By Anjali Deshmukh
Reviewed by Rajeev SatheThe Duckworth Lewis Stern method, or DLS, resets a target when rain cuts a limited-overs match short. It works from resources. That means the balls still to come plus the wickets still in hand, put into one percentage.
Below we show why simple scaling fails. We show how a par score becomes a target. We also show the honest limits. The tables used in world cricket are private. No fan can copy an official figure exactly.
The problem DLS was built to solve
Rain stops a limited-overs match. One side has batted 50 overs. The other has only 30. Who has won?
You cannot just compare totals. You cannot scale by overs either. We will show why, with numbers.
DLS answers that question with one idea: resources. The ICC adopted it as the way to set a fair winner when rain breaks up a match.
The name comes from three men. Frank Duckworth and Tony Lewis built it. Steven Stern later revised it. The published account of the method traces that history.
What a "resource" means
At any point in an innings a batting side has two things left. Balls to receive, and wickets in hand.
DLS squeezes both into a single percentage. A full 50-over innings with all ten wickets standing is 100% of resources.
The percentage falls as either thing runs down. Lose overs and it drops. Lose wickets and it drops.
The clever part: wickets and overs are not separate. Twenty overs left with all ten wickets is worth far more than twenty overs left with one wicket. The tables catch that.
Table 1. Some published Standard Edition resource values
Situation | Resources remaining |
|---|---|
50 overs left, no wickets down | 100% |
40 overs left, 2 wickets down | 77.8% |
30 overs left, no wickets down | 75.1% |
20 overs left, 2 wickets down | 52.4% |
Those four values come from the public Standard Edition table. It is copied in the reference account of the method. Look at the middle two rows. Thirty fresh overs are worth a bit less than forty overs with two wickets gone.
Why it is not a simple proportion
The instinct is to scale. If a side gets 30 overs instead of 50, give it 60% of the runs. That instinct is badly wrong.
Simple scaling against DLS.
Inputs: the first side scored 250 from 50 overs. The second side gets only 30 overs. It has all ten wickets.
Simple scaling: 250 times 30 divided by 50 = 150 runs.
The DLS way: 30 overs and 10 wickets = 75.1% of resources. Par = 250 times 75.1 divided by 100 = 187.75 runs.
Result: DLS asks for 37.75 runs more than scaling does.
Assumptions: the 75.1% figure is a public Standard Edition value. Real world matches use the Professional Edition. Its numbers are not public. So an official figure would differ.
Why the gap? A side chasing in 30 overs still has all ten wickets. It can attack from the first ball. Runs per over rise fast when you can afford to lose wickets.
Plain scaling would hand the chasing side an easy win every time rain fell. DLS shuts that door.
Par score, target and tie
The par score is the number the chasing side should be on. It is set by the resources that side has used. The target is one run more than par.
From par to target.
Inputs: par score of 187.75 runs, from the example above.
Formula: target to win = par rounded up to the next whole run. Score to tie = par rounded down.
Result: target to win = 188. Score to tie = 187. A score of 186 or less is a defeat.
Assumptions: the ICC ODI playing conditions ask for a whole-number target. They also state that one run less is a Tie.
A tie is a real result in its own right. Law 16 of the MCC Laws says a win means passing the other side's completed-innings total. A dead heat is not a win.
Par is also used live. When a match is stopped mid-chase, the umpires compare the score to par at that exact ball. Equal to par is a tie, not a win.
The ODI conditions add a floor. The side batting second must have had the chance to face at least 20 overs. If not, there is no result to work out.
When the chasing side gets more resources, not fewer
Rain sometimes falls in the first innings. The side batting second may then have more resources than the first side had.
The Standard Edition treats the two cases in different ways. If the second side has fewer resources, par is the first side's score times the ratio of resources. If it has more, the extra runs come from a fixed number called G50.
The two Standard Edition formulas.
Notation: S is the first side's score. R1 is the first side's resources. R2 is the second side's resources.
Case R2 less than R1: par = S times R2 divided by R1.
Case R2 greater than R1: par = S plus G50 times (R2 minus R1) divided by 100.
Assumptions: G50 is a set average score for 50 overs. The Professional Edition drops G50. It uses S times R2 divided by R1 in both cases.
G50 has changed over the years. Public values for full members were 225 from 1999, then 235 from September 2002, then 245 from 2009-10. Later values are not in print anywhere we could reach.
We still need the current G50 value used in the Standard Edition for full-member matches. It would come from an ICC or DLS methodology file. The figures above stop at the 2013-14 season.
The honest part: the real tables are not public
We have to be straight with you here. You cannot work out an official DLS target at home.
There are two editions. The Standard Edition uses one public table of resource values. The Professional Edition is used in world cricket. It has a different table for every first-innings total. Those tables sit inside software.
The Professional Edition numbers were withheld for commercial confidentiality. The reference account records that. This is the plain reason our worked example carries a caveat.
The ICC's own DLS page lists three files. There is a methodology document, a Standard Edition table and a list of common questions. The page itself prints no numbers.
So treat our numbers as sums done on public Standard Edition values. The maths is right. The official figure for a real match may not match it.
This is a real openness problem. It is not a flaw in our maths. A fan cannot check a result that decided a World Cup place.
A mid-innings interruption, step by step
Most interruptions are not clean. Play stops, overs are lost, and play resumes. Resources are then counted in pieces.
Resources lost to a rain break.
Inputs: a side starts a 50-over innings, so 100% of resources. Rain comes after 10 overs, with two wickets down. At that point 40 overs are left and 8 wickets stand. That is worth 77.8%. Play starts again with only 20 overs left and the same 8 wickets. That is worth 52.4%.
Formula: resources used = start resources minus resources at the stop, plus resources at the restart.
Result: 100% minus 77.8% plus 52.4% = 74.6% of resources used. The break wiped out 77.8 minus 52.4 = 25.4% of resources.
Assumptions: the public values quoted are for 2 wickets down. We use them for both moments. Real tables are keyed by wickets lost. The official sum would use Professional Edition numbers.
That 74.6% is then set against the other side's resources to give par. More breaks work the same way, one block at a time.
Which matches use DLS
DLS applies to limited-overs cricket. It has no role in a Test. A Test has no target in overs and can end in a draw.
Both ODIs and T20 internationals use it. The ICC playing conditions library holds the rule book for each. The T20I conditions carry the same maths in a 20-over frame.
An innings can end early with no rain at all. Law 13 counts an innings as complete when a side is all out. So a side bowled out inside its overs has used all its resources.
Short formats squeeze the method hard. In a T20 cut to eight overs, one over is a big share of the innings. So par moves in big jumps.
The format rules themselves are on our 50-over guide, our T20 guide and our Test guide.
DLS results in a tournament table
A DLS result still has to feed a group table. Net run rate needs a number for a side that may not have finished its innings.
ESPNcricinfo's rules handle it two ways. Take a match called off with a DLS result. The first side is then credited with the second side's par score off the overs it faced. Now take a match played out on a revised target. The first side is then credited with one run less than the second side's final target.
Our net run rate guide walks through the base formula with the full sums.
How we checked this page
We ran every sum above through a calculator, not from memory. We read the two formulas and the resource values from public sources in August 2026.
We have also flagged the two things we could not confirm. One is the current G50 number. The other is any Professional Edition value. Both gaps come from the system itself, not from us.
Our verification method sets out the standard. Fixes go on the corrections page. All our explainers sit on the guides hub.
Questions people ask
Can I calculate an official DLS target myself?
No. World matches use the Professional Edition. Its resource tables sit inside software. Its numbers were withheld for commercial reasons, as the reference account of the method records. Only Standard Edition values are public.
Why is the DLS target higher than a simple proportion?
A side chasing in fewer overs still has all its wickets. So it can attack from the start. In our example, plain scaling asked for 150 runs from 30 overs. DLS asked for a par of 187.75.
What is the difference between par score and target?
Par is the exact number, often with decimals. The target to win is par rounded up. The score to tie is par rounded down. A par of 187.75 gives a target of 188 and a tie at 187.
What is G50?
G50 is a set average score for a full 50-over innings. The Standard Edition uses it when the chasing side has more resources than the first side. Public values ran from 225 to 245. We could not confirm the current one.
Is DLS used in Test cricket?
No. A Test has no target in a set number of overs. It can also end in a draw. So no rain rule is needed. DLS is for limited-overs cricket only.
How many overs must the chasing side get for DLS to apply in an ODI?
At least 20. The ICC ODI playing conditions call the match a No Result if the second side has not had the chance to bat that long. There are two exceptions: the side is all out, or it has passed the target.
