Just checked here to get an update on Moro. Hope he gets well! He was always my favourite contemporary player with his unique style and tons of imagination..

I believe the 3-1-0 system was suggested by James Mason, circa 1900, maybe a few years earlier. Football later stole the idea. It was meant for round-robin tournaments, of course.

Indeed. Well at the moment we are a long way to this becoming the norm, and frankly I think a 3 point for a win system will only work in all-play-all events. Just think about the problems in a Swiss event. For all-play-all events though it's an interesting system that encourages more adventurous chess, but the rating system will clearly remain unaffected.

I don't understand this conversation at all - there's no reason to change the rating system. A draw is still a draw and a win is still a win, all that changes is the number of points in the tournament standings and leave the ratings alone. There will always be the number crunchers continually watching rating points and the guys who try to win tournaments.

When it comes to using football score in the rating system, even if a solution is found to calculate the ratings, I think the new ratings would be difficult to interpret.

Let's say player A and player B have the same rating with the traditional rating system. However, player A has a much lower drawing rate than player B. Therefore, with a rating system based on football scores, player A should have a higher rating than player B.

But what does that mean? It means, rightly, that in a round robin tournament with football scores, player A has a higher chance to win than player B. But it also means, wrongly, that when the two players play each other, player A would be predicted to have a higher chance of winning the game.

In conclusion, if football scoring system becomes the norm, I think chess would have to abandon the rating system.

If you don't agree with the scoring system don't play in the tournament, all the players knew the system before the start and should have prepared accordingly. And yes Wang's victory was well deserved, he played the type of chess to take advantage. Don't forget that he played a beautiful game against Carlsen with the positional pawn sac for domination but gifted a whole 3 points to Carlsen in the end in a bit of a tragedy, so over-all he clear and away deserved it, not even a shadow of a doubt.

Smyslov_Fan wrote on 08/03/12 at 13:37:22:

I don't think the Elo system could handle football scoring elegantly. It produces rating changes by calculating the difference between the expected result of a player and his actual result. Because the actual results of both players always sum to 1, the rating changes are balanced with each other. (I'm ignoring details like different K factors for now.)

With football scoring, the total number of points are not a constant. Imagine a two-game mini match in which both players won once (so the total score is 3-3), and another one in which they drew both games (so the total score is 2-2). Should the players' post-match ratings be higher in the first case? That can't be right.

So I think you'd have to take the tournament as a whole and scale everybody's football scores so that their sum was equal to the total number of games played. That way you don't get any weird deflation. But now you run into another problem: your rating change depends on other players' results. Say you're playing in a 5-player double round robin and you draw all your games. If everybody else draws all their games too, your 8 football points will become 4 points after scaling, just like everyone else. But if everybody else splits their matches win-loss, then they'll each have 11 football points to your 8, and after scaling you'll end up with 3.08 points instead of 4 (if my calculations are correct), a much worse result for rating purposes. I don't think that's desirable either.

In conclusion, I think we're stuck with the sum of both players' scores for a game being a constant for rating purposes, if we want to avoid cases like the above.

Wondering aloud here:

If ratings were calculated using football scoring, how much different would the rating performances be, not just for this tournament, but historically?