Everyone Chooses to Chase. Our Numbers Say It Barely Matters
Across 963 T20 internationals with the toss and the result recorded, the side that won the toss and bowled first won 48.3 per cent. The side that batted first won 49.5 per cent.
Watch enough T20 internationals and you will hear the same sentence at every toss. The captain wins it, looks at the pitch, and says he will have a bowl. Chasing is easier. You know what you need.
The numbers do not really support it.
What we looked at
Across 963 T20 internationals for which we hold both the toss decision and the result:
- Captains chose to bowl first 563 times. The side that won the toss went on to win 48.3 per cent of those matches.
- Captains chose to bat first 400 times. The side that won the toss went on to win 49.5 per cent of those matches.
Both figures sit within a couple of points of a coin flip, and the decision that is supposed to be the advantage is the marginally worse of the two.
Why the belief persists anyway
Because it is true in the cases people remember. Chasing is genuinely easier when dew is going to arrive, when the pitch is expected to slow down, and in a tournament where a specific ground has a known pattern. In those conditions the captain is right and the data agrees with him.
The problem is that the decision has stopped being a decision. Bowling first is now the default across conditions where it confers nothing, and a default cannot be an edge. If 563 of 963 captains choose the same option regardless of circumstances, the option gets priced in.
What is really going on
Two things. First, the toss winner is not a random sample of the two teams, but close enough to it that a genuine advantage would show up clearly across 963 matches. It does not.
Second, T20 is short. Twenty overs is not long enough for a moderate structural advantage to reliably decide a match. A dropped catch in the fifteenth over moves the result more than the toss does, which is uncomfortable for anyone building a model and liberating for anyone watching.
What this does not mean
It does not mean the toss is irrelevant at every ground. Ground level and tournament level effects are real, and at some venues, in some months, dew decides matches. A number across all T20 internationals averages those conditions away by design.
It means the blanket rule is wrong. A captain who bowls first because the conditions call for it is doing his job. A captain who bowls first because that is what captains do is flipping a coin twice and calling it a plan.
How we use it
Our [batting forecasts](/predictions) treat the innings a player bats in as one input among many, rather than assuming the chasing side is at an advantage before a ball is bowled. On this evidence, it is not.