the athlete.
Bindra tweeted to a journalist when he mentioned that Gagan Narang had the qualification ‘in the bag’ with ten shots left that
The 5 time Olympian knew what he was talking about. In candid accounts in his superb book A Shot At History [co written with Rohit Brijnath] Bindra offers details of why shooting is so difficult – ‘Shooting is different from tennis and swimming, it is not structured towards repetition, it allows for few brilliant encores.’ There is some solid statistical reasoning behind that statement.
Later in the book, he opens up about his failure at London.
‘I was incredibly close in London but impossibly far. Of the three 9s I shot in my last 10 shots two were 9.9s. They were close to a 10 but counted only as 9. They were in fact 0.1mm away from a 10. You know how tiny .1mm is – it is less than the dot that precedes the 1.’ He compares how things unravelled in 2012 as compared to 2008, ‘In Beijing with the rifle put in a vice the grouping of 10 pellets was 5.4mm. Now it was 6mm. It’s not an excuse, it’s just a shooter’s everyday irritation.’
That everyday irritation he describes is what in stats lingo would be called regression to mean. Any ‘normal process’ [an archer shooting an arrow, a shooter at the range aiming at a target 10 meters away, a gymnast trying to stick a landing after a vault, a golfer shooting a round of golf] usually has variations which are grouped around its true average or mean. The variation happens because of two reasons – factors under the control of who is in charge of the process [in this case, the athlete] and factors that are random, i.e. not under the athlete’s control. With practice, an elite athlete can try and eliminate the first cause but there’s really nothing they can do about the second. Thus despite their best efforts, there will be times where they will fall agonizingly short.
Explaining how a ski jumper has two jumps and they tend to do well on their second jump if they had a poor jump and vice versa, Daniel Kahneman in his book ‘Thinking Fast And Slow’ writes – ‘The point to remember is that the change from the first to the second jump does not need a causal explanation’, like the ones sports commentators tend to give. He says it is ‘a mathematically inevitable consequence of the fact that luck played a role in the outcome of the first jump’.
We witnessed this with Aditi Ashok, the 18 year old golfer who shot two great rounds to be in medal contention but could not keep up – regressed to the her mean essentially – in the last two rounds and finished 41st, a commendable achievement still because she was up against the world’s best. But she did not ‘blow a chance to win a medal’ because she really punched above her weight. To truly contend she’ll be the first to tell you that she needs to get better. And given that she is barely 18, she certainly will. There is still a lot of work she needs to do on the ‘variability’ due to the factors under her control – her shot making, putting etc.
Looking at the heartbreak chart again you could say some of the misfortune that our athletes suffered [for which many an unkind word were said to them too] was that ‘mathematically inevitable consequence…of luck’ Another factor that exacerbates this is the fact that with time in short supply Olympics events tend to have very short and sharp formats [eliminations, knockouts etc are the norm].
This effectively means that while skill is important, luck becomes an important factor too. The athletes at the very top will always have enough of a margin above the rest to allow for such variations and maybe still win gold, but like I mentioned, it is a closer and more chance dependent scene lower down the table where athletes of similar calibre are bunched together in a format that does not allow for the full skill advantage to play out.
In his book Moneyball, Michael Lewis explains how in baseball playoff season where teams win or lose based on short 5 or 7 game series rather than a long 162 game season is a ‘giant crapshoot’. He cites Pete Palmer, a sabermetrician who wrote the book The Hidden Game Of Baseball, and his calculation of the role of chance. Palmer calculated that the average difference due to skill between teams is about one run per game, while the average difference due to luck is about four runs a game. The simple translation, as Lewis puts it, is that ‘Baseball science may still give a team a slight edge but that edge is overwhelmed by chance.’ He calls it the ‘sample size problem’.
In most Olympic sports too that is the case. Now the question you might logically have is why, then, do only we [India] seem to have this rotten luck? The short answer to that is that our athletes in general are not sharp or trained enough to have their average levels of performances higher than most elite ones competing in their discipline. A Bindra’s loss is truly heartbreaking because he is an exception but if we had more competing at the level he does, even in the crapshoot of short Olympic formats the dice would have rolled our way at times. Sadly, it does not.
This is something that we have to keep in mind as we watch these sports, which unlike team sports or sports whose formats are different [leagues etc] will keep throwing up such frustrations all the time. The fact that we really pay attention only once in four years or so makes the situation worse. So, how can we battle this demon of regression to the mean that seems hell bent on denying us Olympic glory and keeps feeding troll bait to the likes of Piers Morgan? A few ideas on the next, and last, post in this series – Revelations.