The Logic Of T20

Sporting contests have an internal logic. The value of such a logic is that it makes it possible for us to identify good performances and bad ones - good plays and bad plays - plays that will produce good results most of the time, as opposed to only occasionally. Thus, for example, in a Test Match, a fast bowler who gets a wicket of a specialist batsman caught at point off a full blooded square cut, having bowled a short and wide delivery, is a very different form of dismissal than a fast bowler who traps a specialist batsman on the hook shot, or in the slips after drawing him forward trying to defend his off stump. The reason for thinking that the former (square cut) was a fluke, but that the latter two were qualitatively better is that it is far less likely for the former delivery to result in a wicket than the latter two types of deliveries. We often conclude that the fast bowler got a wicket off a long hop in the first case. There is a fairly good concept of a good delivery and a bad delivery in Test Cricket. Bowling to one's field is considered good, while straying on to leg stump with 3 slips, two gullies, cover and mid-off is just that - straying. These judgments are based on an expectation of how a batsman will play, and what a batsman is likely to find easier or harder. Hence, in Test Cricket, we also have the idea of the rare batsman who can hit the good ball for runs.



What of T20? What type of contest between bat and ball is it? I have previously argued that T20 is a separate sport and discussing it through the lens of Test Cricket, as is common today among professional cricket commentators and journalist is fruitless and substantively wrong. Having said that, does the T20 form, with its severely truncated and skewed contest between bat and ball (in favor of the bat), follow any logic? Are there any patterns of play which make it possible say that a particular type of delivery is better than another? So far, all we have is a truism about the length ball, and bowling in the "block-hole" - the former is easy to line up, the latter is nearly impossible to hit if bowled perfectly.

Lets look at some numbers. I consider the 1st innings of all 131 T20 games played by the top 8 Test playing nations among themselves from January 1, 2008 till date (excluding the most of the T20 World Cup). Chart A shows the share of the overs in which a team lost 0, 1 or 2 wickets, by over in the innings. For example in the 11th over of a T20 1st innings, teams went without losing a wicket just under 70% of the time, they lost a wicket just over  25% of the time, and they lost they lost 2 wickets just under 5% of the time. For the 20th over of a T20 1st innings, teams are as likely to lose a wicket as they are to not lose a single one, while in 1 out 5 innings (about 20% in the chart below), teams lose 2 wickets.

CHART A: Percentage of Overs in which teams lost 0, 1 or 2 wickets from Over 1 to Over 20 of innings

CHART B: Run Rates achieved by teams in overs where they lost 0, 1 or 2 wickets, for over 1 -  20 of innings

The chart suggests that losing a wicket in an over costs a team on average somewhere between a 1-3 runs. It also suggests that the damage caused by losing a wicket is far greater in the first 5 overs of the innings (i.e in the Powerplay) than it is later. Between over 7 and 20 the cost of losing a single wicket, on average, is about 1.5 runs, while during the batting powerplay it is closer to 3 runs.

Of the 2581 total overs in the 131 first innings of these games, 90% were faced by teams that were 5 wickets down or less, while about 70% of the overs were faced by teams that were 3 wickets down or less.

In an earlier post, I posted a chart showing the run rate for a given over in the T20 first innings depending on the number of wickets lost at the start of the over. These numbers suggest the following:

1. Consolidation in any realistic sense is largely a myth. Overs spent consolidating should still accrue 5-6 runs on average. Anything less is a disaster.
2. Batsmen keep attacking no matter what. They are more likely to run out of resources trying to attack than they are to throttle back and play out a few overs. Fielding sides know this and fields rarely change in a sustained way even after wickets fall.
3. Bowlers can basically bowl two types of balls - balls that are easy to hit, and balls that are harder to hit.

The logic of T20, such as it is, boils down to a few specific things in my view:
1. The ability of the fielding side to force the batsman to take risks in order to get boundaries. This means finding bowlers who have the control and the discipline to not concede "free" hits - i.e. run scoring opportunities where the cost of a miscue is minimal (such as a short and wide delivery outside off stump - where the batsman's margin of error is the length of the blade, as opposed to a vertical bat shot where the margin of error would be the width of the blade. This is not the only reason why square cuts are low risk shots while drives, especially on the rise are far higher risk).
2. The ability of batsmen to find ways to mitigate those risks. This amounts, basically, to finding new areas of the field in which to score - finding ways to hit the ball in these areas, without getting into trouble if one misses.

There are, partly due to the rules and partly due to conventions borrowed from Test Cricket, far fewer fielders defending runs behind the wicket than in front of the wicket. Conventional fielding positions behind the wicket are overwhelmingly attacking ones (slips, gullies, leg slips, leg gullies etc). The "scoop", the paddle and the reverse sweep (or even the reverse paddle) - three innovations that have become commonplace in T20, all aim, not surprisingly, at hitting the ball behind the wicket. The risk free way to play these is by getting outside the line of the stumps, so missing the ball doesn't result in LBWs. The way to force the scoop or the paddle or the reverse sweep/paddle to be high risk shots, is to bowl on the stumps (i.e. make sure that you bowl wicket to wicket).

Bad balls in Test Cricket, are still bad balls in T20, unless, the field is specifically set for this type of bowling. For example, a left armer bowling over the wicket to the right hander, but with a long on, a deep mid-wicket, a deep square leg, a deep fine leg, and a short mid-wicket, would be bowling to his field if he bowled a leg stump half volley. Now, this would make the batsman back away to leg if he wanted to hit the vacant off side boundary, and leave a very small margin of error for the bowler if he strayed just a little outside leg. In the first of those cases, the batsman would have to expose his stumps, and bowling straight will always give the bowler a chance.

This is basically what bowling in T20 boils down to. A good ball in T20 is one which gives the fielding side the chance to do some damage should the chance that the batsman is taking not come off. A bad ball in T20 is one which allows a batsman to hit a four or a six with no real cost for miscues or bad execution other than a dot ball.

Now, this is not as simple as it sounds. Neither is it as simplistic as it sounds. A consistent string of good balls has an effect that is greater than the sum of the effect of each of those good balls. Likewise for a string of bad balls. It also depends on a batsman's abilities and attritbutes. Some batsmen are stronger on one side of the wicket, or of one or the other foot (i.e. with shots played with their weight predominantly on the back foot, or with shots played with their weight predominantly on the front foot), some are better at reading spinners from the hand, still others are exceptionally quick at adjusting to spin off the wicket even if they don't read it from the hand. Still others are very comfortable playing horizontal bat shots against fast bowling. For each of these types of players, the risk that is associated with playing certain lines and length is different, on different types of pitches. Someone like Suresh Raina for example, is lethal against fast bowling on flat wickets, but find a wicket where there is some nip off the wicket, and the same strokes prove to be Raina's undoing. A "good" line and length for batsman depends on how tall the batsman is, how he prefers to play - some batsmen get forward by default unless they absolutely can't, others don't commit quite so readily. This is also true of Virender Sehwag as we saw so devastatingly in England and Australia - that Sehwag made runs at flat, hot Adelaide is a case in point.

I predict that teams will become wise to this, and that ball tracking will be used widely to determine lengths and lines for particular batsmen. What we are seeing right now, is games being determined mainly by which teams delivers the fewer "bad balls" (as per the definition developed above), given that most teams have more or less equal batting ability in top level international T20 games (such as the Super Eights of the T20 World Cup).

Like baseball, in which chance plays a massive role, in which the best teams lose about 33% of the games, and the worst teams win about 33% of the time, T20 is amenable to this kind of analysis. Lines, Lengths, records of choices batsmen make with respect to footwork, bowling speed etc. will come into play.

But for this, T20 will have to become a more widely played form of the game. I am increasingly convinced that as a form of limited overs cricket, the 20 overs a side form is probably better (from a business/entertainment point of view), than 50 over cricket, while the quality of cricket in the 50 over game is probably better, at least in some parts of the game. But if the 50 over game is going to become T20-lite, with Powerplays and restrictions, it is merely taking up time and space which could be better dedicated to Tests, or T20s.

Cricket can have two avatars. Three is a crowd. T20 must find its identity as a separate sport, just like baseball did in the 19th century. In order to find its identity, it is essential that its logic - its disciplinary standards which allow the quality of an individual performer to be measured independently of the outcome on a given day, be established. Currently T20 is nascent - its players seem only slightly less clueless about it than the commentators who call these games. Counting dot balls instead of maiden overs is well and good, but dot balls are symptoms - whats really useful to know is why and how they ended up as dot balls.

In a game of chance, an ability to measure risk is gold. In T20, this gold rush is yet to commence. This post has been an effort to work out a way to think about the contest between bat and ball, given the Laws of Cricket, in T20.

Tweet