More on tie-breaks

The discussion on tie breaks used for the 2011 European Championship is bubbling along over at Chessvibes. Of course the debate also covers what is the 'best' tie-break system to be used for various tournaments, a subject that is currently in front of the FIDE Rules and Tournament Regulations Commission.

The major issue is how you differentiate between the performance of players that finish on the same score. The conventional wisdom for swiss tournaments, and one I agree with, is that the player that plays the harder field has achieved 'more' than the other players. The difficulty is what constitutes the 'harder' field.

Now I don't have a problem with Sum of Opponents Scores (Buchholz) as the first tie break. Some may have a preference for Median Buchholz (drop the highest and lowest scoring opponent) due to the effect of first round pairings, but I'm not convinced it is a better system (your first round opponent may have a shocker, or may have a blinder, but then so may your second round or third round opponent).

However one potential defect in the system is: What happens if one of your opponents plays a particularly hard field themselves (and consequently gets a lower score)?

Now while this concerns another level of complexity, GM John Nunn proposed a solution a number of years ago. He suggested a Sum of the Sum of Opponent Scores. In fact he suggested (if I remember correctly) this could be extended further (using an averaging system) until everyone's tie-break scores stabilised.

Interestingly I haven't seen this system formally proposed as an alternative system to the existing tie-break methods. Whether it is because it is seen to be too complicated I'm not sure, but at the next RTRC meeting I get to I'll try and find out.