How to bring singles back into T20

In my last post, I pointed to the fact that West Indies, despite not having the greatest fielding side or being the fastest team between the wickets, have been successful at winning T20 games. Elsewhere I've written about how, over the first eight seasons of the IPL, boundary hitting has been the decisive factor in deciding games. More crucially, scoring singles, twos and threes (non-boundary runs) has not been a factor.

Over 501 IPL games, the side that scored fewer singles has won 58% of the games, while the side that scored fewer non-boundary runs has won 56% of the time. Seventy-six per cent of winning teams have played out fewer dot balls than their opponents.

Two defining facts emerge about T20 cricket. First, power-hitting depth trumps all other factors. Second, singles are usually not contested by the fielding side. This is a counter-intuitive fact that is also revealed in ODI cricket.

These two facts generate two interesting vicious (or virtuous, depending on your point of view) cycles in T20.

Since boundaries (fours and sixes) are so overwhelmingly decisive, individual power hitters and power-hitting depth in teams become that much more vital because they can take the fielders out of the equation.

Power hitters tend not to be fleet-footed (the likes of AB de Villiers are exceptions), thereby reducing the overall quality of outfielding in a power-hitter-heavy side. Do you pick the classy strokemaker who is also a brilliant outfielder who can limit twos to ones, or do you pick the big power hitter who will get out often but can also clear the boundary with ease, and concede the odd two because he is likely to be slow in the outfield?

The quota of four overs per bowler limits the value of quality bowlers, and given the choice between a modest bowler who can hit the long ball and a really good bowler who can't, teams are likely to favour the former more often than not.

Greater hitting depth is more valuable simply because as terrific as a bowler might be, he can only influence at most a fifth of the total deliveries. Lasith Malinga, arguably the finest bowler in T20, averages about 1.5 wickets per 24 balls bowled. Consequently, bowlers are of limited value, so lesser bowlers who can hit the long ball are favoured, which in turn reduces the overall quality of the bowling, which in turn favours the big hitters.

One of the more fascinating developments in cricket in the 21st century has been the ICC's tinkering with the ODI rules regularly since 2005. ESPNcricinfo's scorecards provide player v player information for each game, which is broken down by type of outcome - dot balls, singles, fours, sixes and so on. I collected this data for 3570 T20 games and 1988 ODI games (all in the 21st century, with the exception of the 1999 World Cup). Here is the yearly distribution of non-boundary runs and boundary runs per completed innings (300 balls for ODIs, 120 balls for T20).

At the outset, it is worth keeping in mind that the only difference between the 50- and 20-over contests is the number of overs per innings. The number of fielders, the share of overs per bowler, the number of wickets available to the batting side and the rules of dismissal are all identical. In ODI cricket, between 1.1 and 1.2 non-boundary runs have been scored per boundary run. In T20 cricket, the figure has remained more or less constant at 0.75 non-boundary runs per boundary run.

In 2005, the ICC changed the Powerplay rules to add five additional overs of fielding restrictions. Until 2008, the design of the system meant that most teams left the Powerplay overs for the last ten overs of the innings in ODIs.

The changes in 2008, 2011 and 2012 were designed to counter this tendency. The record shows that they had an interesting effect. Until 2008, 82 singles were scored per 300 balls on average. From 2009 to 2015, this average increased to 91. The effect of forcing fielding sides to play more overs in the first 40 with the extra fielder (or two) in the circle has been to increase the total number of singles scored in 50 overs.

One explanation of this is that when fielding sides are forced to set fields without reference to how well (or poorly) the bat is actually faring against the ball, they tend to sit back on the very edge of the circle and give up the single 30 yards from the bat - as opposed to 60 yards from it when the batting team is doing well. In Test cricket fields are setting according to the merit of the contest.

The IPL trend of singles not mattering also holds true more broadly for T20 games. Fifty-seven per cent of winning teams in T20 games have scored fewer singles than their losing opponents. Forty-five per cent of winning teams in ODIs have scored fewer singles than their losing opponents. This trend has persisted in recent years of ODI cricket, though the number of boundary runs in ODI cricket is now nearly equal to the number of non-boundary runs.

A tentative conclusion is possible so far. Singles and non-boundary runs matter more in ODI cricket than in T20 cricket, in which the high frequency of boundaries (made possible by the large number of wickets available to the batting side over a small number of deliveries - 12 deliveries per wicket, compared to 30 in ODIs) renders non-boundary runs irrelevant.

This irrelevance of singles is also found in the way the number of singles scored correlate with the number of wickets conceded. This data shows the consequences of both singles and wickets are limited. Losing wickets does not seem to stop teams from scoring singles in T20.

In nearly eight out of ten T20 innings, three to eight wickets are lost. In these innings the number of singles scored on average is between 43 and 45. Such consistency is to be found in only about four out of ten ODI innings, where four to seven wickets are lost and the number of singles scored on average lies between 89 and 92. Getting the eighth wicket reduces the average number of singles scored to 82. Getting the ninth reduces it to 76. Teams that are bowled out in ODI innings score 63 singles per 300 balls faced.

In T20, to appreciably affect the number of singles scored, a team has to take the ninth wicket. Teams lose at least nine wickets in only 12% of T20 innings.

What of boundaries (fours and sixes)? When it comes to boundaries, taking wickets is useful in T20. Every additional wicket after the fourth reduces the number of boundary runs scored by five. Every additional wicket after the fifth reduces the number of boundary runs by six in ODIs.

T20 is a terse contest between bat and ball, in that the small grounds (made smaller by the use of boundary ropes), huge bats and line-ups packed with power hitters who are limited batsmen, reduce the range of effective skills required of batsmen and bowlers.

The number of runs scored in twos and threes in the average T20 innings can be made up in three to five boundary hits. The number of twos and threes in an ODI requires between seven and ten boundary hits. The number of runs scored in twos and threes in T20 is about 60% of the number of runs scored in sixes alone. In ODIs the equation is reversed - twos and threes account for twice the runs scored in sixes.

The data suggests that four ideas might help improve matters.

1) Reduce the number of wickets available to the batting side to six.
2) Increase over quotas to six per bowler.
3) Reduce the number of runs that accrue per boundary to three, and abolish the difference between the four and the six.
4) Eliminate the 30-yard circle.

Doing all four would be ideal, but if not, applying the last two will be a start. Batsmen can still hit the ball over the boundary into the second tier, and this will be useful since fielding sides will have the ability to place nine fielders on the boundary. By eliminating the circle, fielders will be placed to actually save singles instead of being positioned due to arbitrary restrictions that do not serve their intended purpose.

As the data from ODIs shows, the forced extra fielder(s) in the circle due to the Powerplay rule has actually increased the number the singles taken.

These modifications will increase the significance of events that occur within the boundary in T20 and make it a less boring sport. It will also bring bowlers into the game. Teams can be composed of eight batsmen and three bowlers, with one of the batsmen capable of covering the extra couple of overs. Speed, field settings, wicket-taking strategy could all have an enhanced place in T20 with these modifications.