Duckworth–Lewis method

This article is about the cricketing term. For the Irish justment to the second team’s target after interrup-
pop group, see The Duckworth Lewis Method. tions in the first innings is often an increase (though
a decrease is possible). Although both teams have
The Duckworth–Lewis method (often written as D/L 10 wickets and the same (reduced) number of overs
available, an increase is fair as, for some of their in-
method) is a mathematical formulation designed to cal-
culate the target score for the team batting second in nings, the first team thought they would have more
overs available than they actually ended up hav-
a limited overs cricket match interrupted by weather or
other circumstances. It is generally accepted to be the ing. They would have batted less conservatively, and
most accurate method of setting a target score. The D/L scored more runs at the expense of more wickets, if
method was devised by two English statisticians, Frank they had known that their innings was going to be
Duckworth and Tony Lewis.[1] shorter than initially thought. Increasing the second
team’s target score neutralizes the injustice done to
The basic principle is that each team in a limited-overs the first team when they were denied some of the
match has two resources available with which to score overs to bat they thought they would get.
runs: wickets remaining, and overs to play. Where overs
are lost, setting an adjusted target for the team batting • If there are interruption(s) before or during the
second is not as simple as reducing the run target pro- second team’s innings, or the second team’s in-
portionally to the loss in overs, because a team with ten nings is cut short, then D/L will reduce the second
wickets in hand and 25 overs to bat can be expected to team’s target score from the initial target set at the
play more aggressively than if they had ten wickets and a end of the first team’s innings, in proportion to the
full 50 overs, for example, and can consequently achieve reduction in the second team’s resources. If there
a higher run rate. The Duckworth–Lewis method is an at- are multiple interruptions in the second innings, the
tempt to set a statistically fair target for the second team’s target will be adjusted downwards each time.
innings, based on the score achieved by the first team, tak-
ing their wickets lost and overs played into account. • If there are interruptions which both increase and
In November 2014, Duckworth-Lewis method was re- decrease the target score, then the net effect on the
named as Duckworth-Lewis-Stern (or D/L/S) method.[2] target could be either an increase or decrease, de-
pending on which interruptions were bigger.

1 Summary of impact on the sec-

2 Examples
ond team’s target
2.1 Stoppage in first innings
• If there is a delay before the first innings starts,
so that the numbers of overs in the two innings are
Increased target
reduced (but still the same as each other), then D/L
will make no change to the target score, as both sides In the 4th India – England ODI in the 2008 series, the
will be in the same position of having the same num- first innings was interrupted by rain on two occasions, re-
ber of overs and 10 wickets available, and knowing sulting in the match being reduced to 22 overs each. India
this throughout their innings. (batting first) made 166/4. England’s target was set by the
D/L method at 198 from 22 overs. As England knew they
• The target score for the second team is first calcu- had only 22 overs the expectation is that they will be able
lated once the first team’s innings has finished. If to score more runs from those overs than India had from
there were interruption(s) during the first team’s their (interrupted) innings. England made 178/8 from 22
innings, or the first team’s innings was cut short, overs, and so the match was listed as “India won by 19
so that the numbers of overs in the two innings are runs (D/L method)".[3]
reduced (but still the same as each other), then (in During the fifth ODI between India and South Africa in
the Professional Edition) D/L will adjust the second January 2011, rain halted play twice during the first in-
team’s target score in proportion to the two teams’ nings. The match was reduced to 46 overs each and South
relative resources, i.e., overs and wickets. The ad- Africa scored 250/9. The D/L method was applied which

1
2 3 THEORY

adjusted India’s target to 268. As the number of overs revised target of 139. Perth won the game by 8 wickets
was reduced during South Africa’s innings, this method with a boundary off the final ball.[8][9]
takes into account what South Africa are likely to have
scored if they'd known throughout their innings that it
would only be 46 overs long. 3 Theory
Decreased target
On 3 December 2014, Sri Lanka played England and bat-
ted first, but play was interrupted when Sri Lanka had
scored 6/1 from 2 overs. At the re-start both innings were
reduced to 35 overs, and Sri Lanka finished on 242/8.
England’s target was set by D/L at 236 from 35 overs.[4]
Although Sri Lanka had less resource remaining to them
after the interruption than England would have for their
whole innings (about 7% less), they'd used up so much
resource before the interruption (2 overs and 1 wicket,
about 8%), that the total resource used by Sri Lanka
was still slightly more than England would have available,
hence the slightly decreased target for England.

2.2 Stoppage in second innings Scoring potential as a function of wickets and overs.

A simple example of the D/L method being applied was The essence of the D/L method is 'resources’. Each team
the first ODI between India and Pakistan in their 2006 is taken to have two 'resources’ to use to make as many
ODI series.[5] India batted first, and were all out for runs as possible: the number of overs they have to receive;
328. Pakistan, batting second, were 311/7 when bad light and the number of wickets they have in hand. At any point
stopped play after the 47th over. Pakistan’s target, had in any innings, a team’s ability to score more runs depends
the match continued, was 18 runs in 18 balls, with three on the combination of these two resources. Looking at
wickets in hand. Considering the overall scoring rate historical scores, there is a very close correspondence be-
throughout the match, this is a target most teams would be tween the availability of these resources and a team’s final
favoured to achieve. And indeed, application of the D/L score, a correspondence which D/L exploits.[10]
method resulted in a retrospective target score of 305 (or Using a published table or computer which gives the per-
par score of 304) at the end of the 47th over, with the re- centage of these combined resources remaining for any
sult therefore officially listed as “Pakistan won by 7 runs number of overs (or, more accurately, balls) left and
(D/L Method)". wickets lost, the target score can be adjusted up or down
The D/L method was used in the group stage match be- to reflect the loss of resources to one or both teams when
tween Sri Lanka and Zimbabwe at the 20/20 World Cup a match is shortened one or more times. The two teams’
in 2010. Sri Lanka scored 173/7 in 20 overs batting first, resource percentages are found, and used to calculate a
and in reply Zimbabwe were 4/0 from 1 over when rain 'par score' for the second team that is usually a fractional
interrupted play. At the restart Zimbabwe’s target was re- number of runs. The target score is this number rounded
duced to 108 from 12 overs, but rain stopped the match up to the next integer, and the score to tie (also called the
when they had scored 29/1 from 5 overs. The retrospec- par score), is this number rounded down to the next inte-
tive D/L target from 5 overs was a further reduction to ger. If the second team reaches or passes the target score,
44, or a par score of 43, and hence Sri Lanka won the then they have won the match. If the match ends when
match by 14 runs.[6][7] the second team has exactly met (but not passed) the par
score then the match is a tie. If the second team fail to
reach the par score then they have lost.
2.3 Stoppages in both innings An example of such a tie was found in the one day interna-
tional between England and India on 11 September 2011.
During the 2012/13 KFC Big Bash League, D/L was used This match was frequently interrupted by rain in the final
in the 2nd semi-final played between the Melbourne Stars overs, and a ball-by-ball calculation of the Duckworth–
and the Perth Scorchers. After rain delayed the start of Lewis 'par' score played a key role in the tactical deci-
the match, it interrupted Melbourne’s innings when they sions made during those overs. At one point, India were
had scored 159/1 off 15.2 overs, and both innings were re- ahead according to this calculation, during one rain delay
duced by 2 overs to 18, and Melbourne finished on 183/2. (and would have won if play was unable to be resumed).
After a further rain delay reduced Perth’s innings to 17 At a second rain interval, England, who had scored some
overs, Perth returned to the field to face 13 overs, with a quick runs (precisely because they were aware of the need
 3

to get ahead in D/L terms) would correspondingly have Tony Lewis said, 'We were then [at the time of the 2003
won if play had not resumed. Play was finally called off Cricket World Cup Final] using what is now known as
with just 7 balls of the match remaining and England’s the Standard Edition. We knew that under normal cir-
score equal to the Duckworth–Lewis 'par' score, there- cumstances it worked well, but we had also known for a
fore resulting in a tied match. while that in case of very big totals the approach was not
This example does show how crucial (and difficult) the really that good. We did have in place a computerised
decisions of the umpires can be, in assessing at exactly version, but it meant that the transparency was lost. You
what point the rain is heavy enough to justify ceasing couldn’t do it manually by looking up the tables. Up un-
til that point the ICC were very happy with the manual
play. If the umpires of that match had halted play one ball
earlier, England would have been ahead on D/L, and so version and the transparency that came with it. But Aus-
tralia got 359 and that showed up the flaws and straight-
would have won the match. Equally, if play had stopped
one ball later, without England scoring off that ball, India away the next edition was introduced which handled high
scores much better. The par score for India is likely to be
would have won the match – indicating how finely-tuned
D/L calculations can be in such situations. much higher now.'[12]
Duckworth and Lewis wrote, 'When the side batting first
score at or below the average for top level cricket..., the
4 Application results of applying the Professional Edition are generally
similar to those from the Standard Edition. For higher
scoring matches, the results start to diverge and the dif-
As with most non-trivial statistical derivations, the D/L ference increases the higher the first innings total. In ef-
method can produce results that are somewhat counter in- fect there is now a different table of resource percentages
tuitive, and the announcement of the derived target score for every total score in the Team 1 innings.'[13]
can provoke a good deal of second-guessing and discus-
sion amongst the crowd at the cricket ground. This can The decision on which edition should be used is
also be seen as one of the method’s successes, adding in- for the cricket authority which runs the particular
terest to a “slow” rain-affected day of play. competition.[14] The ICC Playing Handbook,[15] the of-
ficial handbook for international matches, says 'The Pro-
For 50-over matches, each team must face at least 20 fessional Edition of the Duckworth–Lewis method shall
overs before D/L can decide the game, unless one or both be used in all [international] matches... Where possible,
sides have been bowled out in less than 20 overs and/or arrangements shall be made for the provision of back-up
the team batting second has reached its target in less than capability, in case of computer malfunction, for the oper-
20 overs. For Twenty20 games, each side must face at ation or continued operation of the Professional Edition.
least five overs before D/L can decide the game, unless In the event of computer non-availability or malfunction
one or both sides have been bowled out in less than five where no such provision has been made, the Standard
overs and/or the team batting second has reached its tar- Edition (the method in use prior to October 2003) shall
get in less than five overs. If these prerequisites are not be used.'[16] This also applies to most countries’ national
met, the match is declared a no result. competitions.[14] At lower levels of the game, where use
of a computer cannot always be guaranteed, the Standard
Edition is used.[17]
4.1 Standard Edition and Professional
Edition

Until 2003, a single version of the D/L method was 5 Calculations

in use. This used a single published reference table
of total resource percentages remaining for all possible
Using the notation of the ICC Playing Handbook,[15] the
combinations of overs and wickets,[11] and some simple
team that bats first is called Team 1, their final score is
mathematical calculations, and was relatively transpar-
called S, the total resources available to Team 1 for their
ent and straightforward to implement. However, it had
innings is called R1, the team that bats second is called
a known flaw in how it handled very high first innings
Team 2, and the total resources available to Team 2 for
scores (350+). Following the 2003 World Cup, a sec-
their innings is called R2.
ond version was introduced which overcomes this flaw
by using substantially more sophisticated statistical mod-
elling. However, this version doesn't use a single table of
resource percentages, instead the percentages also vary 5.1 Standard Edition
with score, so the percentages must be obtained from a
computer. Therefore it loses some of the previous advan- For each reduction in overs, the loss in total resources
tages of transparency and simplicity. The original version available to the batting team is found using a published
was named the Standard Edition, and the new version was reference table,[11] then Team 2’s target score is changed
named the Professional Edition. as follows:
4 6 EXAMPLE STANDARD EDITION CALCULATIONS

• If R2 < R1, reduce Team 2’s target score in pro- • If R2 > R1, increase Team 2’s target score in pro-
portion to the reduction in total resources, i.e. S × portion to the increase in total resources, i.e. S ×
R2/R1. R2/R1. The problem of early high scoring rates
potentially producing anomalously high targets has
• If R2 = R1, no adjustment to Team 2’s target score been overcome in the Professional Edition, which
is needed. is essentially 'a different table of resource percent-
ages for every total score in the Team 1 innings.'[13]
• If R2 > R1, increase Team 2’s target score by the Therefore Team 2’s target score can be simply in-
extra runs that could be expected to be scored on creased in proportion to the increase in total re-
average with the extra total resource, i.e. S + G50 sources when R2 > R1,[18] and there is no G50.
× (R2 – R1)/100, where G50 is the average 50-over
total. Team 2’s target score is not simply increased
in proportion to the increase in total resources, i.e. However, the resource percentages used [23]
in the Profes-
S × R2/R1, as this 'could lead to some unrealisti- sional Edition are not publicly available, so a computer
cally high targets if Team 1 had achieved an early must be used which has the software loaded.
high rate of scoring [in the powerplay overs] and
rain caused a drastic reduction in the overs for the
match.'[18] Instead, D/L Standard Edition requires 6 Example Standard Edition cal-
average performance for Team 2’s additional re-
source over Team 1. culations
As the resource percentages used in the Professional Edi-
5.1.1 G50 tion are not publicly available, it is difficult to give exam-
ples of the D/L calculation for the Professional Edition.
G50 is the average score expected from the team batting Therefore examples are given from when the Standard
first in an uninterrupted 50 overs-per-innings match. This Edition was widely used, which was up to early 2004.
will vary with the level of competition and over time. The
annual ICC Playing Handbook[15] gives the values of G50
to be used each year when the D/L Standard Edition is 6.1 Reduced target: Team 1’s innings
applied:
completed; Team 2’s innings delayed
Duckworth and Lewis write, 'We accept that the value
of G50, perhaps, should be different for each country, On 18 May 2003, Lancashire played Hampshire in the
or even for each ground, and there is no reason why any National League.[24][25][26] Rain before play reduced the
cricket authority may not choose the value it believes to match to 30 overs each. Lancashire batted first and scored
be the most appropriate. In fact it would be possible for 231–4 from their 30 overs. Before Hampshire began their
the two captains to agree a value of G50 before the start of innings, it was further reduced to 28 overs.
each match, taking account of all relevant factors. How-
ever, we do not believe that something that is only invoked Hampshire’s target was therefore 221 to win (in 28 overs),
if rain interferes with the game should impose itself on or 220 to tie. They were all out for 150, giving Lancashire
every game in this way. In any case, it should be realised victory by 220 − 150 = 70 runs.
that the value of G50 usually has very little effect on the If Hampshire’s target had been reduced simply in propor-
revised target. If 250 were used, for instance, instead of tion to the reduction in overs, their par score would have
235, it is unlikely that the target would be more than two been 231 x 28/30 = 215.6. This would have kept the re-
or three runs different.'[22] quired run rate the same as Lancashire achieved (7.7 runs
per over), but this would have given an unfair advantage
to Hampshire, as it’s easier to achieve a run rate for a
5.2 Professional Edition shorter period. Increasing Hampshire’s target from 215
overcomes this.
For each reduction in overs, the loss in total resources
available to the batting team is found using a computer,
then Team 2’s target score is changed as follows: 6.2 Reduced target: Team 1’s innings
completed; Team 2’s innings cut short
• If R2 < R1, reduce Team 2’s target score in pro-
portion to the reduction in total resources, i.e. S × On 3 March 2003, Sri Lanka played South Africa in the
R2/R1. 2003 Cricket World Cup Pool B.[27][28] Sri Lanka batted
first and scored 268–9 from their 50 overs. Chasing a
• If R2 = R1, no adjustment to Team 2’s target score target of 269, South Africa had reached 229–6 from 45
is needed. overs when play was abandoned.
6.5 Increased target: Multiple interruptions in Team 1’s innings; Team 2’s innings completed 5

Therefore, South Africa’s retrospective target from their 6.5 Increased target: Multiple interrup-
45 overs was 230 runs to win, or 229 to tie. In the event, tions in Team 1’s innings; Team 2’s in-
as they had scored exactly 229, the match was declared a nings completed
tie.
South Africa scored no runs off the very last ball. If On 20 February 2003, Australia played Netherlands in
play had been abandoned without that ball having been the 2003 Cricket World Cup Pool A.[33][34][35][36] Rain
bowled, the resource available to South Africa at the before play reduced the match to 47 overs each, and Aus-
abandonment would have been 14.7%, giving them a par tralia batted first.
score of 228.6, and hence victory.

• Rain stopped play when they had reached 109–2

6.3 Reduced target: Team 1’s innings from 25 overs (i.e. 22 remaining). At the restart
completed; Team 2’s innings inter- both innings were reduced to 44 overs (i.e. 19 re-
rupted maining for Australia).

On 16 February 2003, New South Wales played South • Rain stopped play again when Australia had reached
Australia in the ING Cup.[29][30] New South Wales batted 123–2 from 28 overs (i.e. 16 remaining), and at the
first and scored 273 all out (from 49.4 overs). Chasing a restart both innings were reduced further to 36 overs
target of 274, rain interrupted play when South Australia (i.e. 8 remaining for Australia).
had reached 70–2 from 19 overs, and at the restart their
innings was reduced to 36 overs (i.e. 17 remaining).
South Australia’s new target was therefore 214 to win (in Australia finished on 170–2 from their 36 overs.
36 overs), or 213 to tie. In the event, they were all out for The Netherlands’ target was therefore 198 to win (in 36
174, so New South Wales won by 213 − 174 = 39 runs. overs), or 197 to tie. It is fair that their target was in-
creased, even though they had the same number of overs
to bat as Australia, as Australia would have batted less
6.4 Increased target: Team 1’s innings cut conservatively in their first 28 overs, and scored more runs
short; Team 2’s innings completed at the expense of more wickets, if they had known that
their innings would only be 36 overs long. Increasing the
On 25 January 2001, West Indies played Netherlands’ target score neutralizes the injustice done to
Zimbabwe.[31][32] West Indies batted first and had Australia when they were denied some of the overs to bat
reached 235–6 from 47 overs (of a scheduled 50) when they thought they would get. The Netherlands were all
rain halted play for two hours. At the restart, both innings out for 122, giving Australia victory by 197 − 122 = 75
were reduced to 47 overs, i.e. West Indies’ innings was runs.
closed immediately, and Zimbabwe began their innings. This formula for Netherlands’ par score comes from the
Zimbabwe’s target was therefore 253 to win (in 47 overs), Standard Edition of D/L, which was used at the time.
or 252 to tie. It is fair that their target was increased, Currently the Professional Edition is used, which has a
even though they had the same number of overs to bat different formula when R2>R1. The formula required
as West Indies, as West Indies would have batted more Netherlands to match Australia’s performance with their
aggressively in their last few overs, and scored more runs, overlapping 72.6% of resource (i.e. score 170 runs), and
if they had known that their innings would be cut short achieve average performance with their extra 84.1% −
at 47 overs. Zimbabwe were all out for 175, giving West 72.6% = 11.5% of resource (i.e. score 11.5% of G50
Indies victory by 252 − 175 = 77 runs. (235 at the time) = 27.025 runs).
These resource percentages are the ones which were in After the match there were reports in the media[34] that
use back in 2001, before the 2002 revision, and so do Australia had batted conservatively in their final 8 overs
not match the currently used percentages for the Standard after the second stop, to avoid losing wickets rather than
Edition, which are slightly different. Also, the formula maximising their numbers of runs, in belief that this
for Zimbabwe’s par score comes from the Standard Edi- would further increase the Netherlands’ par score. How-
tion of D/L, which was used at the time. Currently the ever, if this is true, this belief was mistaken, in the same
Professional Edition is used, which has a different for- way that conserving wickets rather than maximising runs
mula when R2>R1. The formula required Zimbabwe to in the final 8 overs of a full 50-over innings would be a
match West Indies’ performance with their overlapping mistake. At that point the amount of resource available to
89.8% of resource (i.e. score 235 runs), and achieve aver- each team was fixed (as long as there were no further rain
age performance with their extra 97.4% − 89.8% = 7.6% interruptions), so the only undetermined number in the
of resource (i.e. score 7.6% of G50 (225 at the time) = formula for Netherlands’ par score was Australia’s final
17.1 runs). score, so they should have tried to maximise this.
6 9 HISTORY AND CREATION

7 Calculation of an innings’ re- called the D/L par score, may be shown on a computer
printout, the scoreboard and/or TV alongside the actual
source percentage score, and updated after every ball. This can happen in
matches which look like they're about to be shortened by
Although the examples above may make it look like find- the weather, and so D/L is about to be brought into play,
ing the total resource percentage requires a different cal- or even in matches completely unaffected by the weather.
culation for each different scenario, the formula is actu- This is:
ally the same each time − it’s just that different scenarios,
with more or less interruptions and restarts, need to use
more or less of the same formula. The calculations above • To help spectators and players understand whether
just show and use the parts of the formula that are rele- the chasing side are doing better or worse than they
vant for that scenario. would need to do on average to reach the target
score.
The total resources available to a team are given by:[11]
Total Resources • The score the batting team’s score would be com-
resources = start at − pared to to determine which side had won, if the
available innings of match had to be abandoned right then. It is the par
by lost Resources score which is displayed, i.e. the score to tie. The
− target, to win, score is one run more than this. South
interruption first
Africa exited the 2003 World Cup by tieing with Sri
by lost Resources
− Lanka by mistakenly believing the par score on the
interruption second
printout was the target score.[37][38]
by lost Resources
− etc...
interruption third
 
Resources 8.2 Net Run Rate calculation
Resources  remaining Resources 
= start at −  first at − remaining  − It has been suggested that when a side batting second
innings of restart at
interruption successfully completes the run chase, the D/L method
 
Resources could be used to predict how many runs they would have
 remaining Resources scored with a full innings (i.e. 50 overs in a One Day
 
 second at − remaining − International), and use this prediction in the net run rate
restart at calculation.[39]
interruption 
Resources This suggestion is in response to the criticisms of NRR
 remaining Resources
  that it doesn't take into account wickets lost, and that it
 third at − remaining − ... unfairly penalizes teams which bat second and win, as
restart at
interruption those innings are shorter and therefore have less weight
Each time there’s an interruption or a re-start after an in- in the NRR calculation than other innings which go the
terruption, the resource remaining percentages at those full distance.
times (obtained from a reference table for the Standard
Edition, or from a computer for the Professional Edition)
can be entered into the formula, with the rest left blank. 9 History and creation
In the examples above:
The D/L method was devised by two British statisticians,
Frank Duckworth and Tony Lewis, as a result of the out-
8 Other uses come to the semi-final in the 1992 Cricket World Cup
between England and South Africa, where the most pro-
There are uses of the D/L method other than finding ductive overs method was used. Rain stopped play for
the current official final target score for the team batting 12 minutes with South Africa needing 22 runs from 13
second in a match that has already been reduced by the balls chasing England’s 252/6 off 45 overs. The revised
weather. target left South Africa needing 21 runs from one ball,
which was a reduction of only one run compared to a re-
duction of two overs, and a preposterous target given that
8.1 Ball-by-ball par score the maximum score from one ball is generally six runs.[40]
The D/L method avoids this flaw: in this match, the re-
During the second team’s innings, the number of runs vised D/L target would have left South Africa four to tie
a chasing side would expect to have scored on average or five to win from the final ball.[41] Duckworth said, “I re-
with this number of overs used and wickets lost, if they call hearing Christopher Martin-Jenkins on radio saying
were going to successfully match the first team’s score, 'surely someone, somewhere could come up with some-
10.1 2002 7

thing better' and I soon realised that it was a mathematical matches.

problem that required a mathematical solution.”[42][43]
It was first used in international cricket in the second 10.1 2002
game of the 1996–97 Zimbabwe versus England One
Day International series, which Zimbabwe won by seven The resource percentages were revised, and G50 for
runs,[44] and was formally adopted by the International ODI’s was changed to 235, following an extensive anal-
Cricket Council in 1999 as the standard method of cal- ysis of limited overs matches in recent years. These
culating target scores in rain shortened one-day matches. changes came into effect on 1 September 2002.[20] As
of 2014, these resource percentages are the ones still in
use in the Standard Edition, though G50 has subsequently
9.1 Previous methods
changed.
Various different methods had been used previously, with The tables show how the percentages were in 1999 and
the most common being the average run-rate method, and 2001, and what they were changed to in 2002. Mostly
the most productive overs method. they were reduced.
All of these methods have flaws that are easily exploitable:
10.2 2004
• The average run-rate method takes no account of
how many wickets the team batting second have lost, From the 1999 Cricket World Cup match in Bristol be-
but simply reflects how quickly they were scoring tween India and Kenya, Tony Lewis noticed that there
when the match was interrupted, so if a team felt a was an inherent weakness in the formula used at the time
rain stoppage was likely they could attempt to force that would give a noticeable advantage to the side chas-
the scoring rate without regard for the corresponding ing a total in excess of 350. A correction was very soon
highly likely loss of wickets, skewing the compari- built into the formula and the software to correct this, by
son with the first team. including a 'match' factor. However, this minor correc-
• The most productive overs method also takes no ac- tion was not fully adopted by users until the 2004 update.
count of how many wickets the team batting second Updating the source data in its own right would reflect the
have lost, and also has the further effect of penal- overall trend that one-day matches were achieving signif-
izing the team batting second for good bowling, as icantly higher scores than in previous decades, affecting
their best overs are ignored in setting the revised tar- the historical relationship between resources and runs.
get. At the same time as this update, the D/L method was
also split into a Professional Edition and a Standard
An example of this is in the 1988/89 Benson and Hedges Edition.[48] The main difference is that while the Stan-
World Series Cup, where the average run rate method was dard Edition preserves the use of a single table and simple
used: in the third final between Australia and the West calculation – suitable for use in any one-day cricket match
Indies, rain stopped play for one hour and 25 minutes at any level – the Professional Edition uses substantially
with the West Indies needing 180 off 31.2 overs chasing more sophisticated statistical modelling, and requires the
Australia’s 226/4 off 38 overs. The revised target left the use of a computer. The Professional Edition has been
West Indies needing 61 off the 11.2 overs that remained, in use in all international one-day cricket matches since
and the West Indies won the match and the competition early 2004.
with 4.4 overs remaining and eight wickets in hand after
Desmond Haynes hit a Steve Waugh full toss for six. Aus-
tralian fans loudly booed this unsatisfactory conclusion, 10.3 2009
and criticism from the media and both captains led to the
average run rate method being replaced by the most pro- In June 2009, it was reported that the D/L method would
ductive overs method for setting revised targets in inter- be reviewed for the Twenty20 format after its appro-
rupted matches.[45] In this match, the D/L method would priateness was questioned in the quickest version of the
have increased the West Indies target to 232 to take into game. Lewis was quoted admitting that “Certainly, peo-
account a two-hour rain delay during Australia’s innings, ple have suggested that we need to look very carefully and
and then revised the target to 139 after the second inter- see whether in fact the numbers in our formula are totally
ruption. appropriate for the Twenty20 game.”[49]

10 Updates 11 Criticism
The published table that underpins the D/L method is The D/L method has been criticized on the grounds that
regularly updated, using source data from more recent wickets are a much more heavily weighted resource than
8 13 REFERENCES

overs, leading to the suggestion that if teams are chasing [10] Data Analysis Australia’s detailed mathematical analysis
big targets, and there is the prospect of rain, a winning of the Duckworth–Lewis Method daa.com.au.
strategy could be to not lose wickets and score at what
[11] Duckworth/Lewis Method of Re-calculating the Target
would seem to be a “losing” rate (e.g. if the required rate
Score in an Interrupted Match
was 6.1, it could be enough to score at 4.75 an over for
the first 20–25 overs).[50] [12] Tony Lewis, of Duckworth-Lewis, Interview: Journalists
denigrate system by publishing 'rubbish' without under-
Another criticism is that the D/L method does not ac-
standing 27 August 2013
count for changes in proportion of the innings for which
field restrictions are in place compared to a completed [13] espncricinfo D/L FAQ’s Q13
match.[51]
[14] espncricinfo D/L FAQ’s Q14
More common informal criticism from cricket fans and
journalists of the D/L method is that it is unduly complex [15] ICC Playing Handbook
and can be misunderstood.[52][53] For example, in a one-
[16] ICC Playing Handbook 2013/14 Section 6
day match against England on 20 March 2009, the West
Indies coach (John Dyson) called his players in for bad [17] espncricinfo D/L FAQ’s Q1
light, believing that his team would win by one run under
[18] espncricinfo D/L FAQ’s Q4
the D/L method, but not realizing that the loss of a wicket
with the last ball had altered the Duckworth–Lewis score. [19] The dummy’s guide to Duckworth-Lewis
In fact Javagal Srinath, the match referee, confirmed that
the West Indies were two runs short of their target, giving [20] The Duckworth-Lewis Method (2002) from Cricinfo.
the victory to England. [21] ICC Playing Handbook 2006-07
More recently, concerns have been raised as to its suit-
ability for Twenty20 matches, where a high scoring over [22] espncricinfo D/L FAQ’s Q6
can drastically alter the situation of the game and variabil- [23] espncricinfo D/L FAQ’s Q15
ity of the run-rate is higher over matches with a shorter
number of overs.[54] [24] Scorecard

[25] Scorecard

12 Cultural influence [26] Article

[27] Scorecard
"The Duckworth Lewis Method" is the name of a band
[28] Report
formed by Neil Hannon of The Divine Comedy and
Thomas Walsh of Pugwash, which recorded a self-titled [29] Scorecard
concept album of cricket songs.[55][56]
[30] Scorecard

[31] Scorecard
13 References [32] Scorecard

[1] “A Decade of Duckworth–Lewis”. BBC. 1 January 2007. [33] Scorecard

Retrieved 2009-03-21.
[34] Report
[2] “Introducing Duckworth-Lewis-Stern method”.
Cricbuzz. 12 February 2015. Retrieved 2015-03- [35] Report
30. [36] Over by over
[3] Scorecard for the rain-affected 4th ODI between India and [37] South Africa left to lick wounds (BBC)
England on 23 November 2008, from Cricinfo.
[38] Being Duckworth-Lewis: cricket’s weather-break mathe-
[4] Sri Lanka v England Scorecard maticians (The Guardian)
[5] Pakistan v India Scorecard, 6 Feb 2006 [39] SportTaco.com

[6] Scorecard [40] “22 off one ball – A farcical rain rule leaves everyone be-
wildered”, from Cricinfo.
[7] Report
[41] “Stump the Bearded Wonder”, Bill Frindall explains how
[8] Perth v Melbourne scorecard D/L would apply to 1992 WC semi-final

[9] Perth v Melbourne report [42] espncricinfo

 9

[43] BBC • BBC Sport’s explanation of the D/L method

[44] Scorecard of the 2nd ODI between England and Zim- • Web based Calculator for the Standard Edition of
babwe, 1 January 1997, from Cricinfo. the Duckworth Lewis method
[45] 3rd Final, 1988/89 Benson and Hedges World Series Cup • Calculator for the Stern Edition of the D/L method
[46] The Duckworth-Lewis Method (1999) from Cricinfo.
• Alternatives to D/L
[47] The Duckworth-Lewis Method (2001) from Cricinfo.

[48] Rain affected rules from Cricinfo.

[49] Duckworth–Lewis to review their formula for T20

matches

[50] Bhogle, Srinivas, The Duck worth/Lewis Factor, Red-

iff.com.

[51] Booth, Shane, quoted in For a Fair Formula, The Hindu.

[52] Varma, Amit, Simple and subjective? Or complex and

objective?, ESPNcricinfo

[53] Charlie Brooker, AV campaigners have created a stupidity

whirlpool that engulfs any loose molecules of logic, dis-
misses the claim of the simplicity by citing the method’s
formula, The Guardian, 25 April 2011. Retrieved 2011-
04-28

[54] The anomalous contraction of the Duckworth–Lewis

method

[55] BBC news interview with The Duckworth Lewis Method

[56] Interview with band

14 Further reading
• Duckworth, FC & Lewis, AJ “Your Comprehensive
Guide to The Duckworth Lewis Method for Reset-
ting Targets in One-day Cricket”, Acumen Books,
2004. ISBN 0-9548718-0-4
• Duckworth, F “A Role for Statistics in International
Cricket” Teaching Statistics, (June 2001) Volume 23,
No. 2 pp 38–44
• Duckworth, FC & Lewis, AJ “A fair method for
resetting the target in interrupted one-day cricket
matches” Journal of the Operational Research So-
ciety, (Mar 1998) Volume 49, No. 3 pp 220–227
JSTOR 3010471

15 External links
• ICC’s D/L method FAQ
• Cricinfo’s D/L method FAQ
• ICC’s D/L method (standard edition) table of re-
source percentages
• Cricinfo’s explanation of the D/L method
10 16 TEXT AND IMAGE SOURCES, CONTRIBUTORS, AND LICENSES

16 Text and image sources, contributors, and licenses

16.1 Text
• Duckworth–Lewis method Source: http://en.wikipedia.org/wiki/Duckworth%E2%80%93Lewis%20method?oldid=658338127 Contrib-
utors: The Anome, William Avery, Michael Hardy, Netsnipe, Selket, Dale Arnett, Shannonr, Dmmaus, Zzo38, Andycjp, RetiredUser2,
Ukexpat, Chmod007, Kate, Davidbod, Rackham, RAM, Sabretooth, Hesperian, Alansohn, JYolkowski, Arthena, MattDP, *Paul*, Wt-
mitchell, Omphaloscope, Pcpcpc, Novacatz, Thryduulf, Xerol~enwiki, Camw, A.K.A.47, Daniel Case, BillC, Tabletop, Mandarax, Gra-
ham87, Grammarbot, The wub, Tony2Times, CalJW, Chingwakabungya, Hairy Dude, Simon Lieschke, BlackJack, Muhammad sumair,
Miblo, Reyk, Red Jay, Boldra, SmackBot, Solaris smoke, Piccadilly, Newland44, Mauls, Cs-wolves, Rajeevmass~enwiki, Flagman-
tho, Dlohcierekim’s sock, Hongooi, Brianhenke, Flyguy649, Blake-, Ohconfucius, Timdownie, Catapult, Aspirex, Chrisd87, Enda.kelly,
Dacium, The-Pope, Tutunka, Phoenixrod, TheBoogler, Seedybob2, Viennese Waltz, KNM, Phósphoros, MrAngy, GavinTillman, Dumb-
BOT, Jono4174, Nompere, Araghu, Headbomb, Graemec2, Davidhorman, Farncombe, HussainAbbas, Trengarasu, Shirt58, Matford-
david, Myanw, Gökhan, Bigjimr, Responsible?, Manjuap, Henning M, Nimisis, Glennobrien, DrFrench, DoctorMartin, 88888, DanielEng,
Sraywood, Abhijitsathe, Veritywarne, STeamTraen, Nssdfdsfds, SatyrBot, Ktalon, Clarince63, Neerajbhojwani, Mannafredo, Gillyweed,
Jpeeling, SieBot, WereSpielChequers, BotMultichill, Srushe, Wilson44691, Phil Bridger, Superbeecat, Pinkadelica, DutchTreat, Church,
ClueBot, The White Duke, Myrrideon, DragonBot, PixelBot, Qwfp, Paulsummers65, Abomasnow, Osarius, Addbot, Alistair99, Fieldday-
sunday, Tassedethe, Numbo3-bot, NaFenn, Yobot, AnomieBOT, Piano non troppo, Flewis, Materialscientist, Capricorn42, Sylvsie35,
Armbrust, N419BH, Haldraper, BenzolBot, Lalit kohlimca, Amplitude101, DrilBot, Stikko, Full-date unlinking bot, Bgpaulus, ErikvanB,
Satdeep Gill, World8115, Onel5969, Aircorn, EmausBot, Immunize, Satish217 1990, Vickysb4u, RichardPryce, GeorgeBarnick, Efenna,
Subrata Roy, ClueBot NG, MelbourneStar, Skoot13, Rohan636, Widr, BG19bot, Misericord, Shibingeorge, Moaz786, Ayush pokharna,
Peoya, GyaroMaguus, Umang97g, Mmitchell10, Faizan, FrigidNinja, Jodosma, Hiteshstalreja, Dollarnaveen, Jianhui67, Asadhnd, 97rob
and Anonymous: 289

16.2 Images
• File:DuckworthLewisEng.png Source: http://upload.wikimedia.org/wikipedia/commons/8/82/DuckworthLewisEng.png License: CC-
BY-SA-3.0 Contributors: Transferred from en.wikipedia Original artist: Original uploader was Jono4174 at en.wikipedia

16.3 Content license

• Creative Commons Attribution-Share Alike 3.0