A case for Sonneborn-Berger score as tie-breaker in future World Cups

[Swashbuckler]

Now that the group stage is over I think it will be a good idea to focus on the most contentious issue which will have a direct bearing on who lifts the coveted trophy next Sunday.

India and Australia dominated this part of the tournament and earned their qualification quite early, no frills. Hosts were in a precarious situation but sneaked into the third spot by winning their last two matches against good teams. The 4th spot was sealed by New Zealand who finished on the same points as Pakistan but with a far better net run rate.

I won't berate NZ for their luck, nor do I have sympathy for Pakistan. The qualification rules were pre-decided and in this system Black Caps were more deserving to edge ahead of Pakistan. NZ played by the rules and smartly, Pakistan didn't give NRR much thought until late in the tournament and they have only themselves to blame. Now I will proceed to give my point of view about the tie break rule and attempt to put forth a better alternative.

The present rule when two teams are equal on points is to compare the NRR and the team with better NRR automatically advances. But is NRR really the most ideal way to decide such an important question?

This particular rule has a few flaws:

1. NRR doesn't take into account the wickets that fell. If there is a target of 250 in 50 overs and two teams go about the chase in different ways. Team A chases it down in 46 overs with the loss of two wickets while B finishes at 251/9 in 25 overs. Which team had a more dominating win? NRR rule says team B will score higher but anyone with common sense will choose team A. Team B was one wicket away from losing the match, maybe a close shave or umpiring error helped them cross the line. But A always had it under control, 8 wickets and 24 balls to spare. However as per rule book, A's NRR is +0.46 while for B it is a stupendous +5.04. Is this a fair representation? Wickets in ODIs hold less value than tests but should their importance be completely cast away?

2. NRR doesn't take into account the nature of pitches, ground dimensions, toss result, weather, scheduling etc. Conditions vary and it is impossible to assure standardized pitches and ground dimensions for all encounters throughout the world cup. In a tennis tournament, all players play on the same type (speed, bounce, friction) of court with same dimensions. Can we say the same about cricket? It will always be easier to gain NRR on one ground after a particular toss result. A match played on a flat, true wicket that stays same for the entire 100 overs duration, sunny day, 60 m boundaries vs a match where pitch drastically turns rogue after 40 overs, overcast conditions, 90 m boundaries: is it fair for the chasing team in second case? Do these things even out in such a short sample size (9 matches)? Likewise scheduling plays a role, NZ had an easy start and they boosted their NRR column, later against the tougher teams even though they lost they could play to limit damage to NRR. Pakistan had a tough start where the NRR suffered and later when the relatively easier fixtures came, it became more about winning than correcting the NRR.

3. Continuing from the previous point, NRR encourages negative cricket towards the end. Teams that have a headstart often tend to play safe to manage their NRR better. NZ had that advantage, they didn't even have to try and win against Australia and England. This is perfectly understandable from team's point of view but can make for boring cricket. By going hard for a victory there is the possibility of getting a whooping. But by waving the white flag at the beginning it is possible to have a loss with narrower margin. In such cases even the opponent team may relax, experiment because victory is more or less secured.

The popular suggestion is to decide qualification on the basis of head to head result. IMO it is unfair to make such an important decision based on just one match. When there are 9 matches in this stage why should everything hinge on that one particular encounter (Pak vs NZ in this edition) when most probably those two teams wouldn't have known about the importance of that clash. Of course all matches in this format are important but H2H makes it virtually a knockout match, making it more important than other matches. And teams wouldn't know because who can predict which teams will end up in such a scenario. Another problem here is what if there are 3 teams that end up equal and have a rock-paper-scissors kind of result in the group encounter (for instance WI joining Pak, NZ on 11 points)?

So I will try to present an idea to resolve the situation. There is something called Sonneborn–Berger score which is used to break ties in chess tournaments. It is computed by summing the score of each defeated opponent, and half the score of each drawn opponent and more the score better the tie break.The main point is to give more value for a win/draw against a player ranked high, than for a win/draw against a player ranked low in the tournament. Draws are common in chess just like test cricket (this system will work very well in future world test championships) but in ODIs replace the draw with tie/NR. That due to adverse weather/ground conditions, final points table and tiebreak system get messed up is another topic of discussion, hopefully with reserve days for all matches in March 2023 WC in India (when weather is absolutely fine) we don't have a single NR then. By smarter scheduling and reserve days ICC can set right this problem. The SB score is easy to calculate, even laymen will have no problem and more importantly it is a fair reflection of the performance of teams. Wins against top teams should get more value and that is taken care of in this system.

As an illustration, let me present the SB scores of NZ and Pakistan in this WC.

NZ defeated SL, Ban, Afg, WI, SA and the match against Ind was washed out. Only these 6 results will figure in calculation of SB tie-break score: full points of the defeated opponents and half of India because there the points were shared. For NZ it will be 8 (SL)+7 (SA)+7 (Ban)+5 (WI) +0 (Afg) +7.5 (Ind)=34.5

Pak defeated Eng, NZ, SA, Ban, Afg while the match against SL was NR. Pak's SB score will be: 12 (Eng)+11 (NZ)+7 (SA)+7 (Ban)+0 (Afg) +4 (SL)=41

As we can see Pak's SB score is comfortably higher, this despite NZ benefiting from that NR against Ind. NZ got 7.5 against Ind while Pak got 4 against SL. No system is perfect, as I wrote in a previous para it is incumbent on the ICC (and host country/board) to ensure NRs are eliminated in ICC tournaments. We won't have this problem in India 4 years from now so the SB scores will be perfect. No tie break points from beating Afg who finished on zero points, makes sense right? NZ getting punished for failing to win a single match against the top half, Pak getting rewarded for beating Eng and NZ.

These are the SB scores of all teams, conventional points in brackets:

Ind (15) 57.5 Q
Aus (14) 54 Q
Eng (12) 45 Q
NZ (11) 34.5 Q
Pak (11) 41
SL (8) 26
SA (7) 24.5
Ban (7) 16
WI (5) 14.5
Afg (0) 0

As we can see the SB scores correspond with final table standing except in case of Pak/NZ. I will reiterate that NZ deserved to qualify because they came out on top under present system of rules. But if you account for quality over quantity Pak was 4th best.

SB solves most of the problems we have under the present system:

1. Rewards quality of wins more than the net run rate. I will prefer a 4th team that defeated 2 of the semi-finalists in group stage over one that was more ruthless against the laggards but failed to open account against any of the potential SF/F opponents.

2. Wins take precedence over everything else. The teams just have to focus on winning. External factors like nature of pitch, toss result, ground dimensions, scheduling etc are less important. This system demands that you win, not by what margin. Teams instead of playing out of character (keeping NRR in mind) can play their best or most effective cricket to simply get over the line. No worries on the back of minds of chasing batsmen that they have to get the runs in XXX number of overs, none of that clutter.

3. No negative cricket. A team with headstart can't afford to play a tame ending in later stage, they have to go for it in all matches because losses don't count. Crowd will get its share of entertainment, as will viewers.

4. Intrigue right till the end and spectacle for neutrals, broadcasters, media etc. Teams' chances of qualification hinge on the performances of other teams and no side can afford to tank. For instance if applied in this tournament Pak fans would be invested in the success of Eng whom they beat, keeping in mind that NZ has nothing to gain if Eng wins. If the defeated opponents get higher score Pak's SB gets pushed up. Likewise NZ would be cheering WI on right till the end because that was an opponent they beat but whom Pak lost to. In this system of TB there won't be a dull moment, every match assumes importance. It isn't that simple because the final spots get confirmed only at the fag end and there are plenty of permutations and combinations. Nevertheless it will be more fun to follow, since the SB concept is so simple even casual fans can understand it in a jiffy.

What I have noticed in chess tournaments, the SB score is almost always good enough to break ties. But it is always advisable to have other TBs in case the tie can't be broken via this system. If two teams have the same points and also the same SB, it means they have defeated identical quality of opponents. Then we can use NRR, if it is still tied (highly unlikely) individual H2H and if the teams are still equal (0.000000000001% chance) the team that is rated higher (up to decimals) in ICC rankings.

 

[barah_admi]

It is very interesting that you should bring this up as me and my friends were discussing an alternative to NRR for a tie break. The SB method was brought up and although I have never watched a chess tournament in my life, this is a method I am familiar with. From the outset it looks useful, delve deeper and it seems better suited to a tie break than net run rate...however, the problem is that much like many other aspects of the game, there is the use of over complicated mathematics for a simple problem which requires a simple solution.

If there are two teams which are tied on points, then the one with the superior head to head record should go through. That has always been my position. Simple, effective, just.

 

[Napa]

Now that the group stage is over I think it will be a good idea to focus on the most contentious issue which will have a direct bearing on who lifts the coveted trophy next Sunday.

India and Australia dominated this part of the tournament and earned their qualification quite early, no frills. Hosts were in a precarious situation but sneaked into the third spot by winning their last two matches against good teams. The 4th spot was sealed by New Zealand who finished on the same points as Pakistan but with a far better net run rate.

I won't berate NZ for their luck, nor do I have sympathy for Pakistan. The qualification rules were pre-decided and in this system Black Caps were more deserving to edge ahead of Pakistan. NZ played by the rules and smartly, Pakistan didn't give NRR much thought until late in the tournament and they have only themselves to blame. Now I will proceed to give my point of view about the tie break rule and attempt to put forth a better alternative.

The present rule when two teams are equal on points is to compare the NRR and the team with better NRR automatically advances. But is NRR really the most ideal way to decide such an important question?

This particular rule has a few flaws:

1. NRR doesn't take into account the wickets that fell. If there is a target of 250 in 50 overs and two teams go about the chase in different ways. Team A chases it down in 46 overs with the loss of two wickets while B finishes at 251/9 in 25 overs. Which team had a more dominating win? NRR rule says team B will score higher but anyone with common sense will choose team A. Team B was one wicket away from losing the match, maybe a close shave or umpiring error helped them cross the line. But A always had it under control, 8 wickets and 24 balls to spare. However as per rule book, A's NRR is +0.46 while for B it is a stupendous +5.04. Is this a fair representation? Wickets in ODIs hold less value than tests but should their importance be completely cast away?

2. NRR doesn't take into account the nature of pitches, ground dimensions, toss result, weather, scheduling etc. Conditions vary and it is impossible to assure standardized pitches and ground dimensions for all encounters throughout the world cup. In a tennis tournament, all players play on the same type (speed, bounce, friction) of court with same dimensions. Can we say the same about cricket? It will always be easier to gain NRR on one ground after a particular toss result. A match played on a flat, true wicket that stays same for the entire 100 overs duration, sunny day, 60 m boundaries vs a match where pitch drastically turns rogue after 40 overs, overcast conditions, 90 m boundaries: is it fair for the chasing team in second case? Do these things even out in such a short sample size (9 matches)? Likewise scheduling plays a role, NZ had an easy start and they boosted their NRR column, later against the tougher teams even though they lost they could play to limit damage to NRR. Pakistan had a tough start where the NRR suffered and later when the relatively easier fixtures came, it became more about winning than correcting the NRR.

3. Continuing from the previous point, NRR encourages negative cricket towards the end. Teams that have a headstart often tend to play safe to manage their NRR better. NZ had that advantage, they didn't even have to try and win against Australia and England. This is perfectly understandable from team's point of view but can make for boring cricket. By going hard for a victory there is the possibility of getting a whooping. But by waving the white flag at the beginning it is possible to have a loss with narrower margin. In such cases even the opponent team may relax, experiment because victory is more or less secured.

The popular suggestion is to decide qualification on the basis of head to head result. IMO it is unfair to make such an important decision based on just one match. When there are 9 matches in this stage why should everything hinge on that one particular encounter (Pak vs NZ in this edition) when most probably those two teams wouldn't have known about the importance of that clash. Of course all matches in this format are important but H2H makes it virtually a knockout match, making it more important than other matches. And teams wouldn't know because who can predict which teams will end up in such a scenario. Another problem here is what if there are 3 teams that end up equal and have a rock-paper-scissors kind of result in the group encounter (for instance WI joining Pak, NZ on 11 points)?

So I will try to present an idea to resolve the situation. There is something called Sonneborn–Berger score which is used to break ties in chess tournaments. It is computed by summing the score of each defeated opponent, and half the score of each drawn opponent and more the score better the tie break.The main point is to give more value for a win/draw against a player ranked high, than for a win/draw against a player ranked low in the tournament. Draws are common in chess just like test cricket (this system will work very well in future world test championships) but in ODIs replace the draw with tie/NR. That due to adverse weather/ground conditions, final points table and tiebreak system get messed up is another topic of discussion, hopefully with reserve days for all matches in March 2023 WC in India (when weather is absolutely fine) we don't have a single NR then. By smarter scheduling and reserve days ICC can set right this problem. The SB score is easy to calculate, even laymen will have no problem and more importantly it is a fair reflection of the performance of teams. Wins against top teams should get more value and that is taken care of in this system.

As an illustration, let me present the SB scores of NZ and Pakistan in this WC.

NZ defeated SL, Ban, Afg, WI, SA and the match against Ind was washed out. Only these 6 results will figure in calculation of SB tie-break score: full points of the defeated opponents and half of India because there the points were shared. For NZ it will be 8 (SL)+7 (SA)+7 (Ban)+5 (WI) +0 (Afg) +7.5 (Ind)=34.5

Pak defeated Eng, NZ, SA, Ban, Afg while the match against SL was NR. Pak's SB score will be: 12 (Eng)+11 (NZ)+7 (SA)+7 (Ban)+0 (Afg) +4 (SL)=41

As we can see Pak's SB score is comfortably higher, this despite NZ benefiting from that NR against Ind. NZ got 7.5 against Ind while Pak got 4 against SL. No system is perfect, as I wrote in a previous para it is incumbent on the ICC (and host country/board) to ensure NRs are eliminated in ICC tournaments. We won't have this problem in India 4 years from now so the SB scores will be perfect. No tie break points from beating Afg who finished on zero points, makes sense right? NZ getting punished for failing to win a single match against the top half, Pak getting rewarded for beating Eng and NZ.

These are the SB scores of all teams, conventional points in brackets:

Ind (15) 57.5 Q
Aus (14) 54 Q
Eng (12) 45 Q
NZ (11) 34.5 Q
Pak (11) 41
SL (8) 26
SA (7) 24.5
Ban (7) 16
WI (5) 14.5
Afg (0) 0

As we can see the SB scores correspond with final table standing except in case of Pak/NZ. I will reiterate that NZ deserved to qualify because they came out on top under present system of rules. But if you account for quality over quantity Pak was 4th best.

SB solves most of the problems we have under the present system:

1. Rewards quality of wins more than the net run rate. I will prefer a 4th team that defeated 2 of the semi-finalists in group stage over one that was more ruthless against the laggards but failed to open account against any of the potential SF/F opponents.

2. Wins take precedence over everything else. The teams just have to focus on winning. External factors like nature of pitch, toss result, ground dimensions, scheduling etc are less important. This system demands that you win, not by what margin. Teams instead of playing out of character (keeping NRR in mind) can play their best or most effective cricket to simply get over the line. No worries on the back of minds of chasing batsmen that they have to get the runs in XXX number of overs, none of that clutter.

3. No negative cricket. A team with headstart can't afford to play a tame ending in later stage, they have to go for it in all matches because losses don't count. Crowd will get its share of entertainment, as will viewers.

4. Intrigue right till the end and spectacle for neutrals, broadcasters, media etc. Teams' chances of qualification hinge on the performances of other teams and no side can afford to tank. For instance if applied in this tournament Pak fans would be invested in the success of Eng whom they beat, keeping in mind that NZ has nothing to gain if Eng wins. If the defeated opponents get higher score Pak's SB gets pushed up. Likewise NZ would be cheering WI on right till the end because that was an opponent they beat but whom Pak lost to. In this system of TB there won't be a dull moment, every match assumes importance. It isn't that simple because the final spots get confirmed only at the fag end and there are plenty of permutations and combinations. Nevertheless it will be more fun to follow, since the SB concept is so simple even casual fans can understand it in a jiffy.

What I have noticed in chess tournaments, the SB score is almost always good enough to break ties. But it is always advisable to have other TBs in case the tie can't be broken via this system. If two teams have the same points and also the same SB, it means they have defeated identical quality of opponents. Then we can use NRR, if it is still tied (highly unlikely) individual H2H and if the teams are still equal (0.000000000001% chance) the team that is rated higher (up to decimals) in ICC rankings.

Trust you to dig up a particular method of scoring that produces a desired outcome.

The problem with SB score is that it tends to discount losses against weaker teams, and there is no reasonable argument for doing that. Specifically, it would place a team that beat the #1 team and lost to the #8 team would be placed ahead of the team that lost to #1 and beat #8. Losing to the #8 team is pretty awful and shouldn't be excused as the S-B score would.

People need to stop all this whining after the group stage is over. If there was a problem with the NRR system, the complaining should have happened before the group stage began. All this complaining now comes across as a loser mentality.

 

[DeadBall]

It is very interesting that you should bring this up as me and my friends were discussing an alternative to NRR for a tie break. The SB method was brought up and although I have never watched a chess tournament in my life, this is a method I am familiar with. From the outset it looks useful, delve deeper and it seems better suited to a tie break than net run rate...however, the problem is that much like many other aspects of the game, there is the use of over complicated mathematics for a simple problem which requires a simple solution.

If there are two teams which are tied on points, then the one with the superior head to head record should go through. That has always been my position. Simple, effective, just.

If its head to head record in the tournament then yes. Bilateral rankings shouldn't matter in a knock out tournament, especially in a round Robin format because the main thing that differentiates the champions from the rest is the ability to perform under pressure.

They used to have h2h before and then if it was a tie they would use NRR, don't know why they changed it.

 

[Arsalan Pro]

Excellent post, I think ICC used a close to this format in 2003 World Cup, where they gave carried forward points to teams for the 2nd Group Stage because they won't play them again, but double points to the teams who beat a qualified team compared to a eliminated team.

 

[Swashbuckler]

It is very interesting that you should bring this up as me and my friends were discussing an alternative to NRR for a tie break. The SB method was brought up and although I have never watched a chess tournament in my life, this is a method I am familiar with. From the outset it looks useful, delve deeper and it seems better suited to a tie break than net run rate...however, the problem is that much like many other aspects of the game, there is the use of over complicated mathematics for a simple problem which requires a simple solution.

If there are two teams which are tied on points, then the one with the superior head to head record should go through. That has always been my position. Simple, effective, just.

But SB is a very simple method, basic addition. Full points of defeated opponents, half points of opponents against whom you tied or had NR. With just 10 teams it will take less than a couple of minutes to get the SB scores of all teams.

NRR involves more calculation, something like Duckwoth-Lewis I have never even attempted to understand it . Compared to all that SB is a cakewalk and easily understandable for lay people.

I have addressed the H2H TB in one of the paras (just before introducing SB, sorry for the lengthy post). Do you agree with that?

 

[Swashbuckler]

Trust you to dig up a particular method of scoring that produces a desired outcome.

What desired outcome? Do you know me? I certainly don't remember interacting with you before.

SB system is used in chess tournaments which I follow, didn't come out of thin air. I wanted to make the post midway through the tournament but didn't have time, knew the post would be super lengthy and moreover I was too busy to spend much time here.

The problem with SB score is that it tends to discount losses against weaker teams, and there is no reasonable argument for doing that. Specifically, it would place a team that beat the #1 team and lost to the #8 team would be placed ahead of the team that lost to #1 and beat #8. Losing to the #8 team is pretty awful and shouldn't be excused as the S-B score would. .

NRR has many flaws (addressed in my first post), more than SB. No system is perfect but it is desirable to have the least flawed system in place.

Losing to #8 is already punished in the final points table, just like how beating #1 is rewarded. SB isn't to decide final standings, merely to break a tie. Losing to #8 is bad but #8 won't be in the semis. Beating #1 means the team can hang in with the top team, hence more intrigue for the knockout stage. WC is about crowning champions, it is pardonable if teams lose to bottom seeds but win the tournament than the other way round. I want a team that can win it all, not a steady Eddy that loses to top half and wins against bottom half. That isn't the trademark of champions, minnow bashing never is.

Ok so you have identified one problem in SB (logical counter provided), never mind that there are multiple problems in NRR. How to set those right?

People need to stop all this whining after the group stage is over. If there was a problem with the NRR system, the complaining should have happened before the group stage began. All this complaining now comes across as a loser mentality

Nobody is whining here. This is my first cricket related post (over one sentence) here in a while and as I made it clear (more than once) in my post NZ deserved to qualify and Pak deserved to miss out the SF spot. None of the others on this thread are whining. This is a cricket forum and people can have mature, civil discussions about these matters especially when there is no match thread. And this is solely about coming up with a better TB system for the future, how is that complaining? So now fans can't make posts about how to make better rules for 2023 WC, not like ICC will be scanning these posts and implementing the same. If we can't discuss cricket on a cricket forum, what should we do?

 

[Napa]

What desired outcome? Do you know me? I certainly don't remember interacting with you before.

I grouped you together with a hundred other posters complaining about NRR, Dhoni not scoring at 14 an over, washed out game against SL yada yada yada... sorry about that.

SB system is used in chess tournaments

Chess scoring is hardly a model for the rest of the world. Carlsen advanced in the Candidates because he had a larger number of losses. Figure that out)

Beating #1 means the team can hang in with the top team

I would rather that the team that lost to #8 just go home and think about how to improve.

 

[trickyone]

Potw...

 

[Swashbuckler]

I grouped you together with a hundred other posters complaining about NRR, Dhoni not scoring at 14 an over, washed out game against SL yada yada yada... sorry about that.

Ok fine, I am not up to date with recent forum activity.

Chess scoring is hardly a model for the rest of the world. Carlsen advanced in the Candidates because he had a larger number of losses. Figure that out)

And it came in for a lot of criticism as well. They were using 'more number of wins' TB in that event, weird but pre-decided hence fair albeit imperfect. Carlsen edged out Big Vlad fair and square there just like how NZ deservingly got the 4th spot. But focus is not on chess here, it is about the SB system and whether it can be incorporated in future ICC WCs. Whether it is the best system available and the flaws of NRR, H2H etc. Focus should be on the future and not the past, professional sports should evolve and rules should get watertight.

 

[Napa]

Carlsen edged out Big Vlad fair and square there just like how NZ deservingly got the 4th spot.

It was actually not "fair and square" but very controversial. Many people are convinced that Radjabov threw his game against Carlsen to down the Russian. Certainly 48.Rxb7+ makes no sense, it gives up Black's Rook which can create all kinds of threats from the back rank from where it cannot be displaced. 48.Ra8 was a dead draw. Even I can see that.

https://www.chessbomb.com/arena/2013-candidates/13-Teimour_Radjabov-Magnus_Carlsen

Anyway, the point is that chess is not a model for cricket to follow.

 

[Pakpak]

Very good post mate. I've always thought NRR isn't that good especially due to the different weather conditions etc. I thought the same when Australia pipped South Africa in 1999.

 

[Slog]

Trust you to dig up a particular method of scoring that produces a desired outcome.

The problem with SB score is that it tends to discount losses against weaker teams, and there is no reasonable argument for doing that. Specifically, it would place a team that beat the #1 team and lost to the #8 team would be placed ahead of the team that lost to #1 and beat #8. Losing to the #8 team is pretty awful and shouldn't be excused as the S-B score would.

People need to stop all this whining after the group stage is over. If there was a problem with the NRR system, the complaining should have happened before the group stage began. All this complaining now comes across as a loser mentality.

OP is Indian

 

[Swashbuckler]

Very good post mate. I've always thought NRR isn't that good especially due to the different weather conditions etc. I thought the same when Australia pipped South Africa in 1999.

Exactly man. Getting high NRR isn't as straightforward as people say, luck plays a bigger role than it should.

 

[Swashbuckler]

It was actually not "fair and square" but very controversial. Many people are convinced that Radjabov threw his game against Carlsen to down the Russian. Certainly 48.Rxb7+ makes no sense, it gives up Black's Rook which can create all kinds of threats from the back rank from where it cannot be displaced. 48.Ra8 was a dead draw. Even I can see that.

https://www.chessbomb.com/arena/2013-candidates/13-Teimour_Radjabov-Magnus_Carlsen

Anyway, the point is that chess is not a model for cricket to follow.

Spoiler

That is a random speculation. Radjabov was in terrible form in that event and finished dead last. There were many flaws in the organization of that tournament, most important being not employing the SB score for TB. IIRC there was a switch in time control just for that event, from increment after every move to no increment (a departure that time). Don't remember all details but Ivanchuk, Radjabov, Grischuk faced lots of time trouble. Grischuk faces that in all time controls but the other two had torrid time.

Carlsen won fair and square, bottomline. Blame the rules but not the players.

Many moves make no sense, GMs too blunder especially in time trouble. They don't have the comforts of engine evals or stress free environment. Some world champs have missed mate in one positions, walked into beginner forks, lost pieces, there have been so many famous blunders. It is wrong to throw mud on players without proof. The 48th move in this particular instance is not a blunder (but a bad inaccuracy, he should have kept that active rook), though Radja should have saved the game. I have seen Carlsen squeeze such positions against many elite GMs, much greater than this opponent. 2013 Carlsen had that aura and would force opponents to go to passive defence and then relentless torture.

I never said cricket has to follow chess. I talked about SB which is a simple mathematical TB system used in chess. The Candidates you are talking about didn't use SB and was rightly condemned by all including Kasparov. Why would you bring a tournament again and again which never used the system we are discussing here? And enough of chess , this thread is about an entirely different matter. Unfair on fellow posters to discuss other matters (especially 'boring' chess), ok in time pass but not here.

 

[Napa]

Spoiler

That is a random speculation. Radjabov was in terrible form in that event and finished dead last.

Spoiler

The question about Radjabov is due to his saying something to the effect "the right person won the tournament".

 

[Sin Nombre]

SB is not even used as the first tiebreaker in chess in closed tournaments. This is the tiebreak rules used in the Grand Chess tour currently.

a. Number of games won by each of the players in the tie.

b. The results of the games between or amongst the players in the tie.

c. The Sonneborn-Berger System

d. The Extended Koya System, as follows:

i. the number of points scored against players who have scored 50% or more

ii. the inclusion, step by step, of each score group below 50% in turn.

e. If a tie still remains unresolved after the application of Regulation 4 d ii, the Grand Chess Tour Points shall be shared.


Chess in my opinion should use TPR (tournament performance rating) as the tiebreaker for open tournaments. SB has lots of issues especially in opens.

 

[The Viper]

The flaw with nrr is the scheduling of the games and the pitches which can give an advantage to a team. In this case NZ.

NZ had favorable games at the start of the tournament. Taking nothing away, they did well to trash Afghanistan and Sri Lanka. The problem is that they had the perfect conditions to dismiss both teams quickly. The pitch for Afghanistan vs Pak was much slower and it meant that nrr could not be made up. Nrr does not take conditions into consideration. Especially, considering conditions got tougher as the tournament went on batting second.

Pakistan had to play tough games first- big three inside the first five games. When you get beaten comprehensively, you get demoralised also which effects approach in upcoming games. Pakistan just focused on winning games rather than worrying about nrr.

This system is the way to go because quality of wins should be valued more imo. Pakistan beat England which was a big scalp. The West Indies loss was the only somewhat unexpected loss. NZ got away with a tie with India.

 

[Captain caveman]

OP brings up some good points.
But i've sorted this problem out!

1.All matches MUST be played!
Set aside 3 days for washed out games to be played. This can be done between the group stage and knock out stage, as there is usually, at least 3 days between these stages, hence no added days to the length of the tournament!

2. If there are teams tied on points for qualification to the next stage, then the top 2 teams (in case there are more than 2 tied teams) on nrr play an eliminator for qualification to the next stage! This too can take in the 3 day break between stages!

And thats it. Not perfect, if more than 2 teams tied, but the best possible solution in my honest opinion and much fairer than the current system or h2h!

 

[Swashbuckler]

The question about Radjabov is due to his saying something to the effect "the right person won the tournament".

Spoiler

Not enough to call the Candidates controversial or that Radja threw the game, don't you think? What he allegedly said (language depends, most non-Western players struggle to communicate effectively in English) was common view, the chess world (including many Russians) wanted to see the rise of Carlsen. He dominated the chess circuit in 2012/2013 and people believed it was time for him to take the next step. Both him and Aronian were seen as possible challengers but western lobby backed the Norwegian, possibility of first Western champ since Fischer 1972. Radjabov would have probably said the same about Aronian but the Armenian was having a meltdown after a flying start, moreover Azerbaijan-Armenia relations mirror India-Pakistan and those two avoid making public comments, top authority is involved.

SB is not even used as the first tiebreaker in chess in closed tournaments. This is the tiebreak rules used in the Grand Chess tour currently.

a. Number of games won by each of the players in the tie.

b. The results of the games between or amongst the players in the tie.

c. The Sonneborn-Berger System

Closed events in chess and this WC format have a great difference. In cricket world cup more than 90% matches will have results even if weather goes completely haywire, nature of ODI cricket. But GM chess is mostly about draws, when the very best are involved this becomes a bigger problem. To counter the threat of draws the organizers try to incentivize the players to get more victories, likewise preventing short draws by Sofia rules (no draw offer till move 30/40).

There isn't an inherent fault in the SB system. Just that everyone (fans, organizers, broadcasters) wants more blood to be spilt on the boards, especially when the top dogs play each other. This is a commercial decision because of fault in top GM chess rather than anything else.

Chess in my opinion should use TPR (tournament performance rating) as the tiebreaker for open tournaments. SB has lots of issues especially in opens.

Olympiad is the most prestigious open and they use SB as 1st TB. It can get pretty comical because different matches get decided at different times and the final result might depend on who is winning the 3rd board between Peru and Uganda. If you follow the final round of Olympiads you will see the funny side of it. The Russians and Chinese might crowd around the laggard table and silently cheer on 2 unknowns.

I agree TPR should be used as TB in opens.

Anyway I don't want to compare chess with cricket. Just to borrow a system chess uses (or popularized by) and modify it slightly (tie/NR in place of draw) to apply to cricket. We should look to adopt good things from other sports/games if it results in cricket becoming a better, fairer, more logical, entertaining sport.

 

[latecut]

I do not see why one should have such long and convoluted discussion on NRR just because Pakistani team had lower NRR than New Zealand. Is it the first time that a team has missed out on qualification based on NRR ? I am sure there were such cases in the past as well. But I do not find them complaining and whingeing this much.

For instance, I remember South Africa missed out on a semi final place in 2003 world cup by just one run in their match against Sri Lanka because of their miscalculation/ misunderstanding on Duckworth Lewis method. But South Africans did not whinge like this.

 

[Slog]

I do not see why one should have such long and convoluted discussion on NRR just because Pakistani team had lower NRR than New Zealand. Is it the first time that a team has missed out on qualification based on NRR ? I am sure there were such cases in the past as well. But I do not find them complaining and whingeing this much.

For instance, I remember South Africa missed out on a semi final place in 2003 world cup by just one run in their match against Sri Lanka because of their miscalculation/ misunderstanding on Duckworth Lewis method. But South Africans did not whinge like this.

South Africans missed out on Super 6’s not semi finals. Atleast get history right.

 

[latecut]

South Africans missed out on Super 6’s not semi finals. Atleast get history right.

OK. I stand corrected. But this example demonstrates the fact that teams have missed out on qualifications narrowly in the past as well and they have moved on.

 

[Swashbuckler]

OK. I stand corrected. But this example demonstrates the fact that teams have missed out on qualifications narrowly in the past as well and they have moved on.

But if there is a flaw in the system shouldn't there be attempts to move on to a different one? This isn't about India, Pakistan, SA or whingeing, this is about the betterment of the sport, about rules evolving into more logical and watertight versions. Change is constant and necessary, otherwise we would still be playing underarm cricket in whites. After that Eng-SA SF in 1992 WC, DLS method came in. By your argument nothing should have been done back then and the same set of rules should have been persisted with?

 

[Itachi]

This breaks the core concept of odi. If same logic is applied, runs should be added to batting first for incoming wickets with addition to the ones that the team scored in 50 overs.

Cricket isn't chess. In chess, a developed piece will be the hardest one to take and you may sacrifice others in order to save it. In cricket, you can't.

Different game, different Dynamics. Just because it works in one doesn't mean it will work the same way in other.

For that too happen, every cricket match has to be similar in all aspects with that of a chess game.

 

[Swashbuckler]

This breaks the core concept of odi. If same logic is applied, runs should be added to batting first for incoming wickets with addition to the ones that the team scored in 50 overs.

I am not getting what you are saying. In SB it is just about wins and losses, margins don't matter. Doesn't matter how you win, just win. How you optimize your resources, that is totally up to you. Whether you win by 200 runs or 1 wicket on final delivery of the innings, makes no difference in TB. But if you beat top sides you get extra credit when it comes to breaking ties.

Cricket isn't chess. In chess, a developed piece will be the hardest one to take and you may sacrifice others in order to save it. In cricket, you can't.

Different game, different Dynamics. Just because it works in one doesn't mean it will work the same way in other.

For that too happen, every cricket match has to be similar in all aspects with that of a chess game.

No one made the comparison. Topic is SB, do critique the SB method if you want, do you find flaws in it compared to NRR? SB doesn't depend on game dynamics, it is just a basic mathematical system that can be applied in any round robin sport.

 

[Itachi]

I am not getting what you are saying. In SB it is just about wins and losses, margins don't matter. Doesn't matter how you win, just win. How you optimize your resources, that is totally up to you. Whether you win by 200 runs or 1 wicket on final delivery of the innings, makes no difference in TB. But if you beat top sides you get extra credit when it comes to breaking ties.



No one made the comparison. Topic is SB, do critique the SB method if you want, do you find flaws in it compared to NRR? SB doesn't depend on game dynamics, it is just a basic mathematical system that can be applied in any round robin sport.

Let's get it clear. You pointed out the flaws of NRR. And then you proceeded with a system that doesn't address those weakness in the first place?

Even this system suffers the same as NRR. A team who already qualified could experiment in the last few matches. In that case, the deviation will be much higher. A team could go through because it had a schedule of match against top team near the end of the league which will be very different from the other team which had in the beginning.

 

[Swashbuckler]

If the explanation/illustration in OP wasn't clear:


SB has nothing to do with chess, it can be applied to any sport in round robin format, be it cricket, basketball, volleyball or badminton. Enough with the chess analogy already.

Example: There are 4 tennis players: A, B, C and D who play a double round-robin tournament. Each player has to play two matches against each other player. 1 indicates a win, 0 a loss.



Player A and B both have 4 wins, thus the Sonneborn-Berger method is used to determine the tournament winner: For each won game the opponent's final primary score is added to the winner's Sonneborn-Berger score. For example, A won against B, C, C and D, his Sonneborn-Berger score is 4+3+3+1 = 11. B has a Sonneborn-Berger score of 9 (4+3+1+1), thus A is the tournament winner.

 

[mon858]

I think overall nrr is pretty fair. The bigger issue that can be more unfair is rain affected washouts. There should be some provision for replacing washed out games near end of tournament maybe 2 games each over 3 days. Or maybe granting the game with half points of a win to team ranked higher in tournament or based on icc ranking entering world cup. Something that is better than giving 1 pt each to both teams as that os then pure luck and you get 1 pt for washout against Afghanistan or Australia india just based on luck.

 

[Swashbuckler]

Let's get it clear. You pointed out the flaws of NRR. And then you proceeded with a system that doesn't address those weakness in the first place?

There will never be a flawless system, just a less flawed one. SB does address some of the weaknesses (not claiming it is the most perfect) and have included them in OP.

Even this system suffers the same as NRR. A team who already qualified could experiment in the last few matches. In that case, the deviation will be much higher. A team could go through because it had a schedule of match against top team near the end of the league which will be very different from the other team which had in the beginning.

Qualified teams may experiment but they wouldn't want to lose momentum just before the knockouts. History isn't kind towards teams that lost group matches immediately before the SF. On 16 occasions a team has come into the knockout stage of the World Cup off the back of a loss – and none of those sides have ever won the World Cup (https://www.nzherald.co.nz/cricket/news/article.cfm?c_id=29&objectid=12247332). Needless to say they will give it their all and attempt to win.

In fact SB will take care of scheduling disparities, only wins matter. Teams needn't bother about boosting their NRR in the end (this edition: Pak) or trying to stop the bleeding against top teams (NZ against Aus, Eng).

 

[Napa]

There will never be a flawless system, just a less flawed one. SB does address some of the weaknesses (not claiming it is the most perfect) and have included them in OP.

One reason why NRR is superior to SB is that SB (unlike NRR) would treat a 100 run defeat the same as a last ball defeat. All defeats are not equal in cricket.

 

[Captain caveman]

One reason why NRR is superior to SB is that SB (unlike NRR) would treat a 100 run defeat the same as a last ball defeat. All defeats are not equal in cricket.

All pitches in cricket are not the same, means nrr is irrelevant!
All cricket grounds and their boundaries in cricket are not the same, means nrr is irrelevant!
The weather in every match in cricket is not the same, means nrr is irrelevant!

 

[Napa]

All pitches in cricket are not the same, means nrr is irrelevant!
All cricket grounds and their boundaries in cricket are not the same, means nrr is irrelevant!
The weather in every match in cricket is not the same, means nrr is irrelevant!

Your bar for irrelevance is too low, some sort of caveman logic.

 

[Captain caveman]

Your bar for irrelevance is too low, some sort of caveman logic.

You don't seem to have grasped the fact that ALL factors are not the same in every match and deciding to elect a single factor i.e. run rate to differentiate between matches is completely flawed!!

 

[bujhee kom]

Sour grapes!

 

[Napa]

All pitches in cricket are not the same, means nrr is irrelevant!
All cricket grounds and their boundaries in cricket are not the same, means nrr is irrelevant!
The weather in every match in cricket is not the same, means nrr is irrelevant!

You don't seem to have grasped the fact that ALL factors are not the same in every match and deciding to elect a single factor i.e. run rate to differentiate between matches is completely flawed!!

Okay, let me explain it to you.

Understand that it is NET Run Rate, so something has to favor just one team for there to be a bias.

1) "All pitches in cricket are not the same". If the pitch favors run scoring for both sides, then the NET rate won't be biased.
If it favors run-scoring for only one side (say the team batting first), then S-B will also be biased as the team batting first is more likely to win.

Same is true for weather.

2) "All cricket grounds and their boundaries in cricket are not the same". If you have a ground with short boundaries then it will favor both sides when it comes to run scoring. So NET Run Rate will not be biased.

 

[barah_admi]

But SB is a very simple method, basic addition. Full points of defeated opponents, half points of opponents against whom you tied or had NR. With just 10 teams it will take less than a couple of minutes to get the SB scores of all teams.

NRR involves more calculation, something like Duckwoth-Lewis I have never even attempted to understand it . Compared to all that SB is a cakewalk and easily understandable for lay people.

I have addressed the H2H TB in one of the paras (just before introducing SB, sorry for the lengthy post). Do you agree with that?

I just read your paragraph on H2H. I disagree that it is unfair, or any less fair than the use of NRR or SB depending upon team situations, for example, one of the problems you highlighted regarding NRR is that it does not take into account pitches and match situations; for SB, there can be the argument as to why ICC Rankings, which are based on bilateral series should matter in a tournament? and so on.

All methods are flawed in one way or another, so my solution to any such problems is always the same - the simplest method should be used. In this case it is H2H, now if there was a situation where 3 teams are all on equal points and have a "rock papers scissors" issue, then we can go to a second method of breaking the tie. In this instance it should be SB, which I believe is better than NRR.

But the first method should always be H2H.

 

[warcry]

why not have a tie breaker match between two teams level on points. (like in tennis)
it would be like a quarter final and would generate interest.
if 3 or more teams are tied on points, just arrange the team with respect to ranking. this will mean that rankings actually matter.

 

[Swashbuckler]

for SB, there can be the argument as to why ICC Rankings, which are based on bilateral series should matter in a tournament?

I think you are mistaken, SB doesn't take into account ICC rankings. It takes into account the final points table of that tournament only.

 

[barah_admi]

I think you are mistaken, SB doesn't take into account ICC rankings. It takes into account the final points table of that tournament only.

Ahh ok...either way, simplicity is the best option.