Some thoughts on the Bilbao scoring system in London

At the 2nd London Chess Classic, as in the first, the so-called Bilbao scoring system was used, which is identical to the soccer systen and in the official tournament rules is formulated as follows:

While we are on the subject it is appropriate to mention another rule that was introduced to try and get exciting games and fewer disappointing grandmaster draws:

Both of these rules (in addition to perhaps the choice of players) led to an extraordinarily exciting tournament, with something dramatic, sensational or scene-changing happening in each and every round. We can recall just one single game that could be termed "boring" – one out of 28, which is a highly commendable ratio. So apparently the formula was correct, and the organisers and chess public got what they were looking for.

According to the rules given above Magnus Carlsen was the sole winner of the tournament, having scored four wins to two by his closest rivals. Magnus had lost two games and drawn one, while they had drawn all their remaining encounters. This left the young Norwegian two points ahead on the Bilbao system.

Magnus Carlsen received his trophy from Garry Kasparov at the closing ceremony [photo: Ray Morris-Hill]

Final standings – official Bilbao scoring system

If we look at the traditional cross table, which scores one point for a win, half a point for a draw and zero for a loss the final situation would appear as follows:

Some traditionalists were uncomfortable with the official result. In his Huffington Post column GM Lubomir Kavalek wrote: "Ask the traditionalists, who for several centuries counted one point for a win and a half point for a draw, and they will tell you that in the year 2010 three players shared first place in London: McShane, Anand and Carlsen." This, he said, is what would go down in the annals of chess history. There were many similar postings in the chess forums and blogs, and we too received quite a few letters. Here is a small selection that reflects the arguments presented by many other readers:

Paul Lillebo, Asheville, NC, USA
GM Kavalek is really unfair in suggesting that "In the annals of chess history the final results will be noted as follows: Carlsen, Anand, Luke McShane - 4.5 points in 7 games." Quite the contrary: In the annals of chess history the result will be shown as the organizers have shown it, with Carlson winning. Kavalek seems to think that the traditional scoring system is part of the game of chess. It isn't. Consider matches, for example: Some matches have been scored by wins only, others by the traditional point system. So have the "annals of chess history" changed the results of, as Kavalek might suggest? Of course not. You play the match or tournament by the rules everyone has agreed to.

GM Kavelek's error is that he seems to dismiss the idea that the scoring system can affect the way the players play the game. But of course it can, and it should, because the values of the various outcomes change. Magnus Carlsen seemed to do precisely what the organizers hoped for: He took extra risks, very possibly because of the scoring system, and the result was a delightful six decisive games out of seven. No annals of history will change that, or the final result.

M. van Veen, Westervoort, Netherlands
I find it really strange, that a player who has lost against both other players still beats them with an equal score. Moreover, both other players have not lost a game – shouldn't that be worth more than the two losses from the winner of the match? This is a typical case that shows tha the rule three points for a win, and one point for a draw results in incomprehensible rankings. Still, my congratulations for Magnus Carlsen. He is very lucky indeed.

Karl Storm, Fribourg, Switzerland
So Mr. Kavalek thinks that if the London Classic's point system were different then the final table would be too? How clever! Well, Mr. Kavalek, the point system is what it was and so is the final table.

Henrik Carlsen, Oslo, Norway
If Magnus had lost to Kramnik in the penultimate round the two would have swapped classical scores and Magnus would be at four points – half a point behind Kramnik and Anand and McShane, not a full point, as initially reported. And yes, that makes perfect sense as his extra win equates to the value of Kramnik's extra three draws in football score used in London. Magnus would be one point ahead of Anand and McShane as his two extra wins outpace their five extra draws. The story about Magnus losing to the no 2 and 3 on the final list, and the fact that he would have won even if he had lost to the no 2, 3 and 4 (in case Kramnik had won against Carlsen) does not really say anything meaningful. They would have ended 2nd, 3rd and 4th exactly because they won against Carlsen. Substitute a loss against Carlsen for one of them and replace it with Nakamura or Adams beating Carlsen; they would swap places on the final table. Basically Magnus beat the bottom two, and among the rest, those who beat Carlsen were close to him on the table (Anand, McShane and could-have-been Kramnik), those Carlsen did beat (Nakamura and Adams) were further down.

Jouko Äijälä, Iitti, Finland
The Bilbao rule is principally OK – it is right to reward fighting chess. But the ratio 3 to 1 is not reasonable. In my opinion, the best solution – both in chess and in football and some other games – would be five points for a win and two for a draw. But, as Mr Dorfman pointed out, this doesn't suit well to double round robin tournaments.

Anthony Andrea, London
They all signed the contract for London and they all knew what it would take to win, and magnus won – end of story. But lets have a proper tournament, with five points for a win, zero points for a draw and negative three points for a loss then we will see the men and the boys.

Derek Grimmell, Clinton, Iowa, USA
The whole argument over the traditional scoring system versus the soccer system is just plain silly. Traditional and London scoring are nothing but two different ways of describing (or summarizing) the underlying dataset, which is the actual games. Traditional scoring more accurately reflects each player's overall solidity against the entire field. London scoring more accurately reflects a player's ability to defeat top-level opponents. Which is the more impressive feat? Given the leveling of information, opening preparation, and especially the improvement in defensive technique in the last 50 years, it seems logical to conclude that the ability to defeat top-level players should be valued above the ability to hold draws against them. London scoring highlights this ability. Or, put another way, traditional scoring says that defeating Anand or Kramnik is about as impressive as holding draws against them twice. London scoring says that defeating them is as impressive as holding three draws against them. For my part, I vote for the London system as more accurate. Winning a game at this level is far more impressive than it was 50 or 100 years ago, and a scoring system that recognizes this fact is probably more accurate.

Johannes J. Struijk, Terndrup, Denmark
I don't like the Bilbao system. It is too arbitrary. I can't see what is wrong with: "we fight for the point and in the case of a draw we split it." However, if you want a tiebreaker for the cases of equal final scores, then the number of games won is as good as anything else. In that case the final result would still have been a first place for Carlsen!

Michael Bacon, Louisville, KY USA
Certain players to 'fight'. Players like Carlsen and Nakamura, for example, do not need that incentive while others do, unfortunately. That said, a player should obtain more points for a win with Black, as opposed to a win with White. The same goes for a draw with Black. That would be of great benefit in tournaments with an odd number of rounds.

Simon Knight, Toluca, Mexico
I believe Carlsen fully deserved to be crowned in London despite falling to both his nearest rivals, and his jousting approach justified the scoring system. If, however, he had also succumbed to Kramnik and still emerged as king I think the revolting peasants of tradition may have torn down the battlements of Bilbao despite the fierce stone throwing of the revolutionaries. An possible peace treaty could be an adjustment of the traditional tiebreak 3x (sum of vanquished opponents' scores) + sum of battered but not felled opponents' scores. On the battlefield of London Carlsen's crown would have remained untarnished, but under other circumstances .

Malcolm Pein, Director London Chess Classic
I see there has been a lot of huffing and puffing on the three points for a win question. It's great to have a debate, even if some of the participants may be a little stuck in the past and want to put the kibosh on 3-1-0. Actually, I was not a great fan of the idea at the outset and even now, I am not totally convinced. But from a sporting perspective we were absolutely delighted with the outcome. I suspect the no draws rule might be at least as influential in producing the fantastic fighting chess we had.

I should mention a couple of points: more than one player contacted me before the event to ask which system we were employing as it was going to effect which openings they prepared. Also: look at the last round pairings and scores, with five players in contention for first place. I can't remember an all-play-all with more than four players in which every board had a potential winner and one board (Vlad v Vishy) even had both players able to secure first. Last year we invited an expert in sports sponsorship to the tournament. As a non chess player, he said the key point for him in the debate had to be the increase in the number of players in contention near the end.