Champions Trophy 2013 June 10, 2013

Why net run rate doesn't work

The method of ranking teams which are level on points through NRR was designed to reward comprehensive wins, but it surely isn't working at the Champions Trophy

Over the last few of days, there have been a couple of close, low-scoring games in the Champions Trophy. Both have been won by the team chasing with plenty of overs to spare, but they've been nail-biters because in both instances the winning teams had to rely on their last couple of wickets to do the job. West Indies squeezed past Pakistan's 170 with 56 balls to spare but only two wickets in hand, while New Zealand were very nearly unsuccessful in their quest for 139 against Sri Lanka, winning only by one wicket even though they had 13.3 overs in hand. The other two results in each of the two groups - India beating South Africa in group B and England trouncing Australia in A - were clearly more convincing wins.

However, you wouldn't know that if you looked at the points table, for New Zealand are on top in group A and West Indies in B. That's because of the net run rates, which is the method used to break the deadlock if teams are level on points in multi-team tournaments. The NRR takes into account only the run rates of teams, and is calculated as the difference between the batting run rate of a team and the bowling economy rate over the entire tournament. A team which is all out is considered to have faced the full quota overs. However, in non-all-out situations, wickets lost isn't factored in at all. Thus, New Zealand benefit because the method considers the fact that they won with 81 balls to spare, but ignores the fact that they were nine down when they did so. Hence, their NRR of 1.048 is superior to England's 0.960. Even their most ardent supporter would admit, though, that England's win was far more convincing. Group B's scenario with West Indies and India is exactly the same. This method of ranking teams which are level on points was designed to reward comprehensive wins, but that surely isn't working here.

Limited-overs cricket is clearly a game where teams need to juggle with two sets of resources - overs and wickets in hand. Depletion of either of those resources, with respect to the target before the team, is a sign that the team's in trouble. Any method which is used to differentiate between teams on equal points should therefore consider both these factors when judging how comprehensive the victory was. The NRR method fails to do that. There have been other debates and arguments on the shortcomings of the NRR, but this is clearly the greatest one.

A possible solution here is to use the rain rule to decide the margin. In the case of the New Zealand-Sri Lanka match, the par score for New Zealand when nine down in 36.3 overs is 132. Since they won the match at this stage, they were seven runs ahead of the par score, which thus becomes the margin of victory. Since England's margin of victory over Australia was 48 runs, they would clearly be the group leaders. However, if New Zealand had won in the same number of overs for the loss of four wickets, then the margin by the D-L method would have been 52 runs.

Similarly, the margin of victory for West Indies over Pakistan would have been 20 runs by this method. Since India beat South Africa by 28, they'd have been the group B table-toppers at this stage. The victory margins in games which are won by the team batting first are anyway in terms of runs, so this option allows all match results to be expressed in terms of runs. It can be further argued that these margins should further take into account the target, so that a ten-run margin in a low-scoring game counts for more than the same margin in a high-scoring one.

As things stand in this tournament, there could be a scenario - however unlikely it seems at the moment - in which New Zealand and England finish with the same number of points, and are fighting for second place in the group. If New Zealand stay ahead of England on NRR based on their one-wicket victory in Cardiff, it'll surely be a travesty.

S Rajesh is stats editor of ESPNcricinfo. Follow him on Twitter

Comments have now been closed for this article

  • Dummy4 on June 17, 2013, 14:15 GMT

    Remove NRR, get head to head match ups to break deadlocks. Using DL to solve this is just more reason to get confused...

  • Sreejith on June 11, 2013, 15:49 GMT

    Calculate each team's Effective Run Rate (ERR) as Number of runs scored/Number of overs played (50 if all wickets lost)/Number of wickets lost. In NZ vs SL game, NZ's ERR (for) = 139/36.5/9 = 0.4231 (SL's ERR against) and SL's ERR (for) = 138/50/10 = 0.276 (NZ's ERR against). Now NZ's NRR = 0.4231 - 0.276 = 0.1471 and SL's NRR = -0.1471. Likewise, in the Eng vs Aus game, Eng's ERR (for) = 269/50/6 = 0.8967 (Aus's against) and Aus's ERR (for) = 221/50/9 = 0.4911 (Eng's against). So Eng's NRR = 0.8967 - 0.4911 = 0.4056 and Aus's NRR = -0.4056. Lets say in the next match (NZ vs Eng), NZ score 250/5 in 50 overs and Eng score 252/4 in 43.3 overs. Then NZ's new ERR for = 389/86.5/14 = 0.3212 and ERR against = 390/93.5/14 = 0.2979. So NZ's NRR = 0.0233, whereas Eng's ERR for = 521/93.5/10 = 0.5572 and ERR against = 471/100/14 = 0.3364. So Eng's NRR = 0.2208.

    This way, teams value their wickets and overs while both chasing or setting a total. NRR's also fair taking both wickets and overs lost

  • c on June 11, 2013, 14:04 GMT

    My response to @Bishop and other comments here that continue to engage in a meaningless effort to outsmart each other with rhetoric:

    Firstly, if it is called a Net Run Rate (NRR) - what is that? It is the RATE at which the batting side scored their runs off of a specific number of overs. Period! Forget about the REST! : )

    @Bishop: Yes, it IS POSSIBLE for a team to have a Plus NRR to the side it lost to - here in the case of SL losing to NZ. YES, if in the course of a tournament it so happens that a team only needs to focus on the NRR (the whole point of it being CONSIDERED A FACTOR in a tournament such as this Champions Trophy) then that is what they should do! This Trophy from the beginning had EVERYONE focusing on the NRR! - Why? As I said before, Go figure! And I say it with respect to you and others cos if there is one things humans are pretty good at it is COMPLICATING things for themselves and others!

    ICC is conducting a Trophy that defies logic. Not smart.

  • anil on June 11, 2013, 10:52 GMT

    The Goal Difference method used in football leagues around the globe is based on two important tenets that cricket has long overlooked. Firstly, it rewards teams that ENTERTAIN by scoring lots of goals. Secondly, its SIMPLISTIC enough for even a young fan to work out what his/her team needs to accomplish on the field. This makes for a gripping viewing experience for the entire duration of the game. I was entertained a lot more watching the NZ/SL & WI/PK clashes than the other so called 'comprehensive' wins by IN & AU. Today, the excitement of watching 4's & 6's has been significantly devalued, while the value in taking wickets remains least hit by the scourge of T20 cricket. So I would leave it to Ramesh and you whizzes to determine whether a system based on wickets taken and wickets conceded is feasible. The way limited overs cricket is headed, bowlers would soon prefer safer defensive non-wicket taking tactics like targeting the tram-lines than attempting yorkers. My 2cents

  • Dummy4 on June 11, 2013, 10:11 GMT

    I think every wicket should be valued as 5 overs in a run chase. When a team chases the total, the overs played should be num_wickets_lost*5 or actual overs played whichever is higher. So in the nz-sl case, nz would have got a nrr of 0.33 as follows 1. sl run rate: 138/50 = 2.76 2. nz run rate : 139/45 (9*5) = 3.03 3. nz net run rate: 3.03 - 2.76 = 0.33

    This way teams will value their wickets and run rate both.The wickets lost factor needs to changed based on number of overs played in a rain affected match. For ex, if the match is of 35 overs, the factor should be 3.5 (35/10).

    Also, this should not be called as net-run-rate. It can be called as net-average, as average is runs scored per wicket lost.

  • Dummy4 on June 11, 2013, 9:28 GMT

    The problem here is that some matches decided by runs and some by wickets and there is no direct relation between wickets & runs. If you look from Sri Lanka's point of view it is fair since they didn't bat entire 50 overs.

    Using D/L method will be tricky since D/L method does have many problems of its own. Instead of reducing it using D/L, it could be tried to project the score (as if the team chasing has played full 50 overs).

    Another alternative is, if 2 teams have same points, it should be decided on head-to-head record. Since in this tournaments, each team plays the other only once there is no issue of having a 1-1 lockout (as in when you have home-away system and each team wins 1).

  • Harmon on June 11, 2013, 9:28 GMT

    The article over-complicates a very simple concept. I agree that chasing 180 with 1 wicket in hand is no great chase, whether it is done in the 25th or in the 50th over but that is how the game of cricket is supposed to be played. The win or loss is defined in terms of the runs scored. The wicket parameter has never been considered except in some cases of tie or when the situation is too crowded in some table standings.

    So since that team is deemed to have won that scored more runs even if it chased 300 for 9 while the other team might have scored 299/1 how can we talk of wickets in NRR?

  • Dummy4 on June 11, 2013, 9:21 GMT

    A simpler system would be to just compare the true run rate. Runs given away divided by balls bowled and compare it to runs scored divided by balls faced. Simple and effective. I believe wickets lost should not be factored in. Wickets are a resource that the teams have to exploit and can and should use them as they please. All eleven players are expected to bat in cricket. Teams should not be penalised for using what they have available. And no D/L method, it has caused too much pain already in cricket. Why open it to up to cause more misery.

  • Ralf on June 11, 2013, 9:01 GMT

    I completly disagree! Let´s just avoid the pointless rocket science and consider the matter as such: 1.The team bowling first restricted the team batting to a considerably low total. 2. As a result, it is up to that chasing team, to dispose their ressources at will. 3. Due to their bowling efforts in the first innings, they gained the advantage to establish whatever tactics they want. 4. If it means to chase down the target quickly at the risk/expense of losing wickets - be it so! 5. If they chose to bat carefully, thus using the whole range of overs - be it so! 6. The most important thing is to reach the target, regardless of losing on wicket or nine. Remember: you have got 9 wickets in order to do so, there´s no need to save seven or eight of them. 7. Hence - the NRR-thing looks completely ok with me!

  • Dummy4 on June 11, 2013, 9:00 GMT

    I tend to disagree that Eng won more comprehensively. New zealand did a better job at restricting Sri Lanka at a low score and if you just concentrate on the run chase, you are considering on the Batting. You dont want to give credit to bowlers doing a good job? As somebody commented , Sri Lanka bowling out Zim for 38 runs was one of the most one-sided match.