Is the Sonneborn-Berger (S-B) tiebreak system fair? (From Leslie Eastoe)
Alack, alack, a woe is me, I’ve been undone by the cursed SB, Equal points with Dr Brown, Yet it’s he alone who wears the crown…. I have never approved of the SB tie break system, but it has never before affected me in such a significant manner. In the 2007 – 2008 Championship final, the SB placed me in 5th place, above Leslie Ellis in 6th spot. Obviously there is not too much to get excited about here, but I believe we should have finished in equal 5th position and not 5th/6th.
In the 2008 – 2009 Championship final, I finished on equal points with Dr Brown, but the SB split put me in 2nd place. The tie split denied me a share of the Championship title, and also of the financial reward that goes with it. Now study the final tables for these two seasons. In one we have joint champions and in the other there is a champion and a runner up. This does not look at all reasonable.
The BCCA have referred to the William Mason Championship final as the “Flagship event” of the organisation, but I feel that the use of SB to determine the Champion, devalues the event. SB is extremely convenient for tournament controllers, but in an all play all event is quite absurd. The fact that the SB method is a simple mathematical formula does not mean that it is a sensible one. It is completely arbitrary, and has no more justification than the tossing of a coin or in counting which player castled queens side the most frequently. There is no basis in logic, mathematics or common sense in awarding the title by use of SB.
Study the internet and we can find many different tie split methods. The trouble is that in an all play all event, they are all flawed. In chess the result of a game is 1 – 0 or draw. There is no goal difference to help us. This means that when players finish on equal points, there is no fair or just means of splitting them. Any method used will be biased. In the case of a finish as in the 2008 – 2009 season, the BCCA should have joint champions or the provision for a play-off for the title.
The bias in the SB tie split method is to regard wins against players that finished higher in the table as better than wins against those finishing lower in the table. At first glance this sounds reasonable, but it is not. For in an all play all event, if one player has scored better against those finishing higher in the table, and yet still scored the same total of points as another player, then the player must have underperformed against those finishing lower in the table.
To be specific to 2008 – 2009, and with apologies to players named, the SB method regarded Dr Brown’s win v Keith McLaughlin and draw with Doug Finnie as better than my win v Doug and draw with Keith. It rewarded Dr Brown’s “over-performance” v Keith, but ignores his “under-performance” v Doug. I would mention that if there is a difference in values to be attributed to wins v Keith or Doug, then there should be exactly the same difference attributed to draws against them! I also note that my and Dr Browns grading performance for these 2 games is identical.
I call upon the BCCA to abolish the use of SB in the Championship final. There is absolutely no merit in finishing above anyone on SB. In an all play all event, equal points really does mean equal! Chess is black and white (no pun intended), with no arguments about whether or not all of a ball is over a line or if any of the pieces were offside or deemed to be interfering with play.
There is no arguing your way out of the truth of a checkmate. Yet at the end of a long hard season, after immense effort and toil, the final placing in the score table is allocated arbitrarily where points are equal. The fact that one player has a higher SB than the other is meaningless, and as so the SB should not form part of our beautiful game. Older members will know that chess, like love, like music, has the power to make men happy. The use of SB makes this particular life member extremely unhappy. To misquote Prime Minister Gordon Brown, let us consign this ridiculous tie split to the dustbin of history, where it belongs.
Now I have finished ranting and raving about the SB system I would like to confirm that Dr Brown won the title fairly under the rules of the BCCA, and is to be congratulated on doing so. My complaint is with the use of a ludicrous tie split method, and not with our new champion.
Editor: Do you agree with Leslie? Or disagree? The BCCA Executive Committee is to discuss this matter and would like to hear the views of any member who feels strongly either way. If it helps, the Sonneborn-Berger tie-break system is used to split players finishing on an equal number of points. It works as follows: – Add the scores of every opponent the player beats and half of the score of every opponent the player draws. The system was named after William Sonneborn and Johann Berger, but it was invented by Oscar Gelbfuhs in 1873. The system is the main tie- breaking system in use in round robin tournaments.
{ 2 comments… read them below or add one }
Completely agree with Leslie’s comments. Makes no sense at all to use any tiebreak system in an all play all. If there is a strong need to have only one winner, not sure why that would be, then some system of playoff would be best and therefore let Chess decide the result.
Peter Mulleady
This issue was discussed at length and although the point about unfairness is well made, every tie break sytem is flawed! My own opinion is that places should be shared. However there are occasions when we do need to split players (for promotion or relegation) and we will use the same system used in international chess. To continue the debate by say looking at football, is the most goals scored the best sytem, surely the team scoring fewest goals has had to work harder?!
You must log in to post a comment.