Author: Heinz van Kempen
Date: 08:04:24 06/19/05
Hi all , since the release yesterday morning CEGT testers are running a big Fruit 2.1 test with CEGT time control 40/40 adapted to 2 Ghz via Crafty benchmark. So far we did not play with full power but can give 173 games played on seven machines. Christian and Charles this afternoon decided to interrupt their CEGT WCC for a while to give more games and Chris also joined now, so that we are testing currently on ten fast machines. So here is what we have so far with still huge error bars: Code: 2 Fruit 2.1 : 2720 41 41 173 63.9 % 2621 38.7 % 16 Fruit 2.0 : 2604 17 17 1176 46.3 % 2629 30.6 % Here is the CEGT rating list calculated a few hours ago: Code: Program Elo + - Games Score Av.Op. Draws 1 Shredder 9 : 2750 12 12 2387 69.9 % 2604 28.9 % 2 Fruit 2.1 : 2720 41 41 173 63.9 % 2621 38.7 % 3 Fritz 8 : 2703 16 16 1365 61.4 % 2623 30.3 % 4 Junior 9 : 2691 14 14 1777 61.3 % 2612 31.3 % 5 Hiarcs 9 : 2670 13 13 1941 57.4 % 2619 34.7 % 6 Gandalf 6.0 : 2655 13 13 1919 55.4 % 2617 33.7 % 7 Ktulu 7.0-7.0a : 2655 16 16 1279 58.3 % 2597 31.8 % 8 Chess Tiger 15.0 : 2647 12 12 1895 54.1 % 2618 37.8 % 9 Ruffian 2.1.0 : 2635 14 14 1679 52.5 % 2618 34.0 % 10 CM 10000 Steadfast : 2633 22 22 608 50.7 % 2628 36.3 % 11 Ruffian Leiden2003 : 2629 30 30 326 58.7 % 2568 37.7 % 12 List 512 : 2628 16 16 1283 49.8 % 2629 33.0 % 13 Pro Deo 1.1 : 2627 16 16 1203 49.8 % 2629 30.4 % 14 CMX Yoda : 2618 23 23 578 49.0 % 2625 33.4 % 15 Spike 0.9a : 2605 16 16 1127 51.2 % 2597 34.7 % 16 Fruit 2.0 : 2604 17 17 1176 46.3 % 2629 30.6 % 17 SOS 5 for Arena : 2599 16 16 1246 45.3 % 2632 34.2 % 18 Pseudo 0.7c : 2594 19 19 793 48.5 % 2605 37.7 % 19 Deep Sjeng 1.6 : 2591 23 23 621 43.9 % 2634 29.8 % 20 SlowChess Blitz WV : 2591 16 16 1086 47.3 % 2609 36.7 % 21 Aristarch 4.50 : 2587 17 17 1161 44.0 % 2629 31.8 % 22 Ktulu 5.1 : 2570 26 26 510 40.8 % 2635 28.6 % 23 Zappa 1.0 : 2568 17 17 1051 46.1 % 2596 34.3 % 24 Anaconda 2.0.1 : 2568 17 17 1101 44.1 % 2609 33.2 % 25 Thinker 4.7a : 2567 20 20 716 42.6 % 2619 36.6 % 26 Pharaon 3.2 : 2554 17 17 1037 42.2 % 2608 35.4 % 27 Delfi 4.5 : 2552 17 17 1110 41.7 % 2610 32.5 % 28 Glaurung 0.2.3 : 2551 21 21 749 42.0 % 2607 32.2 % 29 DanChess CCT7 : 2550 17 17 1050 43.3 % 2597 32.1 % 30 GLChess 3.01.2.2 : 2545 42 42 179 41.9 % 2602 33.5 % 31 AnMon5.50 : 2536 18 18 1049 41.1 % 2598 31.4 % 32 Naum 1.71 : 2528 22 22 618 38.8 % 2608 39.0 % 33 Naum 1.7 : 2526 27 27 442 41.5 % 2586 31.4 % 34 Patriot 1.3.0 : 2511 27 27 510 32.4 % 2639 25.5 % 35 Amyan 1.595 : 2503 18 18 1042 36.3 % 2600 29.7 % 36 Yace 0.99.87 : 2495 26 26 510 36.3 % 2593 29.4 % 37 Movei 00.8.310 : 2493 40 40 226 35.6 % 2596 26.1 % 38 AnMon5.53 : 2493 61 61 90 47.2 % 2512 30.0 % 39 Ufim 6.00 : 2488 38 39 226 34.5 % 2599 31.9 % 40 Amyan 1.597 : 2483 41 41 169 40.8 % 2547 39.1 % When I remember correctly it was Uri who stated (when Fruit 2.0 was out) that Fruit will be stronger than Shredder in the year 2007. Will it really last so long? Congratulation to Fabien! Even when this version drops by 40 points it is still sensational. Those who want to follow the rating progression for Fruit can do this here (updated three or four times daily with games from all testers, thanks to Michael Koppel for this fantastic sheet where ratings are calculated automatically when giving in the results, you can even test how a result would have to be, to improve ELO a bit and so on): http://www.husvankempen.de/nunn/Fruit_21.htm Games are available for download with evaluation, search depth, etc.(updated daily): http://www.husvankempen.de/nunn/downloads.htm When you want to compare winning percentages for the engines relatively to Fruit 2.0 you can do this in the statistics below the rating list: http://www.husvankempen.de/nunn/rating.html I give some results for Fruit 2.1 below: Code: Fritz 8 vs. Fruit 2.1 --- (1,0:3,0) ---Christian Junior 9 vs. Fruit 2.1 --- (2,0:6,0) ---Christian + Heinz Hiarcs 9 vs. Fruit 2.1 --- (10,0:12,0) --- Michael Ktulu 7.0a vs. Fruit 2.1 --- (1,5:3,5) ---Christian Gandalf 6.0 vs. Fruit 2.1 --- (1,5: 2,5) ---Christian Chess Tiger 15.0 vs. Fruit 2.1 --- (3,5:4,5) ---Christian + Heinz Ruffian 2.1.0 vs. Fruit 2.1 --- (2,5:1,5) --- Christian List 5.12 vs. Fruit 2.1 --- (5,5:14,5) ---Heinz Pro Deo 1.1 vs. Fruit 2.1 --- (9,5: 10,5) --- Heinz Spike 0.9a vs. Fruit 2.1 --- (6,0:16,0) --- Heinz SOS 5 vs. Fruit 2.1 --- (7,0:13,0) --- Michael SlowChess Blitz WV vs. Fruit 2.1 --- (9,0:11,0) --- Heinz Glaurung 0.2.3 vs. Fruit 2.1 --- (2,5:3,5) --- Uschi Ufim 6.00 vs. Fruit 2.1 --- (0,0:5,0) --- Uschi http://www.husvankempen.de/nunn/ http://www.chessfighters.de/cegt/ http://www.miko42.de/ http://uk.geocities.com/bloodhound66@btinternet.com/index.htm Best Regards Heinz
This page took 0.01 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.