Author: José Carlos
Date: 13:46:49 01/21/02
When CCT3 happend, I was close to finishing my complete rewrite of Averno, so I had to use an old version (0.32) that had almost nothing to do with new Averno. So I felt sad and faced the tournament with less interest. For CCT4 I had my new version ready. The new design allows me to code many things very cheaply so, even with the lack of time due to work (conversion over to euro has been a big mess for us who write programs for companies) I had a lot of new features in Averno. The week before CCT I implemented book learning and tried an agressive approaching, giving the learned knowledge a big score so that it'll influence the book choices strongly. But it wasn't enough, and in the first 6 rounds Averno made some mistakes in the opening. Round 1. Pharaon - Averno: 1-0 After the opening Averno played the strange g5, creating a weak position for an inexistent attack. Pharaon outplayed Averno clearly, giving it no chance. Round 2. Averno - Hiarcs: 0-1 Another loss. Bad way to start a tournament... In the opening, Hiarcs sacrifaced a exchange for a pawn, which Averno took happily. I was analyzing with tiger in another computer and tiger agreed with Averno for some moves, believing Hiarcs was wrong. But Hiarcs was right, it outplayed Averno and Tiger and went over Averno's position. Not a chance for Averno. I was really impressed at how these super-strong programs "understand" activity and pressure. I (~2150 FIDE) had a hard time trying to understand black compensation, but it was there, and Hiarcs used it. Fantastic. My congratulations to Marc Uniacke. Round 3. Celes - Averno: 0-1 In an even position out of book, Celes played a losing move. Immediatly, Averno showed +2.xx and there was no game after that. I'm sorry for Celes programmer; he had a lot of problems during the tournament and I hope he can fix his bugs this week. Come on! you still have 5 rounds! Round 4. Averno - Arasan: 1-0 This time luck was on my side. The opening was very comfortable for white and also, Averno's eval is incomplete. It is acceptable in some areas, and absolutly blind in others. This opening lead the game to a position where Averno had some knowledge to use. [D]rnq2rk1/pp3p1p/1n4p1/3pP3/3P1P2/P3QN2/6PP/R1B2RK1 w - - 0 19 f5 was a nice move I dind't expect Averno to find. I was very happy because Arasan is a strong program against whom Averno usually has a lot of problems. I expect Arasan to finish with more points that Averno at the end. Round 5. PostModernist - Averno: 1-0 Again a bad opening choice. A gambit which led nowhere. Averno went out of book with a bad position and lost chanceless against a very strong opponent. So far I can't be sad; losing against Pharaon, Hiarcs and PostModernist is "normal". Averno has played some good games on ICC vs PM, but most of time PM beats Averno. Round 6. Averno - Tinker: 1-0 Tinker is more or less of the same strength. I expected a good fight and I hoped I could win this one. After the opening, Tinker was unhappy with the postion, so it decided to move it's major pieces towards white king. My not very well tuned null move probably hidded black threats somehow, and so Tinker got this nice position: [D]5rk1/2pb1ppp/2pp3r/P4P2/2P1Pn1q/5B2/6P1/R1BQ1RK1 b - - 0 21 I'm not sure how, but I believe black must win here. Instead, Tinker wasn't unable to kill Averno's king, and so Averno pushed the a pawn so that at certain point, both programs realized white was winning. A very disappointing game, because Averno was totally outplayed by Tinker and only luck (since both programs were blind, it's a matter of luck) gave Averno a full point. After six games, Averno has 3 points, which is a good result, but so far I'm only satisfied with the game vs Arasan. In the other games, Averno showed nothing to be proud of. Good luck to all the participants for next week! José C.
This page took 0.03 seconds to execute
Last modified: Thu, 07 Jul 11 08:48:38 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.