Glaurung: Lodz report and other news

[Tord Romstad]

Hello all,

I have spent this evening finishing my tournament report from Lodz, and doing various other little changes and additions to my web site:

• The main page has been slightly redesigned. It still looks ugly and amateurish, of course.
• The tournament report from Lodz is up, including my own comments as well as a PGN file of all Glaurung's games with evaluations and search depths for all moves. A link to the report can be found at the front page.
• An FAQ for Glaurung has been added.
• The Mac OS X GUI has been updated with a new version, containing numerous bug fixes. All PGN files now seem to load correctly, and the GUI no longer crashes when trying to start a new game after an engine vs engine match is finished. The player names in the move list window are now correctly updated when the main engine is switched.
• Salvo Spitaleri's opening book for Glaurung 2 unfortunately turned out to be broken. A new and corrected book has been added to the web site.

http://www.glaurungchess.com

Tord

[Stan Arts]

Hi Tord,

Thanks for the report, nice read.

In the position: 3k4/2n2ppp/p7/4p3/1P2B3/P5PP/1P1r1P2/1R4K1 w - - 0 1

Maybe Bxh7 seems to work from a computer's viewpoint because white pawn h4 h5 is possible untrapping it again, sort of.. That might explain the move.
That's one of the annoying habbits of alpha-beta, play really dangerous stuff if it figures it'll just make it at the end of a variation.

[Tord Romstad]

Hi Tord,

Thanks for the report, nice read.

Thanks! I'm glad you liked it.

In the position:

[diag]3k4/2n2ppp/p7/4p3/1P2B3/P5PP/1P1r1P2/1R4K1[/diag]
3k4/2n2ppp/p7/4p3/1P2B3/P5PP/1P1r1P2/1R4K1 w - - 0 1

Maybe Bxh7 seems to work from a computer's viewpoint because white pawn h4 h5 is possible untrapping it again, sort of.. That might explain the move.

It might - except for the fact that Glaurung searched deeply enough to actually see that the bishop would get captured. The main line was 40. Bxh7 g6 41. Re1 Rd5 42. Rc1 Ne8 43. Rc6 Ke7 44. Rxa6 Nf6 45. Ra7+ Ke6 46. Bxg6 fxg6 47. Ra6+ Ke7 48. Ra7+ Ke6 49. Rg7 Kf5 50. Kg2 Rd2 51. b3, with a score of +1.88. Notice that the bishop is lost at move 46.

I am therefore quite sure that the problem is in the evaluation, and not in the search. This is also confirmed by the fact that Glaurung remained very happy about its position for many moves after the capture of the bishop happened in the game.

Tord