Cricket’s stat gurus14 min read . Updated: 19 Oct 2015, 01:13 PM IST
A short history of India's cricket statisticians, how they went from record-keeping to analysis, and why we love the numbers
A short history of India's cricket statisticians, how they went from record-keeping to analysis, and why we love the numbers
What gives Sudhir Vaidya springs? Here he is, bouncing around his home, clambering up on beds, reaching shelves that 77-year-old arms are not meant to reach. The objects of his pursuit: ancient items. Heavy box files, bound notebooks and registers sheathed in mosaic-printed cardboard, all filled with numbers, relics put out of business by the weightless Internet. But the heft of these thousands of pages is what lightens Vaidya’s step.
“Records of every single Test, One Day International (ODI) and Twenty20 international cricket match to date." Pat. “Scores and averages of every Ranji Trophy cricketer." Pat. “Individual records of every cricketer to have played international cricket, organized by team." Pat. Descriptions are delivered to the rhythm of books snapping shut and proud hands smacking their covers in approval.
It must take a lot of time to maintain this flood of numbers. “Lot of time? No, no. See, this takes all my time. I am retired now and spend my days and nights with the figures of the game."
Yes, but to what end? “Here, a list of all the Test match innings of over 500 runs." “Here, all the innings of less than 100."
What is the purpose of this white and blue sea?
“It is thrilling."
Pat. I am back in Vaidya’s sea, but no longer as a mere observer. I recognize how, as a cricket lover, my life has been dotted with dips in this ocean. Would I like a dip now? Sunil Gavaskar’s Ranji Trophy record? A spring, and there it is, in the register containing data on all players whose surname starts with G.
This is not an intimate raft we share, me and Vaidya, India’s oldest living cricket statistician. This is a large ship on which most of India’s cricket fans have journeyed.
Statistics, whether in the form of mere trivia or deep analysis, are an indelible part of the fan experience. They fuel our arguments: If Shane Warne is better than Muttiah Muralitharan, why does Murali have a better bowling average even if you exclude games against minnows, and why has he got more top-order batsmen out than Warne did? They cover our television screens and social media feeds during tense moments in games: Ah look, India have chased 300-plus runs 15 times in ODIs; if enough of us retweet that, perhaps we will transmute the past into the future. They gave us collective springs when a short man from Mumbai began to hack down gigantic records with a heavy bat. Most Test hundreds, most Test runs, 100 hundreds, we shivered with delight in anticipation and celebration.
While this obsession with statistics is a phenomenon that permeates many sports, cricket seems to heighten it. By its very nature, the sport churns out numbers by the bucketful. Every act on a cricket field can be quantified in numbers. Unlike, say, in football, in which it is difficult to define in figures the importance of one single pass in a complicated team move. In cricket, every ball changes a bowler’s economy rate, the innings run rate, the batsman’s strike rate.
Vaidya, though, has not created his sea for selectors or coaches. In his mind, his creation needs no purpose other than its existence itself. For believers, there need not be a utilitarian purpose to the secrets whispered by statistics. We quietly accept that no matter how much we bleat about how the figures suggest that a particular player be selected, the men making the decisions have access to the same data and more, and will choose their paths regardless. And there is certainly no conceivable practical use in comparing a modern-day hero to a past great.
There is not much that can be achieved by our cheers, tears and prayers when we are watching a game on television either. It becomes clear that the question of why we love sport statistics is linked inextricably to that eternal question: What makes us so crazy about sport in the first place?
Perhaps clues will present themselves in the stories of how the providers of numbers found their occupations. Gopala Krishna leaked into the world of sports statistics through the nib of a pen that formed perfect letters. In the 1950s, when he was in school, his excellent handwriting earned him the position of scorer for inter-school matches. He would go on to become official statistician for the Karnataka State Cricket Association and cover several games for All India Radio.
S. Rajesh, now the statistics editor at the ESPNcricinfo website, first embraced numbers as vital ammunition for sibling rivalry. In the mid-1980s, his elder brother and he would argue about whether Vivian Richards or Greg Chappell was the better batsman. “It was not easy to find details such as series averages then, so I’d go look at old Sportstar magazines and find the relevant data. I eventually began keeping my own records."
For Vaidya, a single photo caption became a window to a world he would completely immerse himself in. In 1958, when he was 20, a line across a picture of Hanif Mohammad mentioned that the great Pakistani batsman had broken Sir Donald Bradman’s record for the highest Test score. It aroused Vaidya’s curiosity. By the next year, he was spending his evenings memorizing the location of every book in the Mumbai Cricket Association’s Dr Kanga Memorial Library, springing up and down among the bookshelves whenever other library members asked for advice on which books to read.
The beginnings of lives spent with cricket’s numbers seem varied. A craze for cricket is, of course, essential, but is there a particular personality type that gravitates towards this meticulous job? “You have to be a little obsessive in nature," says Atul Kahate, who has been writing articles about cricket statistics since he was 14. The word triggers a memory. I am a pre-teen and, like all children smitten by cricket, daydreaming about scoring hundreds and taking hat-tricks for India. I am adding up the exact scores I have made in my imaginary matches and calculating my sensational average. It’s a habit that still creeps into my reveries.
Kahate’s home in Pune swirls back into focus. He too seems wistful. “Also, stats lovers are people who want to discover the unknown," he muses.
What is this unknown? Navjot Singh Sidhu’s average is higher than Tendulkar’s. A 10-year-old’s secret. But what does that tell us? That Sidhu is better than Tendulkar, vindication of my support for the swashbuckling sardar? A petty comparison. Or perhaps something more? A quest to test the limits of human performance?
“We like to compare players across eras and think about what the pinnacle of achievement could be," Rajesh explains. “That is what drives a lot of statistical analysis. People like to think about things like how much Bradman would have scored if he had played in this age of bigger bats and flatter pitches."
Anantha Narayanan, who has written 150 pieces of statistical analysis for ESPNcricinfo and is currently working on a book that will feature similar work, explains that to quantify a cricketer’s performance, one has to look at several parameters: the period he played in, the opposition he took on, the conditions he contended with, the support he had, and more. Consider this: In a list of the greatest Test match innings ever that Narayanan, 70, is preparing—an update to the Wisden 100 list he put together in 2001—Graham Gooch’s 154 against the West Indies at Headingley, in 1991, occupies the top spot. It is far from the highest score in a Test match, but due to myriad factors—the influence of the innings on the match, the significance of it to the overall team score, and more—it is, according to Narayanan’s statistical model, the best display of batsmanship ever.
Vaidya, Rajesh, Kahate, and Krishna all started off maintaining records by hand, copying scores from the daily newspapers and weekly magazines. This was between the 1950s and the 1990s, long before a click could pull up all the details of a match. Statisticians lugged their books into press boxes at stadiums and fed information to radio announcers and, later, television commentators. These proclamations were the only access to even the basic records of our heroes.
Every now and then, a book would be released with detailed numbers from the game, and we believers would rummage through them looking for key disclosures. Aha, Sidhu’s average. And here is Tendulkar’s.
In this time of limited information, cricketers themselves flocked to the statisticians to obtain proof of their deeds. Ranji Trophy players would often ask Krishna for their season averages. Kahate once blew Amol Muzumdar away with a hand-drawn batting pie chart that illustrated where he had scored his runs during the innings of 260 he played on his first-class debut. Vaidya’s sea of numbers entranced Kapil Dev—he, in fact, told Dev about his incredible feat of having missed just one Test in a 131-Test career—and Gavaskar, from whose face he once erased a sullen look with a simple statistic.
“I was in Pakistan in 1978, covering the Test series there for AIR." says Vaidya. “Gavaskar was out in the 90s in Lahore and he was very upset with the decision. At the Indian high commissioner’s party that evening, he was sulking near the hotel’s entrance. I told him, ‘Don’t worry, you’ve got out in the 90s for the first time after so many innings.’ He bounded into the party telling everyone he had created another record."
Quickly, ties were formed among India’s pioneering cricket statisticians. The late Anandji Dossa, a revered figure not just among statisticians and cricket historians but the entire cricket-loving world, played mentor to several young mavericks. His collection of figures was always made available to Krishna and Vaidya.
There was even a short-lived formal group, the Association of Cricket Statisticians and Scorers of India, formed in 1987 and defunct by the mid-1990s. Statisticians and historians from England and Australia, including the renowned Bill Frindall, began contacting their Indian counterparts and asking them to contribute to well-known international publications.
The statistics were still fairly basic. Cricket was just about shedding its image of being a gentleman’s game, and such crude elements as statistics had to be ushered in slowly. A batting or bowling average was simply a lowly supplement to a cricketer’s worth, measured principally by the grace of his strokes or the fluidity of his bowling action.
But our appetites were not to remain so easily satiated for long. We wanted more nuanced fodder. What did all these records mean? Where did our titans really stand, against their peers and in history? We needed more information, more quickly and easily. And, of course, teams and players wanted more data too. Cricket was now being played for big money, and winning was far more important than aesthetics.
By 1993, a cricket-loving South African engineer based in Oregon had created an open-source software named Dougie that assisted in scoring matches live on the Internet. Sites such as ESPNcricinfo were archiving scorecards on the Web, and individuals were writing computer programs to mine the data from them. Kahate, a software engineer himself, had created a database of scorecards on his personal computer and a program to crunch the numbers. “I was working for a firm in Europe when I created my own archive of scores and stats," he says. “But by the time I came back to India in 1999, all the information you could want was available for free on the Internet."
Painstakingly putting together these figures has clearly given Vaidya an unnerving ability to remember stray incidents from decades-old matches. “When Sunil Gavaskar was out stumped in Colombo in 1985, the radio commentator asked me if it had ever happened before (Vaidya was helping AIR cover the Test). I immediately remembered it had, against New Zealand in 1976, and sifted through my books to confirm it." A simple search on ESPNcricinfo’s famous statistics engine, Statsguru, gives me the same information within seconds.
The pieces of trivia that had required such careful curating are now being thrown up by sophisticated computer engines and displayed graphically on screens. But this does not mean statistics have lost their sheen or statisticians their purpose. As figures became easier to find and more of them became easily calculable, they began to be analysed in far greater detail than before.
Us believers, of course, were not going to allow the backrooms of teams to hoard all these exciting new numbers. We wanted them broken down and explained to us. This gave birth to what Narayanan calls the cricket analyst. “A statistician is someone who sits in the commentary box during games and digs up facts and figures when the commentators request him to. But an analyst is someone who spends time trying to explain more elaborate concepts through stats." Narayanan is adamant on the distinction between the two and places himself in the latter category.
Records and trivia still capture attention. Mohandas Menon, perhaps India’s most popular cricket statistician, has over 93,000 Twitter followers tuning in for updates, which are mostly odd records and interesting facts. But this deepening of discourse has opened up fascinating possibilities. We can quantify our heroes, and compare them, based on increasingly varied and sophisticated parameters. Have a hunch that M.S. Dhoni performs well under pressure? People such as Narayanan have created indices to measure such things. Convinced that Rahul Dravid performed better than his peers against good fast-bowling attacks? You can break down his statistics to find out.
In the age of the analyst, men such as Narayanan, Rajesh and Date write columns that use statistics as a means to highlight trends and predict results. Professors of statistics weigh in with alternative methods of calculating averages. The discoveries are now not simply unknown nuggets of information, but truths about players and the sport itself.
Information on every ball is being analysed. “Now that we have ball-by-ball data, we know what exact shot was played to a particular ball," Rajesh explains. “We can even determine things such as what a batsman’s ‘release’ shot is, based on which one he opts to play after he has faced a few balls without scoring."
And we believers, we have warmed to this new age. Narayanan’s inbox is filled with responses to his columns, many of which contain elaborate suggestions on what he should analyse next.
But amid this excitement, there are whispers of trepidation. For one, it is important to remember that complex statistical analysis is a specialized field, which will not yield to a mere enthusiasm for sport. Arunabha Sengupta, chief cricket writer and historian at CricketCountry.com and a postgraduate from the Indian Statistical Institute, says one needs an education in mathematics or statistics to attempt meaningful in-depth analysis of cricket statistics. If you don’t have one, you are likely to make false assumptions and base your analysis on faulty premises.
Sengupta finds a lot of what is trotted out as deep analysis of cricket statistics baseless. “For example, averages and centuries in won matches and lost matches, etc., are meaningless and can be shown to be so using statistical methods," he says.
We return to our initial question: What purpose do these statistics, so rigorously processed by teams and players, serve those who watch the game? To become a follower of any sport is to construct a parallel world for one’s self, one in which the accomplishments of other men can be transferred to you. In this world, there are heroes and villains, and miracles and disasters. These must all be measured. We must know as much as we can about this world, its trends and their meanings, its little nuances and opposing ideologies. For this world and all the events in it truly mean something to us.
A 10-year-old can, just for a moment, cease to exist in his own cage of insecurities. For now, all that matters is that Sidhu is the hero of his world. And he has a secret. Sidhu’s average is better than Tendulkar’s.
Statistics that defy commonly held beliefs :
Not so sunny days
Sunil Gavaskar’s record against the much-feared West Indies team of the 1970s and 1980s—he averaged 65.45 against them in 27 Tests—is the stuff of legend. But many of those Tests were against weak, spin-heavy West Indian attacks. In the only series Gavaskar played against the famous pace quartet of Michael Holding, Joel Garner, Andy Roberts and Malcolm Marshall, he averaged 30.
—From an article that helped S. Rajesh get a job at Wisden.
The chink in the wall
You’d expect a man nicknamed the wall not to get bowled very often, but Rahul Dravid was in fact out bowled more than 20% of the times he was dismissed, making him the fourth likeliest batsman to be bowled among all top-order batsmen who have been out at least 100 times.
—From an article by Anantha Narayanan on ESPNcricinfo.com.
The rough days of uncovered pitches
It is commonly believed that batting in the 1920s and 1930s was harder than now, adding even more weight to the achievements of Don Bradman. But in fact, batsmen averaged less in the 1990s than in that period of uncovered pitches. Bradman himself averaged only 20 on wickets affected by rain.
—From an article by Arunabha Sengupta on CricketCountry.com