An approach for negating white's advantage

Sort:
ThePeanutMonster
I was toying around with some numbers tonight, to try and see how to manage the result inequality between white and black. Lets assume that overall game results in chess are: 38% white, 32% draws 30% black. Games with a victor show 54% victory for white and 46% for black. If we assume this is because of an inherent advantage in white (which, I should add is not yet mathematically proven, so this is pure speculation of course!), then black should, in theory be compensate for the disadvantage. Example: Imagine two equal players. If there were 100 games between black and white theoretically there are two perfect outcomes: 1) 54 wins for white and 46 for black = 54 points for white and 46 points for black. or 2) 100 draws = 50 points each. Because of the difference in scores, yet the equality of the players, relatively speaking, a win for black is worth more than a win for white.  How much more? 11% more.  In order to equalize the points to an even 50 each (which reflects their equality) after 54 wins for white and 48 for black, a win for white should be worth 0.92 and a win for black worth 1.08, rather than the current 1 point each. This would give each player 50 points after 54 wins for white and 46 wins for black. (1/54)*50 =0.92; (1/46)*50=1.08 Assuming that the same advantage operates in draws, then this scoring should also change because it is harder for black to draw than for white.  If we merely adopt the same formula above, then a draw for black is worth 0.54, while for white 0.46. What this means is that in terms of draws, 100 games (each ending in a draw) would end up with a result of 27 points for black, and 23 for white.  This higher score is expected: it’s harder work for black to equalize given white’s advantage and black should be rewarded for it. I’m not necessarily advocating this approach, just thought it would be interesting to run the numbers and see what came out. Thoughts?
mrguy888

Or just play as both black and white... That is what they do in tournaments to balance the advantage.

ThePeanutMonster

Well sure, if there are multiple games. I'm talking about generally, on the individual game level, how to recognise the advantage...

mrguy888

In an individual game what is the point of scores other than win, lose, and draw? Scores are for tounraments which have multiple games.

ThePeanutMonster

Well, maybe there is no point! As I said, I was just fooling around to see what the points should, in theory, be worth, I'm not advocating an overhaul of tournament scoring... which obviously is that way for a reason!

Vance917

The point seems to be that there could be multiple games, as in a tournament, but with each pair playing only one game.  Maybe 11 players, 10 games, round robin format, each player plays five games as black and five as white.  Sure, you could double it so that each combination plays both ways, but if time is short, then this may not be possible.  Well, now that I think about it, the balance means that any advantage you get by winning as black is compensated by losing as white, so maybe here it would not matter.  But in a larger pool, as we have here in Chess.com, with no guarantee of balance, it could be worth further consideration.

mrguy888

Fair enough.

The biggest problem I can see with something like that is that it is really difficult to actually quantify the advantage of white moving first. Even with statistics, theory changes over time which can change the actual value of going first. Even if you could quantify it you would have to requantify it every now and then to keep the point of equalizing the colors valid which is much more work than just making sure each player plays both colors equally.