Explainer · evergreen
The DLS method, explained: how rain actually changes a cricket target
In one line: The Duckworth-Lewis-Stern method in plain terms — resources, par scores and why “20 overs, 150 to win” is arithmetic, not guesswork.
Every rain-affected cricket match produces the same argument: the revised target looks impossible, or suspiciously easy, and someone declares the maths a joke. The Duckworth-Lewis-Stern method is not a joke and not an opinion — it is a resource model, and once you see the two resources it tracks, the targets stop looking arbitrary.
The two resources
DLS says a batting team has two things that matter: overs remaining and wickets in hand. A team with 10 overs and 10 wickets left is richer than a team with 10 overs and 3 wickets left, because wickets buy the right to attack. The method models this with a resources table (originally derived from hundreds of completed matches): any game state — 25 overs left, 7 wickets standing — maps to a percentage of a full innings remaining. A full 50-over innings with all 10 wickets = 100% of the resource. Rain removes overs; wickets remove the other thing; the table prices both. The official overview and current regulations live at the ICC's DLS page, and the derivation history is in Wikipedia's DLS entry.
Where a rain target comes from
Suppose Team A scores 250 in 50 overs, and rain arrives with Team B 10 overs into the chase. DLS reads Team A's innings as a completed 100%-resource effort (in the standard formulation), reads Team B's remaining resources from the table at the interruption — say 62% — and sets the target proportionally: 250 × 62% ≈ 155. That is the par score you see on screen. If more rain follows, the resources recalculate again at each stoppage. The famous steep final-over targets are the same maths at the resource cliff: when overs run out while wickets remain, the marginal value of each ball is enormous.
Why it beats the alternatives
The pre-DLS methods (run-rate averages, percentage-of-score rules) had fatal flaws: they ignored wickets entirely and incentivised slow scoring before rain. DLS-Stern (the current version, tuned for high-scoring modern play) removes the perverse incentives — the optimal strategy at every point is simply to score as fast as you safely can, which is what a spectator would want anyway. The residual complaints are real but smaller than folklore suggests: par-score displays confuse target with level, and shortened games carry genuine uncertainty that no model can price perfectly.
How to read a rain break
When the covers come on, ignore the punditry and read three numbers: the par score at the interruption, the overs each side will have left, and the wickets standing. If the chasing side is above par with wickets in hand, they are ahead of the model; below par with wickets in hand, the model thinks the bowling side has done well. That is the whole language of rain in cricket — a resource argument, updated whenever the weather and the rules allow.
Par is not the target
Broadcast graphics often show a par score, which is the score the chasing side would need at that moment to be level with the first innings under the model. The target is the final number they must reach to win, usually one more than the par score once the innings is shortened. Confusing these two numbers creates much of the apparent absurdity. The same distinction applies when officials calculate a result after a match is abandoned: the scheduled overs, the overs actually available and the wickets lost all matter, and a minimum-overs rule may determine whether a result is possible at all.
Why no simple run-rate rule works
A team at 100 for two after 20 overs is not in the same position as one at 100 for eight, even though the run rate matches. The second side has fewer batting resources left and cannot attack with the same freedom. A fair revision therefore has to price both time and wickets, not merely multiply the current rate by the overs remaining. That is why the formula looks less intuitive than a straight rate calculation: it is answering a harder question. Spectators can disagree with a particular outcome while still seeing why a resource-based method is fairer than the alternatives.
Sources
Next