Hi
,Tournament B with positions 15 and 16 is uploaded.
Just look at the crosstables there and you will see that Thinker won pos. 15 with a huge edge, only for scoring then 50% with pos. 16. So you think statistics is weird? Not that much, when you look at the overall score it is pretty clear that Thinker is having a breakthrough with version 4.5.
Then look at Amateur, scoring fantastically with both positions. That is fine and I feel happy for Will, but when you look at the overall score it cannot be denied that Amateur is still not amongst the favourites in this strong tournament.
More weird statistics? Look at the Quarks. Up to position 12 including both very close after 320 games already. Now 480 games were needed and adding the test games we can finally be almost sure that there is a considerable improvement, as the new version scored 16.5 points more so far and this despite of the individual result Quark 2.05b vs. Quark 2.35. Thomas might have another explanation for that. I can only tell...take two engines, no matter of different versions of the same engine or not, that are probably 40 ELO points apart and run several 30 or 40 games matches between them...you will get all kinds of results.
Hope that no one again wants to tell me that 100 or 200 games are enough to prove if a new version is better or not
.Finally midnight is the deadline for the pyramid. Hope that there will be some more challenges from new participants. This event is still in danger to be cancelled for lack of interest and we would never see Ruffian drop to rank 100
.Best Regards
Heinz
http://www.husvankempen.de/nunn/