John Saunders wrote:One peculiarity which I had not picked up on previously (apologies if it has already been debated here) was that, for tie-break purposes, a player's results against their highest and lowest rated opponent were ignored in calculating the TPR used for the primary tie-break in Aix. So, though Gawain's actual TPR for the event was 2653, the TPR used to calculate his primary tie-break was 2611 because his win against David Navara (2722) in the last round (a tremendous result, btw) was discounted from the tie-break calculation. Apparently there was almost a mutiny when this tie-break methodology became known to the players.
It makes sense to discount one (or perhaps two) game(s) against the lowest rated player(s), since it is bad luck on a player who, for whatever reason, is paired against a very low-rated opponent at some point, but it is hard to see the logic in discounting games against their strongest opponents. Perhaps someone could try explaining this tie-break methodology or give some information on its background. Has it been used before?
Gawain had an excellent finish - 2½/3 against Gallagher, Sargissian and Navara. Jovanka also deserves praise for a good result. A win against talented Italian GM Daniele Vocaturo (who recently won Tata Steel C ahead of Nyzhnyk) in the final round was a great way to finish.
The FIDE Regulations regarding tie-breaks are at
http://www.fide.com/fide/handbook.html? ... ew=article. As you suggest, they include the discounting of results against the lowest rated players, but not those against the highest rated ones. However, these Regulations may be out of date, as to the best of my knowledge the decisions of the 2010 FIDE Congress in this area haven't yet been reflected in the Handbook.
When using Buchholz (Sum of Opponents' Scores) as a tie-break, it's common to use Median Buchholz, discounting the highest and lowest opponents' scores. That way you don't benefit or suffer if you happen to have played someone who went on to achieve an extreme result.
When using Tournament Performance Ratings, I suppose the logic of discounting the result against your highest rated opponent is that you get a boost to your TPR from a loss to a very highly rated player compared with an "ordinary" loss.
This logic breaks down, of course, when you have defeat your highest rated opponent and then find your best result has been discounted.
Whatever the merits and demerits of the system, it's obviously unfortunate if the manner of calculation wasn't known to the players until after the last round. I suspect, however, than the information was available, but no-one really looked at it (any more than we did here).
The tie-break involved removing each player's games against their lowest and highest rated player so if Gawain had been involved he would have had to remove his win against Navara 
