War of CricketDuckworth Lewis Method
The Duckworth-Lewis-Stern (DLS) method is the formula the International Cricket Council uses to set a fair, revised target when a rain-affected One Day International or Twenty20 match cannot be completed as scheduled. It replaced the original Duckworth-Lewis (D/L) method this page used to describe: statistician Steven Stern revised the model in 2014, and the ICC has used DLS in international cricket since the 2015 World Cup.
The problem it solves
Before D/L existed, rain-shortened chases were settled by methods such as Average Run Rate, which simply compared scoring rates and ignored how many wickets a team had lost. That let a team batting with wickets in hand chase a rate no side batting first could sustain for a full innings. Frank Duckworth and Tony Lewis, two English statisticians, introduced their method in 1997 and the ICC adopted it officially in 1999. Its insight is that a batting side has two resources to work with, not one: the overs it has left to face, and the wickets it has in hand. Lose either resource and the fair target changes; the method's job is to work out by how much.
How the resources table works
D/L publishes a reference table showing the percentage of batting resources a team has remaining for any combination of overs left and wickets in hand. A team starting its innings fresh has 100% of its resources (50 overs, 10 wickets). A team with 20 overs left and 10 wickets in hand has roughly 56.6% of its resources remaining — more than a simple 20/50 overs ratio (40%) would suggest, because batsmen can accelerate more freely with wickets intact. If a match is shortened, the revised target is Team 1's score multiplied by the ratio of Team 2's resources to Team 1's resources. Losing overs and losing wickets both reduce a team's resources, and reducing resources lowers the target — not the other way around; a side that loses more wickets before an interruption has fewer resources left, so its revised target falls, not rises.
A worked example
On 18 May 2003, Lancashire played Hampshire in the ECB National League. Rain before play reduced the match to 30 overs each; Lancashire batted first and scored 231-4. Before Hampshire's innings, the match was cut further to 28 overs. Using the Standard Edition resource table below, the calculation ran as follows.
| Overs remaining | 10 wkts | 8 wkts | 6 wkts | 4 wkts | 2 wkts |
|---|---|---|---|---|---|
| 30 | 75.1% | 67.3% | 54.1% | 33.6% | 11.9% |
| 28 | 71.8% | 64.8% | 52.6% | 33.2% | 11.9% |
Lancashire's resources: 30 overs, 10 wickets = 75.1%. Hampshire's resources: 28 overs, 10 wickets = 71.8%. Hampshire's par score = 231 × (71.8/75.1) = 220.85, so their target was 221 to win in 28 overs. Hampshire were bowled out for 150, and Lancashire won by 220 − 150 = 70 runs. A simple pro-rata cut (231 × 28/30 = 216) would have undervalued the resources Lancashire had used and handed Hampshire an easier chase.
Standard Edition vs Professional Edition
The table above is the Standard Edition, which uses one published table and can be calculated by hand — it is what the original version of this page described. International One Day Internationals and Twenty20 Internationals instead use the Professional Edition, adopted from 2004 after the 2003 World Cup final exposed a flaw in how the Standard Edition handled very large first-innings totals. The Professional Edition varies its resource percentages by score as well as by overs and wickets, so it requires software rather than a printed table — match officials run it on a laptop pitchside.
Why Stern revised it
By the early 2010s, one-day and T20 scoring rates had risen well beyond the levels the original D/L tables were built on. The old model under-valued how aggressively modern sides could bat with wickets in hand, which meant a team chasing a big target could sometimes protect wickets at what looked like a losing rate and still stay ahead of the D/L par score. Stern, who became custodian of the method after Duckworth and Lewis retired, rebuilt the resource curves to reflect contemporary scoring and increased the rate teams need to start at when chasing a large total. The method was renamed Duckworth-Lewis-Stern in November 2014 and used for the first time at a World Cup in 2015.
Limits of the method
DLS is not without critics. Because wickets carry heavy weight in the resource calculation, sides can still profit from batting cautiously and preserving wickets rather than chasing the required rate outright. Studies using ball-by-ball databases have found the DLS par score is a weaker predictor of the eventual winner in some match situations, and pundits regularly note that few players or spectators can explain the mechanics mid-game. For an ODI to be decided by DLS, both sides must face at least 20 overs (five overs in a T20); if the weather prevents that, the match is a no result rather than a DLS result.