Substack

Sunday, June 24, 2007

Sport and statistics

In the late nineties during his stint with the South Africans, the late Bob Woolmer brought the laptop into fashion for a cricket coach. In recent years, we have moved by leaps and bounds in employing statistical techniques for improving performance of sportsmen. In an age when professional sport at the highest level is exceptionally competitive and lucrative, and only a small difference separates the top few players and teams, it is understandable that sportsmen and their coaches try to squeeze out even the very minute marginal advantage.

By analysing the strokes of tennis players, we can gather a wealth of information about their styles of play. The service of a player can be studied to discern trends about when (different stages of the match) serves from deuce court (or advantage court) hits the tramline or centre line or the mid court. Other examples include, the probability of the player changing direction on a cross court rally by going down the line, the situations a player is likely to play drop shots, the general nature of service returns of a player on different angles of the serve etc

In football, both penalty kicks and freekicks, are fertile grounds for statistical analysis. An inventory of all Zinedine Zidane spot kicks may throw up interesting trends, including the direction he generally prefers - top left or right, bottom left or right, or centre top. Similar study of Oliver Kahn's movements may give some insights into the direction in which he prefers to move while defending spot kicks. We may even be able to discern interesting trends in the ball passing movements of individual players, as well the team. Thus a player marking Diego Maradona could gather useful information from the team statistician about his movements.

In cricket, one can possibly draw interesting conclusions about a batsman's response to different types of deliveries, not just short rising deliveries. It is possible that a particular batsman will get targetted by a specific category of bowlers with a particular type of delivery. Similarly with bowlers, detailed analysis may reveal specific trends in their deliveries. A Stuart Clarke may prefer to not follow up a nasty bouncer with a yorker, but with a loose half volley inviting an indiscretion from the shaken up batsman. There are many other insights that can be drawn by studying statistics - how a batsman tries to get off the mark, how he reacts to balls outside the off stump etc

Similar statistical analysis have substantial role to pay in most team games. Table Tennis, Hockey, Baseball, Rugby, Badminton are some of the game likely to be enriched by the use of statistical techniques. The coaches will have to identify those vital parameters or variables in their sport and then analyse it for useful information.

These statistics reveal a lot about the mental make-up of players and how they respond to different situations. We are not far from the day when a statistical analyst could program a ball-by-ball strategy for an innings which the captain could execute with the help of a PDA. Of course we will never have a Plan executed to clockwork precision. (A Sachin Tendulkar or a Brian Lara can, on their day, easily destroy any plan. But most journeymen cricketers can be constrained) These methods are only aimed at enhancing performance outcomes by optimizing our repsonses to different situations, players and teams.

The purists will be shocked by these developments and there will be heated debates about whether it is for the long term good of the game. But sport is after all entertainment in its different dimensions, and nothing else. To the extent that these techniques have made modern day sport more competitive than ever before, we should have no dearth of entertainment. An example is Chess, which in contrast to a decade back is today far more technically analysed with sophisticated software, but none the less popular.

This provides a sliver of opportunity for our perpetually losing sportsmen. Get some of our accomplished software talent into analysing sport!

1 comment:

Hari(何瑞理) said...

Pardon this very late response to your post -- I was researching for some data and in the process came across your website.

I would like to know if there's any research done on the correlation between state level sporting performance and the availability of sporting infrastructure. For instance, Kerala, where I'm from, has always done reasonably well in national athletics. Yet it has only two world class athletic tracks (one of which is so broken it can't be used). Also, is there a national plan to build more sporting infrastructure in the country? Which state currently fares the best in this regard? I searched online quite a bit, but could not find anything.

I have lived abroad for quite many years and every time I see a high school equipped with a good gym and a running track, my heart goes out to those hundreds of kids who train barefoot in the dry dusty clay surfaces of India.

What is surprising is that with so much private equity in the country, none of the big business houses seem to have taken any active interest in this area. Taken purely in commercial sense, this seems like too good an opportunity for an investor with deep pockets to pass up.

It's really sad that people like Anju Bobby George, who once was placed 6th in the world, has to resort to a career with the customs department rather than devoting the rest of their life transferring their hard earned knowhow to promising youngsters while earning a good living while doing so.