Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Scottish Chess Championships 2013
Alex McFarlane Wrote:Finally, I hope everyone will again join with me in thanking the anonymous benefactor.

Yes indeed! A lot of chess activity has been made possible by their generosity.

Alex McFarlane Wrote:TPR favours the lower rated player because they are lower rated. Consider A v B where A is lower rated and they draw. A has the better performance because he scored against a higher rated player whereas B scored against the lower rated player.”


Ah yes, thanks Alex. In a 9 round Swiss that game would make a difference of 400/9 =44.4 presumably. Though, the effect of the opponents that are 400 below on the TPR, the way it’s calculated, is higher than that. For example, if a player had enough such opponents, then even with 9 out of 9 they couldn’t reach a TPR of 2400.

Still, such 400-below pairings do make it much more feasible for the player to reach an outright winning score, and that would outweigh the reduction in their chance of winning a (possibly multiple) tie break. So, no complaints would really be justified on those grounds.

Your example does highlight a very fundamental flaw in TPRs. What is (presumably) wanted is the level of rating performance that is the most likely given a player’s results; whereas the TPR merely attempts to estimate the rating which would predict their actual score

There seems to be no science concerning performances in small numbers of games – it even seems to be a concept lacking in definition. But as it’s supposed to relate to the rating system when over 30 games, wouldn’t some kind of scaling towards the present rating (based on the number of games) help? Eg, if the tournament was 9 games, take 9/30 of the difference between the TPR (as calculated) and the present rating.

I understand that the appeal of the TPR is that it is independent of the existing rating; but as that’s what seems to make it unsound, wouldn't bringing in the present rating rectify that - when the purpose is to do with measurement!?

Re playoffs, I don't believe a later full chess playoff is worthwhile when the championship has been embedded. If the aim is to produce the player with the strongest performance then it's is most likely to met by a 'sound' TPR. With the TPR as it stands, I think a rapid play contest would be better.
Cheers
Reply


Messages In This Thread

Forum Jump:


Users browsing this thread: 1 Guest(s)