In a recent interview, English cricketer, Graeme Swann, expressed a preference for doing away with 50-overs-a-side one-day international (ODI) matches.
Last week, it was Rahul Dravid's turn to say something similar. In his Bradman Oration lecture at Canberra, India's technically most-accomplished batsman referred to the empty stands at the recently-concluded ODI home series against England and West Indies. Attributing it to too much of context-less cricket and spectator fatigue, Dravid sought a reduction in the number of ‘meaningless' ODI matches, while claiming that the game itself is at a crossroads.
Indeed, cricket has undergone dramatic changes in the last fifty years. It's no longer the gentleman's game, played in the lush green countryside of England in five days; cricket today is quite a different cup of tea. It first happened with the 50-over version in the early 1970s. In 2003, an even-shorter format, consisting of 20-over innings and dubbed T20, was introduced. Its most popular avatar is the Indian Premier League (IPL), which combines the excitement of limited-overs cricket — involving domestic as well as international players — and celebrity glamour into a single evening of entertainment.
What do we already know about the different formats? Well, most fans probably enjoy the one-evening fanfare with lots of action and a result at the end of the day. Players, too, possibly find the shorter version less stressful, and so can pack more into their schedules. On top of it, IPL is more lucrative as well.
So, is all this enough to say that, perhaps, ODI is on its way out, if not already?
THE DISMAL SCIENCE VIEW
In his rather erudite address, Dravid stated that he didn't want to be a doomsday prophet or an economist. We, however, happen to be economists (albeit in the US, where cricket is only slightly more popular than baseball in India). Our attempt here is to provide, hopefully, a different perspective on the issue through some number crunching.
Our statistical analysis uses data from both ODI and T20 matches for 2008 and 2009. The data consists of innings-specific aggregate information on a number of variables for each team. For T20, we also separately take into account 20-over International (T20I) matches and IPL games. Our dataset covers 276 ODIs played by 18 teams during 2008 and 2009. This period also includes 118 IPL games involving 18 teams and a total of 77 T20I matches between 16 teams in 2008 and 12 teams in 2009.
We use a ‘production function' approach that is common in the analysis of cricket by economists and statisticians. This approach considers the outcome of the game to be a function of batting, bowling, fielding and other relevant ‘inputs' and, thus, enables us to determine the ideal winning strategy. Since it is widely believed that batting is the driving force in the shorter formats of the game, we construct three different batting measures from our aggregate data — the average number of runs a team scores for each wicket it loses; the average number of runs scored by a team for each over bowled; and the average number of boundaries (fours and sixes) struck by a team in a game.
Measures for bowling are constructed from a similar perspective (though some would claim there can be no ideal bowling strategy really in T20 and batting is all that decides outcomes!). Here, we use the following — the number of opposition runs scored for each wicket taken; the number of opposition runs scored per over bowled; and the number of overs bowled per wicket taken. We also include some other relevant variables — namely, fielding (the sum of dismissals from catches, run-outs and stumpings), home team advantage, and toss-winning benefit.
Next, assuming that the above ‘inputs' determine the outcome at least at the aggregate level, we use econometric techniques — a conditional logistical regression — to rank the different measures of batting and bowling in terms of their impact on winning. A logistical regression predicts the probability of occurrence of an event (in this case, winning) by fitting data (the relevant inputs) to what is called a logistic function. The conditional part takes the uniqueness of each game into account. Our regressions, further, incorporate the performance of the rival team. After all, in any competitive sport, performance that doesn't account for how the other contestant played makes little sense.
THE FINDINGS
We find that runs per over is the ideal batting strategy for both ODI as well as T20 games, and the number of opposition runs scored per over is the ideal bowling strategy for both formats. While some may find this unsurprising, from our perspective, it is quite unexpected. If the same set of strategies determines outcomes in both formats, it turns out that a difference of 180 deliveries doesn't really matter in how to play the game!
The variable ‘fielding' plays an important role in improving the odds for winning in both formats, though it is conditional upon a team adopting the optimal strategy combination. Otherwise, fielding doesn't seem to make any difference. That is, catches win matches only when you score plenty of runs per over and restrict the other team's batsmen to fewer runs per over. The odds for winning are also unaffected by other inputs such as winning the toss and home team advantage.
For those truly interested, the detailed analysis can be found at http://bus.lsu.edu/McMillin/Working_Papers/ pap11_14.pdf.
WHAT THE FUTURE HOLDS
In the current context, where both the 50-over and 20-over formats co-exist and in the light of the growing popularity of T20 cricket, we can now conjecture about some future directions of the game. T20 is simply a faster-paced game, making for more excitement. When it comes to IPL, the presence of Bollywood stars and entertainment are an added draw.
Given the popularity and significant monetary benefits that come with the game's shortest format, as pointed out earlier, players are more likely to shift allegiance to this format. Such a transition becomes all the more feasible, since our analysis reveals that at the team-level, there is no unique skill-set required to excel in the T20 format. The bottom-line is that the additional 30 overs do not change a team's winning strategy!
Our speculations are supported by a recent survey conducted on 45 overseas players by the Federation of International Cricketers' Association, which noted that 40 per cent indicated a preference to play in the IPL even over their country. Moreover, 32 per cent of the respondents reported a willingness to retire prematurely, to keep playing unconditionally in these lucrative leagues.
At the same time, there will always be purists and lovers of the fine game of cricket in its pristine form. It means test cricket will remain. As Dravid speculated, it is the ODIs that face a threat. His solution of creating specialised tournaments can probably help get through this rough patch. Also, given that the majority of the ICC's revenues currently derive from ODIs, its immediate demise is certainly unlikely. But the fans have spoken and there could be some truth in it.
(Colin Cannonier, a former firstclass cricketer, currently teaches Economics at Belmont University, while Bibhudutta Panda and Sudipta Sarangi do the same at Wabash College and Louisiana State University, respectively.)