How to calculate net run rate in cricket: the NRR formula and the four rules that break it

Net run rate is the run rate a team scores at across a tournament minus the run rate it concedes across the same tournament. Add up every run a side has scored and divide by the overs it faced, do the same for the runs it has given up and the overs it bowled, then subtract the second number from the first. That is the whole formula. Everything else people find confusing about NRR comes from four rules that change what goes into it.

The net run rate formula

NRR = (total runs scored / total overs faced) – (total runs conceded / total overs bowled)

The result is expressed in runs per over and carries a sign. A side scoring at 8.40 an over and conceding at 7.90 has an NRR of +0.50, meaning it has been half a run an over better than its opponents across the tournament. A negative figure means the reverse.

Two things about the formula catch people out straight away. It is calculated over the whole tournament, not per match, so a team does not have a net run rate for a single game. And it is a team figure only. Individual players do not have one, though a bowler’s economy rate is worked out the same way.

Run rate, required run rate and net run rate are three different things

Run rate is runs divided by overs in a single innings. A side on 90 after 15 overs is scoring at 6.00. Required run rate is what the chasing side still needs, taken from the runs left and the overs left. Net run rate is the tournament figure above, and it exists to separate teams level on points.

The plain run rate is also how a bowler’s economy is calculated: runs conceded divided by overs bowled. It is the same arithmetic pointed at one bowler instead of a team.

Overs are not decimals, and this is where most calculations go wrong

Cricket writes 18.5 to mean eighteen overs and five balls. Typing 18.5 into a calculator means eighteen and a half overs, which is eighteen overs and three balls. Every ball after the decimal point has to be converted to sixths before it goes into the formula.

Scorecard	Balls	Use this
.1	1	0.167
.2	2	0.333
.3	3	0.500
.4	4	0.667
.5	5	0.833

So a score of 107 in 17.2 overs is 107 divided by 17.333, which is 6.17 an over, not 107 divided by 17.2. The gap looks small on one innings and compounds badly across fourteen of them. Points tables use the same notation in their overs columns: a season total printed as 254.2 means 254 overs and two balls, or 254.333.

A worked example, two matches at a time

Take Punjab Kings’ first two wins of IPL 2026. On 19 April they made 254 in their 20 overs and held Lucknow Super Giants to 200 in theirs. On 25 April they chased down Delhi Capitals’ 264 for 2, reaching 265 in 18.5 overs, which remains the highest successful run chase in IPL history.

Column	Match 1	Match 2	Total
Runs scored	254	265	519
Overs faced	20	18.833	38.833
Runs conceded	200	264	464
Overs bowled	20	20	40

Scoring rate is 519 / 38.833 = 13.365. Concession rate is 464 / 40 = 11.600. Net run rate is 13.365 – 11.600 = +1.765.

Notice what the chase did. Punjab got there with seven balls to spare, finishing on 265 for 4, and those seven unused balls do real work. Had they taken the full 20 overs to reach the same 265, their figure would have been +1.375 instead of +1.765. Finishing early is how a chasing side moves its net run rate, which is why teams already through to the playoffs still bat as though the margin matters.

The four rules that break a hand calculation

Work through a season with nothing but the scorecards and the answer will usually be close to the official one without matching it. These are the reasons.

A side bowled out is charged its full quota. If a team is dismissed for 140 in 13.4 overs, the calculation still uses 20 overs, not 13.667. The rule exists so that being bowled out cheaply cannot flatter a team’s run rate. It is the single largest distortion in the formula and it is applied to both sides of the sum: the batting team is charged the full quota, and the bowling team is credited with having bowled them.

Matches with no result are excluded entirely. An abandoned game contributes nothing to either team’s runs or overs, even though it still shows in the matches-played column and still hands out a point each. Rule 16.10.2 of the ICC playing conditions is explicit that only matches in which a result is achieved count towards net run rate. Plenty of explainers, including an earlier version of this page, have it the other way round.

Matches decided by DLS use a par score, not the actual one. Where the Duckworth-Lewis-Stern method settles a match, the side batting first is credited with the par score the chasing side was set, which is its target minus one, over the chasing side’s revised allocation. The runs actually scored in that first innings do not go into the sum.

A shortened match uses the revised allocation. In a game cut to 15 overs a side, a team bowled out is charged 15, not 20.

We recalculated every IPL 2026 net run rate from the ball-by-ball data

To check the rules above actually behave as described, we rebuilt all ten net run rates for the IPL 2026 league stage from ball-by-ball records of the 70 matches, then compared them with the published table.

Team	Pts	Runs for	Overs	Runs against	Overs	NRR
Royal Challengers Bengaluru	18	2,642	254.2	2,619	272.4	+0.783
Gujarat Titans	18	2,558	270.4	2,434	278.0	+0.695
Sunrisers Hyderabad	18	2,854	276.1	2,675	272.4	+0.524
Rajasthan Royals	16	2,619	259.4	2,649	267.4	+0.189
Punjab Kings	15	2,711	250.0	2,718	258.0	+0.309
Delhi Capitals	14	2,559	272.4	2,581	257.1	-0.651
Kolkata Knight Riders	13	2,284	251.1	2,384	258.0	-0.147
Chennai Super Kings	12	2,530	275.0	2,512	263.1	-0.345
Mumbai Indians	8	2,553	268.4	2,636	261.2	-0.584
Lucknow Super Giants	8	2,514	275.3	2,616	265.1	-0.740

Eight of the ten reproduced to the digit on the first attempt. The two that did not were Lucknow Super Giants and Royal Challengers Bengaluru, and both were out by exactly three runs, in opposite directions.

They are the two sides in the only DLS result of the league stage. At the Ekana Stadium on 7 May, Lucknow made 209 in a match reduced to 19 overs and Bengaluru were set a revised 213, finishing on 203 to lose by nine. Because DLS decided it, Lucknow are credited with the par score of 212 rather than the 209 they actually scored, and Bengaluru are charged with conceding 212. Adding those three runs moves Lucknow from -0.751 to -0.740 and Bengaluru from +0.794 to +0.783, and both then match the published table.

Three runs in one match out of a season of 25,821, and it is the whole difference between a hand calculation and the official one.

Two checks you can run on any points table in ten seconds

Runs scored across a complete group must equal runs conceded across it, because every run one side scores is a run another concedes. In IPL 2026 both columns total 25,821. The overs columns must match for the same reason, and both come to 2,653.833. If a table’s two halves do not balance, a match is missing from it.

The second check is on the sign. Add up every team’s NRR in a complete group and the total will be close to zero but will not be exactly zero, because each team divides by its own overs rather than a common denominator. A group whose net run rates sum to something far from zero has a data problem.

The bowled-out rule was worth more than second place

Gujarat Titans bowled a side out four times in IPL 2026, more than any other team, and were bowled out once themselves. Chennai Super Kings for 140 in 13.4 overs, Sunrisers Hyderabad for 86 in 14.5 and Rajasthan Royals for 152 in 16.3 were each charged the full 20, which is exactly fifteen overs of credit Gujarat never had to bowl. Their fourth, Bengaluru for 155 in 19.2, added another four balls.

Recalculating the table with that rule switched off, and using only the overs actually bowled, changes Gujarat’s figure from +0.695 to +0.320 and Sunrisers Hyderabad’s from +0.524 to +0.577. All three of the sides on 18 points were separated by net run rate alone, so on the raw numbers Sunrisers rather than Gujarat would have taken second. Punjab Kings are the control case: they were neither bowled out nor bowled a side out all season, and their NRR is identical either way.

What net run rate cannot do

NRR only ever breaks a tie on points. It never overrides them. Delhi Capitals finished sixth in IPL 2026 on 14 points with a net run rate of -0.651, above Kolkata Knight Riders on 13 points with a far healthier -0.147. A season of better margins does not buy a single place if the wins are not there.

It is also not the only tie-break in use. Pro Kabaddi separates teams level on points by score difference, the raw sum of points scored minus points conceded, with no division by time played at all. Our guide to the kabaddi points table covers how that column behaves. Football’s goal difference works the same way.

And it says nothing about how a team got there. A side that wins six close games and loses four badly will sit below a side with the same record built the other way round, which is one reason net run rate is a better guide to the conditions a team has played in than to how well it has played.

Quick answers

What is the full form of NRR? Net run rate.

Can net run rate be negative? Yes. Any team conceding at a faster rate than it scores has a negative NRR, and in a ten-team league roughly half of them will.

What counts as a good NRR? In T20 leagues anything above about +0.500 is strong and above +1.000 is rare over a full season. No IPL 2026 side finished above +0.800.

Does the margin of victory matter more than the number of wins? Only after the points are level. Within that, yes: the overs left unused in a chase move the figure faster than the runs do.

Do the powerplay overs count differently? No. Every over counts the same in the formula, whatever the fielding restrictions. Our explainer on powerplay rules covers where those restrictions apply.

Lead photo: ANDREW CABALLERO-REYNOLDS / AFP

More cricket explainers and features