How DLS Works: The Formula Behind Rain Interruptions
The Duckworth-Lewis-Stern (DLS) method is the mathematical system used to set revised targets when rain interrupts limited-overs cricket. This article explains how DLS works conceptually, what 'resources' means in the system, and why DLS sometimes produces controversial targets.
What DLS Measures
The DLS method is based on the concept of 'resources' — the combined value of wickets remaining and overs remaining that a batting team has available to score runs: a team starting an innings has 100% of their resources (all 10 wickets and all overs); as wickets fall and overs are used, the remaining resources decrease; the specific insight: wickets remaining and overs remaining are not independently valued — 5 wickets and 20 overs is worth more than just the sum of '5 wickets' plus '20 overs' because the interaction of wickets and overs creates the scoring opportunity; and when rain reduces a match's over count, the DLS tables calculate what percentage of resources each team has had access to and adjusts the target proportionally.
How Revised Targets Are Set
The DLS revised target calculation for a rain-interrupted match: if Team A bats first and scores 250, then rain reduces the second innings to 40 overs instead of 50, Team B doesn't simply need 250 × (40/50) = 200; instead, the DLS tables calculate what percentage of resources Team A had (they had 100% — they batted 50 complete overs with all 10 wickets available); then calculate what percentage of resources Team B will have (40 overs with all 10 wickets — approximately 90% of a team's resources in a 50-over match); and Team B's target is set so that Team B needs the score that corresponds to Team A's scoring ratio, adjusted for the resource percentage difference; if Team A used 100% of resources to score 250, Team B should need 250 × (B's resource%) / (A's resource%) to win. The specific tables used in DLS are calibrated from extensive analysis of completed matches across multiple formats.
Why DLS Sometimes Produces Controversial Targets
DLS targets are occasionally criticised for being 'unfair' — specific scenarios produce results that feel counterintuitive: when the first team bats in easy conditions and the second team faces difficult conditions after rain, DLS doesn't adjust for conditions — it adjusts only for resources; when rain interrupts the first innings rather than the second (affecting DLS in a different way), the target setting is more complex and occasionally produces targets that are significantly higher than what seems reasonable; and when a team is performing very well at the moment of interruption (e.g., batting at 12 runs per over when rain comes) the DLS par score at that moment may be below the team's actual scoring rate, causing the batting team to feel they would have scored more without the interruption. DLS is the best mathematical solution available given the information it uses, but it cannot perfectly model all possible match scenarios.
Frequently asked questions
When does DLS apply vs when is a match simply abandoned?
DLS applies when a match can continue with at least a minimum number of overs: ICC playing conditions specify minimum over counts that must be completed for a result to be valid — in ODIs, a minimum of 20 overs per side must be bowled for a DLS result to count (slightly different minimum for specific tournaments); in T20Is, the minimum is typically 5 overs per side; if rain prevents reaching the minimum over count, the match is abandoned (no result) — this outcome is specific to weather that prevents any meaningful cricket from occurring; and mid-match interruptions that allow more than the minimum overs use DLS to produce a result. The minimum over threshold exists because DLS is less accurate when applied to very short samples — a DLS calculation based on 5 overs of data is significantly less reliable than one based on 25 overs.
How accurate is DLS compared to actual match outcomes?
DLS is the most mathematically validated rain-rule system in cricket, but its accuracy is limited by what it can model: academic studies comparing DLS targets to what teams would have been expected to score without rain interference show that DLS overestimates targets in approximately 45% of cases and underestimates in approximately 40% — the remaining 15% are considered accurate within a small margin; the specific scenarios where DLS is least accurate: very high first-team totals (DLS underperforms at the extremes of scoring distributions); matches on pitches with unusual playing conditions (very slow, very fast) that don't match the historical dataset used to calibrate the DLS tables; and tail-end interruptions (rain after 45 overs of a 50-over first innings — the DLS resource tables are less calibrated for these late-first-innings interruptions). The statisticians who maintain DLS (David Stern has updated the method in the 'Stern' portion of the DLS name) regularly recalibrate the resource tables as new match data accumulates.
Is there a different system from DLS used anywhere?
Yes — some cricket boards have used or proposed alternative rain-rule systems, though none have achieved the global adoption of DLS: the V/Jayadevan (VJD) method is an alternative mathematical system proposed by Indian engineer Jayadevan that uses different resource curves from DLS — it is specifically used in some domestic Indian competitions (BCCI has applied it in Ranji Trophy matches) and was proposed as an alternative to DLS for international cricket (the BCCI formally requested the ICC consider VJD as an alternative); the ICC evaluated VJD and decided to retain DLS for all international cricket, citing DLS's broader validation and acceptance; and the original D/L method (before Stern's updates) was a simpler version of the current system. At the international level, DLS is universal — national boards that use alternatives do so only in domestic competition.