yes, that's is right!
Wrote to the same time of the same message
Best
Frank
SWCR-64: Houdini 1.5 x64 started in around 15 hours ...
Moderator: Ras
-
Frank Quisinsky
- Posts: 6931
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: Houdini 1.5 x64 impressions ...
Hi,
yes, that's is right!
Wrote to the same time of the same message
Best
Frank
yes, that's is right!
Wrote to the same time of the same message
Best
Frank
-
wsogpr
Re: Houdini 1.5 x64 impressions ...
Hi Robert,
I use engines exclusively for analysis to improve my human play - openings and general game analysis (which I'm sure is an important use for many people). Houndi 1.5 generally gives a very good assessment of opening and middle game positions and so It is now my primary analysis engine in most of these types of positions.
The one problem I have with Houndini 1.5 that should be easy to solve is that it currently only uses the values for the pieces that are more suitable for computer play rather than human play. For human play, as I'm sure you know, the balance is 3.25 pawns for a B/N, 5 Pawns for a R, 9.75 Pawns for a Q. This would greatly improve the usefulness of your engine if there was an option for it to evaluate material imbalances in this way (perhaps including the material values in the list of configurable parameters).
Would this be possible?
I use engines exclusively for analysis to improve my human play - openings and general game analysis (which I'm sure is an important use for many people). Houndi 1.5 generally gives a very good assessment of opening and middle game positions and so It is now my primary analysis engine in most of these types of positions.
The one problem I have with Houndini 1.5 that should be easy to solve is that it currently only uses the values for the pieces that are more suitable for computer play rather than human play. For human play, as I'm sure you know, the balance is 3.25 pawns for a B/N, 5 Pawns for a R, 9.75 Pawns for a Q. This would greatly improve the usefulness of your engine if there was an option for it to evaluate material imbalances in this way (perhaps including the material values in the list of configurable parameters).
Would this be possible?
-
Houdini
- Posts: 1471
- Joined: Tue Mar 16, 2010 12:00 am
Re: Houdini 1.5 x64 impressions ...
Sean,wsogpr wrote:Hi Robert,
I use engines exclusively for analysis to improve my human play - openings and general game analysis (which I'm sure is an important use for many people). Houndi 1.5 generally gives a very good assessment of opening and middle game positions and so It is now my primary analysis engine in most of these types of positions.
The one problem I have with Houndini 1.5 that should be easy to solve is that it currently only uses the values for the pieces that are more suitable for computer play rather than human play. For human play, as I'm sure you know, the balance is 3.25 pawns for a B/N, 5 Pawns for a R, 9.75 Pawns for a Q. This would greatly improve the usefulness of your engine if there was an option for it to evaluate material imbalances in this way (perhaps including the material values in the list of configurable parameters).
Would this be possible?
The values Houdini uses are actually very close to the classical values you suggest, i.e. 325 cp for minor pieces, 500 cp for Rook and 975 for Queen.
It's important to understand that a piece value is not just a static value, it's significantly influenced by dynamic factors: mobility, pawn structure, open lines, strong fields or weak fields, attacking potential.
I could provide a host of UCI options for tweaking the piece values and all dynamic factors, but honestly, you would find it very difficult to improve upon the default values.
Robert
-
Frank Quisinsky
- Posts: 6931
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: SWCR after 75.080 / 902 of 1.000 Houdini 1.5 games!
Hi there,
now I can look in the results. Not at home the most time at xmas.
1a.
In 5 games I must make a correction!
All are Komodo 1.2 games, remis end position and Komodo lost on time. Corrected to remis. Same I do all the time because I need a clear database for my statistics. Lost on time games in unclear postions will be replayed in SWCR, but all of this 5 games vs. Komodo 1.2 was clear remis positions (well known Komodo 1.2 bug).
1b.
In one game a little bit trouble with Junior 12.0. Junior 12.0 crasht the GUI, the most of the time only 7 / 8 matches are still running in the latest two days. With other words, no problems in more as 900 games with Houdini 1.5 x64.
2a.
I made some statistics with the Houdini 1.5 games.
29x won under 50 moves up to mate (in this statistic now the new number 1 too, Spark 1.0 x64 now on place 2.
2b.
One game lost Houdini under 50 moves up to mate (vs. Shredder 12). Only Spark 1.0 don't lost a game under 50 moves up to mate. Second place here. In all other middlegame / endgame statistics I create so far Houdini is on place 1. Perhaps I will give my 39 little statistics to the SWCR database later free too. I will work a little bit on it to make the Excel table better.
Very nice short games Houdini 1.5 plays.
Houdini 1.5 is really clearly improved. For me not to see that Houdini have to do with one of the latest IvanHoe compiles. To many differents. Only the late middlegame is a point. Here all of the IPP family engines are very strong, comparable to Rybka. Can be random or not, I don't know.
From my point of view Houdini will get now his own rating place in my list and will not longer in the group of IPP family engines. Houdini will later start in SWCR-32 too (in SWCR-32 IPP family is not playing).
Congratulation to the author!
OK, let us follow the latest 100 games before Rybka 4 Exp. 61 started in SWCR.
Best
Frank
If a person have interest I can create a download link of Houdini games with Shredder GUI comments (time, evaluation, ply).
now I can look in the results. Not at home the most time at xmas.
1a.
In 5 games I must make a correction!
All are Komodo 1.2 games, remis end position and Komodo lost on time. Corrected to remis. Same I do all the time because I need a clear database for my statistics. Lost on time games in unclear postions will be replayed in SWCR, but all of this 5 games vs. Komodo 1.2 was clear remis positions (well known Komodo 1.2 bug).
1b.
In one game a little bit trouble with Junior 12.0. Junior 12.0 crasht the GUI, the most of the time only 7 / 8 matches are still running in the latest two days. With other words, no problems in more as 900 games with Houdini 1.5 x64.
2a.
I made some statistics with the Houdini 1.5 games.
29x won under 50 moves up to mate (in this statistic now the new number 1 too, Spark 1.0 x64 now on place 2.
2b.
One game lost Houdini under 50 moves up to mate (vs. Shredder 12). Only Spark 1.0 don't lost a game under 50 moves up to mate. Second place here. In all other middlegame / endgame statistics I create so far Houdini is on place 1. Perhaps I will give my 39 little statistics to the SWCR database later free too. I will work a little bit on it to make the Excel table better.
Very nice short games Houdini 1.5 plays.
Houdini 1.5 is really clearly improved. For me not to see that Houdini have to do with one of the latest IvanHoe compiles. To many differents. Only the late middlegame is a point. Here all of the IPP family engines are very strong, comparable to Rybka. Can be random or not, I don't know.
From my point of view Houdini will get now his own rating place in my list and will not longer in the group of IPP family engines. Houdini will later start in SWCR-32 too (in SWCR-32 IPP family is not playing).
Congratulation to the author!
OK, let us follow the latest 100 games before Rybka 4 Exp. 61 started in SWCR.
Best
Frank
Code: Select all
75.080 SWCR games
Last update: December 26th, 2010 (23:00)
At the moment *19* updates will be added!
119. x64 Rybka 4 Exp. 61 Update Scheduled: 27.12.2010
117. x64 Houdini 1.5 Update Started: 23.12.2010 ** still running
116. x32 Booot 5.1.0 NEW Started: 15.12.2010 * ended
115. x64 Spark 1.0 Update Started: 15.12.2010 * ended
114. x64 IvanHoe B49jA Update Started: 10.12.2010 * ended
113. x64 GullChess 1.1 Update Started: 06.12.2010 * ended
112. x64 Critter 0.90 Update Started: 03.12.2010 * ended
111. x64 Umko 1.1 x64 Update Started: 29.11.2010 * ended
110. x64 Rybka 4 Exp. 42 Update Started: 11.11.2010 break up to January 01st, 2011
109. w32 Zarkov 6.44 NEW Started: 11.11.2010 break up to January 01st, 2011
108. x64 Tornado 4.25 Update Started: 11.11.2010 break up to January 01st, 2011
107. x64 Crafty 23.4 JA Update Started: 10.11.2010 break up to January 01st, 2011
106. w32 Sjeng c't 2010 Update Started: 10.11.2010 break up to January 01st, 2011
105. x64 Stockfish 1.9.1 JA Update Started: 08.11.2010 break up to January 01st, 2011
104. x64 Equinox 0.87t Update Started: 07.11.2010 break up to January 01st, 2011
103. x64 Junior 12.0 31.10.10 Update Started: 07.11.2010 break up to January 01st, 2011
102. x64 Gaviota 0.80 NEW Started: 07.11.2010 break up to January 01st, 2011
101. x64 Protector 1.3.6 Rev. 370 Update Started: 07.11.2010 break up to January 01st, 2011
100. w32 Jonny 4.00 Again Started: 07.22.3020 break up to January 01st, 2011
Rank Name Elo + - games score oppo. draws
1 Houdini 1.5 x64 3002 23 23 902 82% 2742 25% NEW + 53
2 Rybka 4 x64 Exp. 42 2959 23 22 897 78% 2729 26% NEW + 20
- Houdini 1.03a x64 2949 22 21 920 79% 2719 29%
- Rybka 4 x64 2939 17 17 1520 80% 2700 29%
3 IvanHoe B49jA x64 2931 19 19 1116 77% 2730 33% NEW + 9
- IvanHoe B52aC x64 2922 21 21 880 77% 2728 34%
4 Stockfish 1.8.0 JA x64 2906 18 18 1200 75% 2715 33%
- Rybka 3 x64 2905 23 22 840 78% 2687 28%
- Stockfish 1.9.1 JA x64 2902 21 21 896 72% 2730 34% NEW - 4
- Fire 1.31 x64 2901 21 20 920 75% 2721 36%
- Stockfish 1.7.1 JA x64 2896 19 18 1120 76% 2706 34%
- Stockfish 1.9.1 JA 2893 21 20 960 78% 2689 31%
- Rybka 4 2891 18 18 1200 76% 2695 32%
- Stockfish 1.7.1 JA 2873 18 18 1200 75% 2688 31%
5 Critter 0.90 x64 2873 18 18 1157 70% 2733 37% NEW + 37
- Stockfish 1.8.0 JA 2861 20 20 920 74% 2695 37%
- Rybka 3 2858 16 16 1520 74% 2688 31%
- Critter 0.80 x64 2836 16 16 1525 67% 2705 33%
6 Naum 4.2 x64 2833 13 13 2322 66% 2714 36%
- Stockfish 1.6.3 JA 2825 18 18 1080 71% 2678 36%
- Naum 4.2 2822 15 15 1680 68% 2695 35%
- Critter 0.80 2814 18 18 1040 67% 2694 37%
- Naum 4.1 2810 20 19 920 68% 2684 35%
- Critter 0.70 x64 2807 20 20 880 65% 2701 38%
7 Komodo 1.2 JA x64 2805 13 13 2080 63% 2712 39%
- Stockfish 1.6.0 JA 2800 19 19 960 68% 2686 39%
8 Shredder 12 2800 9 9 4521 63% 2705 36%
9 Sjeng c't 2010 2792 19 19 896 58% 2733 38% NEW + 105
- Komodo 1.0 JA x64 2791 20 20 840 64% 2693 40%
10 GullChess 1.1 x64 2787 18 17 1116 58% 2735 37% NEW + 16
- Shredder 12 x64 2787 15 15 1600 63% 2690 34%
- Naum 4.0 2783 19 18 960 65% 2680 38%
11 Deep Fritz 12 2779 14 14 1760 62% 2698 41%
- Critter 0.70 2775 19 19 920 61% 2701 39%
- GullChess 1.0a x64 2771 16 16 1326 57% 2717 38%
- Komodo 1.2 JA 2758 17 17 1200 58% 2700 41%
12 Spark 1.0 x64 2755 17 17 1076 53% 2736 41% NEW + 16
- GullChess 1.0a 2745 19 18 960 58% 2695 39%
- Fritz 12 2744 17 16 1160 59% 2686 44%
13 Thinker 5.4d Inert x64 2740 12 12 2322 53% 2716 39%
- Spark 0.5 x64 2739 13 13 1964 54% 2707 37%
14 Hiarcs 13.1 2735 11 11 2722 52% 2718 38%
- Stockfish 1.5.1 JA 2729 19 19 840 59% 2670 43%
15 Zappa Mexico II x64 2724 12 12 2320 51% 2716 40%
- Spark 0.4 x64 2717 20 20 840 53% 2696 40%
- Komodo 1.0 JA 2714 16 16 1200 53% 2693 40%
16 Junior 12.0 x64 2713 20 20 896 47% 2735 33% NEW + 31
17 Protector 1.3.4 JA x64 2712 14 14 1720 50% 2711 37%
- Thinker 5.4d Inert 2710 12 12 2240 52% 2698 42%
18 Fruit 09_07_05 x64 2707 13 12 2322 49% 2717 33%
- Spark 0.5 2707 16 16 1200 51% 2701 41%
- Critter 0.60 x64 2700 20 20 840 50% 2697 38%
- Protector 1.3.6-370 JA x64 2700 20 20 897 45% 2736 34% NEW - 12
19 Booot 5.1.0 2696 18 18 1036 43% 2745 38% NEW
- Doch 1.3.4 JA 2689 19 19 840 51% 2685 44%
- Sjeng WC-2008 x64 2687 14 14 1760 47% 2711 36%
- Critter 0.60 2687 19 19 920 50% 2687 39%
- Junior 11.2 2685 17 17 1200 48% 2702 33%
- Protector 1.3.5 x64 2685 20 20 840 47% 2707 39%
- Spark 0.4 2685 19 19 880 49% 2690 42%
20 Hannibal 1.0a x64 2683 14 14 1762 44% 2727 35%
- Junior 11.2 x64 2682 15 16 1520 47% 2707 30%
21 Onno 1.2.70 x64 2679 13 13 2081 45% 2714 37%
- Protector 1.3.4 JA 2678 14 14 1600 47% 2700 39%
- Cyclone xTreme Wrath 2674 17 17 1080 47% 2696 41%
- Protector 1.3.2 2672 17 16 1160 47% 2694 41%
- Onno 1.1.1 x64 2670 20 20 840 46% 2698 40%
- Protector 1.3.5 JA 2668 19 20 840 44% 2710 42%
- Junior 2010 2667 16 16 1240 47% 2689 36%
- Hiarcs 12.1 2666 19 19 880 47% 2688 41%
- Protector 1.3.1b 2665 19 19 840 47% 2688 42%
- Sjeng WC-2008 2664 12 12 2240 45% 2699 37%
- Doch 1.2 JA 2664 19 19 840 48% 2677 40%
- Hiarcs 12.1 Sharpen PV 2663 16 16 1280 45% 2697 39%
- Zappa Mexico II 2660 12 12 2240 44% 2699 42%
- Spark 0.3a 2654 17 17 1120 44% 2697 41%
- Doch 09.980 JA 2650 19 19 840 46% 2674 42%
- Junior 11.1a x64 2647 20 20 840 43% 2699 32%
- Junior 11.1a 2644 19 19 960 44% 2690 36%
- Spark 0.3 2641 19 19 880 43% 2689 42%
- Hannibal 1.0a 2635 18 18 1080 40% 2704 36%
- Bright 0.5c 2634 19 19 960 41% 2700 35%
- Onno 1.2.70 2634 16 16 1200 40% 2704 40%
22 Umko 1.1 x64 2633 17 17 1196 35% 2744 34% NEW + 23
- Onno 1.1.1 2629 15 15 1520 40% 2694 41%
23 Loop M1-T x64 2629 19 19 960 36% 2735 36%
24 Crafty 23.4 JA x64 2626 20 20 900 34% 2738 32% NEW + 28
- Loop 2007 x64 2623 13 13 2122 37% 2718 34%
25 Jonny 4.00 2616 14 14 1860 36% 2719 31% NEW
- Fruit 05/11/03 2614 12 12 2240 38% 2700 41%
26 Equinox 0.87t x64 2614 20 21 900 33% 2738 32% NEW + 14
- Loop 13.6 2613 15 15 1520 38% 2694 39%
- Critter 0.52b 2612 18 18 1040 38% 2697 37%
- Twisted Logic 20100131x x64 2611 18 18 1120 35% 2716 32%
- Umko 1.0 x64 2610 17 17 1200 34% 2725 37% (ponder not possible)
- Glaurung 2.2 JA 2608 17 18 1080 37% 2698 36%
- Loop 2007 2608 17 17 1200 36% 2705 35%
27 Ktulu 9.03 2607 14 14 1760 37% 2702 31%
28 SmarThink 1.20 x64 2601 13 13 2318 34% 2718 33%
- Equinox 0.83 x64 2600 18 18 1160 32% 2734 32%
- Crafty 23.3 JA x64 2598 17 18 1200 33% 2725 34%
- SmarThink 1.20 2593 12 13 2240 35% 2700 37%
- Equinox 0.83 2578 19 19 960 33% 2702 34%
- Twisted Logic 20100131x 2572 15 15 1600 32% 2705 30%
29 Spike 1.2 Turin 2571 15 16 1480 31% 2706 34%
- Cipollino 3.25 x64 2568 20 20 960 28% 2737 30%
- Crafty 23.3 JA 2563 19 20 960 31% 2702 32%
30 BugChess2 1.7 x64 2560 18 18 1200 27% 2732 30%
- Crafty 23.2 JA x64 2555 18 19 1120 28% 2718 30%
31 Scorpio 2.6 JA x64 2554 18 18 1200 27% 2727 31%
32 Chronos 1.99 x64 2552 18 19 1120 27% 2719 33% (ponder not possible)
32 Tornado 4.25 x64 2552 23 24 744 28% 2719 25% NEW + 72
- Crafty 23.3 JA x64 NP 2548 20 21 960 25% 2738 30% (test without ponder)
34 Daydreamer 1.75 JA x64 2521 19 19 1120 24% 2720 30%
- Tornado 3.6.7 x64 2480 23 24 840 19% 2726 24%
35 Zarkov 6.44 2476 27 27 624 20% 2713 25% NEW
36 Gaviota 0.80 x64 2370 31 33 624 12% 2718 15% NEW
-
Sedat Canbaz
- Posts: 3018
- Joined: Thu Mar 09, 2006 11:58 am
- Location: Antalya/Turkey
Re: SWCR after 75.080 / 902 of 1.000 Houdini 1.5 games!
Hello Frank,
Great work-many thanks for your efforts...
Its interesting to note if we compare my rating with yours,H1.5 has almost same elo difference than R4 default or R4 exp settings
Best,
Sedat
Great work-many thanks for your efforts...
Its interesting to note if we compare my rating with yours,H1.5 has almost same elo difference than R4 default or R4 exp settings
Best,
Sedat
-
Frank Quisinsky
- Posts: 6931
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: Houdini 1.5 x64 final results ***
Hi Sedat and the others,
Sedat, I know that our results are very comparable. I am looking often in your work ... thanks for your good work!
At first:
I added the 1.000 Houdini 1.5 x64 games with GUI comments (comes from Shredder Classic 4.0 GUI - ply, evaluation, time - in my German "Aktuell Page, entry 103. The complete uncommented SWCR database with 75.158 games can be found in my download area. My German Actual page can be translated in x languages with google, see link in front of this page.
At second:
You can see ...
Shredder Classic 4.0 ELO calculation = 3.014 and Bayesian 0056 = 3.005 = -9 that's normaly. Furthermore, you can see in download file with Shredder GUI comments the five games vs. Komodo 1.2 x64 I made the corrections (lost on time games in clear remis positions, I wrote it in a message before).
Here are the final results:
New SWCR rating list:
Best
Frank
Sedat, I know that our results are very comparable. I am looking often in your work ... thanks for your good work!
At first:
I added the 1.000 Houdini 1.5 x64 games with GUI comments (comes from Shredder Classic 4.0 GUI - ply, evaluation, time - in my German "Aktuell Page, entry 103. The complete uncommented SWCR database with 75.158 games can be found in my download area. My German Actual page can be translated in x languages with google, see link in front of this page.
At second:
You can see ...
Shredder Classic 4.0 ELO calculation = 3.014 and Bayesian 0056 = 3.005 = -9 that's normaly. Furthermore, you can see in download file with Shredder GUI comments the five games vs. Komodo 1.2 x64 I made the corrections (lost on time games in clear remis positions, I wrote it in a message before).
Here are the final results:
Code: Select all
Houdini 1.5 x64
Games played December 23th, 2010 - December 27th, 2010
Houdini 1.5 x64 - Rybka 4 x64 Exp. 42 (2963) 26.0 - 14.0 65.00% Perf=3070
Houdini 1.5 x64 - IvanHoe B49jA x64 (2931) 24.0 - 16.0 60.00% Perf=3001
Houdini 1.5 x64 - Stockfish 1.9.1 JA x64 (2905) 30.0 - 10.0 75.00% Perf=3095
Houdini 1.5 x64 - Critter 0.90 x64 (2871) 24.5 - 15.5 61.25% Perf=2950
Houdini 1.5 x64 - Naum 4.2 x64 (2832) 26.0 - 14.0 65.00% Perf=2939
Houdini 1.5 x64 - Komodo 1.2 JA x64 (2804) 28.5 - 11.5 71.25% Perf=2961
Houdini 1.5 x64 - Shredder 12 (2800) 31.5 - 8.5 78.75% Perf=3027
Houdini 1.5 x64 - Sjeng c't 2010 (2793) 32.5 - 7.5 81.25% Perf=3047
Houdini 1.5 x64 - GullChess 1.1 x64 (2787) 32.0 - 8.0 80.00% Perf=3027
Houdini 1.5 x64 - Spark 1.0 x64 (2757) 35.5 - 4.5 88.75% Perf=3115
Houdini 1.5 x64 - Thinker 5.4d Inert x64 (2739) 33.5 - 6.5 83.75% Perf=3023
Houdini 1.5 x64 - Hiarcs 13.1 (2736) 35.0 - 5.0 87.50% Perf=3074
Houdini 1.5 x64 - Zappa Mexico II x64 (2725) 35.5 - 4.5 88.75% Perf=3083
Houdini 1.5 x64 - Junior 12.0 x64 (2716) 36.5 - 3.5 91.25% Perf=3123
Houdini 1.5 x64 - Fruit 09_07_05 x64 (2708) 37.5 - 2.5 93.75% Perf=3178
Houdini 1.5 x64 - Protector 1.3.6-370 JA x64 (2699) 33.5 - 6.5 83.75% Perf=2983
Houdini 1.5 x64 - Booot 5.1.0 (2696) 35.5 - 4.5 88.75% Perf=3054
Houdini 1.5 x64 - Hannibal 1.0a x64 (2682) 33.0 - 7.0 82.50% Perf=2951
Houdini 1.5 x64 - Onno 1.2.70 x64 (2680) 37.0 - 3.0 92.50% Perf=3116
Houdini 1.5 x64 - Umko 1.1 x64 (2634) 37.0 - 3.0 92.50% Perf=3070
Houdini 1.5 x64 - Crafty 23.4 JA x64 (2628) 37.5 - 2.5 93.75% Perf=3098
Houdini 1.5 x64 - Loop 2007 x64 (2624) 37.5 - 2.5 93.75% Perf=3094
Houdini 1.5 x64 - Jonny 4.00 (2617) 38.0 - 2.0 95.00% Perf=3128
Houdini 1.5 x64 - Equinox 0.87t x64 (2612) 35.5 - 4.5 88.75% Perf=2970
Houdini 1.5 x64 - SmarThink 1.20 x64 (2601) 35.5 - 4.5 88.75% Perf=2959
828.5 - 171.5 82.85% Perf=3014
1000 of 1000 games played
Conditions:
Windows XP Prof. x64 Edition
SClassic 4, Fritz 11 GUI
S12: Random-Book 4.1
F12: Perfect 2010
40 moves in 10 minutes
ponder = on
each engine = 1 core
resign / learning = off
hashtables = 256Mb
4TB Nalimov, GaviotaB., egbb,
ShredderB., TotalB. & TripleB.
Intel® Core™2 Quad
4x Q9550, 2,83GHzCode: Select all
75.178 SWCR games
Last update: December 27th, 2010 (07:30)
At the moment *19* updates will be added!
118. x64 Rybka 4 Exp. 61 Update Started: 27.12.2010 ** still running
117. x64 Houdini 1.5 Update Started: 23.12.2010 * ended
116. x32 Booot 5.1.0 NEW Started: 15.12.2010 * ended
115. x64 Spark 1.0 Update Started: 15.12.2010 * ended
114. x64 IvanHoe B49jA Update Started: 10.12.2010 * ended
113. x64 GullChess 1.1 Update Started: 06.12.2010 * ended
112. x64 Critter 0.90 Update Started: 03.12.2010 * ended
111. x64 Umko 1.1 x64 Update Started: 29.11.2010 * ended
110. x64 Rybka 4 Exp. 42 Update Started: 11.11.2010 break up to January 01st, 2011
109. w32 Zarkov 6.44 NEW Started: 11.11.2010 break up to January 01st, 2011
108. x64 Tornado 4.25 Update Started: 11.11.2010 break up to January 01st, 2011
107. x64 Crafty 23.4 JA Update Started: 10.11.2010 break up to January 01st, 2011
106. w32 Sjeng c't 2010 Update Started: 10.11.2010 break up to January 01st, 2011
105. x64 Stockfish 1.9.1 JA Update Started: 08.11.2010 break up to January 01st, 2011
104. x64 Equinox 0.87t Update Started: 07.11.2010 break up to January 01st, 2011
103. x64 Junior 12.0 31.10.10 Update Started: 07.11.2010 break up to January 01st, 2011
102. x64 Gaviota 0.80 NEW Started: 07.11.2010 break up to January 01st, 2011
101. x64 Protector 1.3.6 Rev. 370 Update Started: 07.11.2010 break up to January 01st, 2011
100. w32 Jonny 4.00 Again Started: 07.22.3020 break up to January 01st, 2011
Rank Name Elo + - games score oppo. draws
1 Houdini 1.5 x64 3005 22 21 1000 83% 2741 25% NEW + 57
2 Rybka 4 x64 Exp. 42 2959 23 22 900 78% 2730 26% NEW + 20
- Houdini 1.03a x64 2948 22 21 920 79% 2718 29%
- Rybka 4 x64 2939 17 17 1520 80% 2700 29%
3 IvanHoe B49jA x64 2931 19 19 1120 77% 2730 33% NEW + 9
- IvanHoe B52aC x64 2922 21 21 880 77% 2728 34%
4 Stockfish 1.8.0 JA x64 2906 18 18 1200 75% 2715 33%
- Rybka 3 x64 2905 23 22 840 78% 2687 28%
- Stockfish 1.9.1 JA x64 2901 21 20 899 72% 2731 34% NEW - 5
- Fire 1.31 x64 2901 21 20 920 75% 2720 36%
- Stockfish 1.7.1 JA x64 2896 19 18 1120 76% 2706 34%
- Stockfish 1.9.1 JA 2893 21 20 960 78% 2689 31%
- Rybka 4 2891 18 18 1200 76% 2695 32%
- Stockfish 1.7.1 JA 2873 18 18 1200 75% 2688 31%
5 Critter 0.90 x64 2872 18 18 1160 70% 2734 37% NEW + 36
- Stockfish 1.8.0 JA 2861 20 20 920 74% 2695 37%
- Rybka 3 2859 16 16 1520 74% 2688 31%
- Critter 0.80 x64 2836 16 16 1525 67% 2705 33%
6 Naum 4.2 x64 2833 13 13 2325 66% 2715 36%
- Stockfish 1.6.3 JA 2825 18 18 1080 71% 2679 36%
- Naum 4.2 2822 15 15 1680 68% 2695 35%
- Critter 0.80 2814 18 18 1040 67% 2694 37%
- Naum 4.1 2810 20 19 920 68% 2684 35%
- Critter 0.70 x64 2807 20 20 880 65% 2701 38%
7 Komodo 1.2 JA x64 2805 13 13 2084 63% 2712 39%
- Stockfish 1.6.0 JA 2800 19 19 960 68% 2686 39%
8 Shredder 12 2800 9 9 4525 63% 2705 36%
9 Sjeng c't 2010 2792 19 19 900 58% 2734 38% NEW + 105
- Komodo 1.0 JA x64 2790 20 20 840 64% 2692 40%
10 GullChess 1.1 x64 2787 17 17 1120 57% 2735 37% NEW + 16
- Shredder 12 x64 2787 15 15 1600 63% 2690 34%
- Naum 4.0 2783 19 18 960 65% 2680 38%
11 Deep Fritz 12 2779 14 14 1760 62% 2698 41%
- Critter 0.70 2775 19 19 920 61% 2701 39%
- GullChess 1.0a x64 2771 16 16 1326 57% 2717 38%
- Komodo 1.2 JA 2758 17 17 1200 58% 2700 41%
12 Spark 1.0 x64 2755 17 17 1080 53% 2737 41% NEW + 16
- GullChess 1.0a 2745 18 18 960 58% 2695 39%
- Fritz 12 2744 17 16 1160 59% 2686 44%
13 Thinker 5.4d Inert x64 2739 12 12 2326 53% 2716 39%
- Spark 0.5 x64 2739 13 13 1964 54% 2707 37%
14 Hiarcs 13.1 2735 11 11 2726 52% 2719 38%
- Stockfish 1.5.1 JA 2730 19 19 840 59% 2670 43%
15 Zappa Mexico II x64 2724 12 12 2324 51% 2717 40%
- Spark 0.4 x64 2717 20 20 840 53% 2696 40%
- Komodo 1.0 JA 2714 16 16 1200 53% 2693 40%
16 Junior 12.0 x64 2713 20 20 901 47% 2737 33% NEW + 32
17 Protector 1.3.4 JA x64 2711 14 14 1720 50% 2711 37%
- Thinker 5.4d Inert 2710 12 12 2240 52% 2698 42%
- Spark 0.5 2707 16 17 1200 51% 2701 41%
18 Fruit 09_07_05 x64 2706 13 13 2326 49% 2717 33%
- Critter 0.60 x64 2700 20 20 840 50% 2697 38%
- Protector 1.3.6-370 JA x64 2700 20 20 901 45% 2737 34% NEW - 11
19 Booot 5.1.0 2696 18 18 1040 43% 2746 38% NEW
- Doch 1.3.4 JA 2689 19 19 840 51% 2685 44%
- Sjeng WC-2008 x64 2687 14 14 1760 47% 2711 36%
- Critter 0.60 2687 19 19 920 50% 2687 39%
- Junior 11.2 2685 17 17 1200 48% 2702 33%
- Spark 0.4 2685 19 19 880 49% 2691 42%
- Protector 1.3.5 x64 2684 20 20 840 47% 2707 39%
20 Hannibal 1.0a x64 2683 14 14 1766 44% 2727 35%
- Junior 11.2 x64 2681 15 16 1520 47% 2706 30%
21 Onno 1.2.70 x64 2679 13 13 2085 45% 2715 37%
- Protector 1.3.4 JA 2678 14 14 1600 47% 2700 39%
- Cyclone xTreme Wrath 2674 17 17 1080 47% 2696 41%
- Protector 1.3.2 2673 16 16 1160 47% 2694 41%
- Onno 1.1.1 x64 2670 20 20 840 46% 2698 40%
- Protector 1.3.5 JA 2668 19 20 840 44% 2711 42%
- Junior 2010 2667 16 16 1240 47% 2689 36%
- Hiarcs 12.1 2666 19 19 880 47% 2688 41%
- Protector 1.3.1b 2666 19 19 840 47% 2689 42%
- Sjeng WC-2008 2664 12 12 2240 45% 2699 37%
- Doch 1.2 JA 2664 19 19 840 48% 2677 40%
- Hiarcs 12.1 Sharpen PV 2663 16 16 1280 45% 2697 39%
- Zappa Mexico II 2660 12 12 2240 44% 2699 42%
- Spark 0.3a 2654 17 17 1120 44% 2697 41%
- Doch 09.980 JA 2651 19 19 840 46% 2674 42%
- Junior 11.1a x64 2647 20 20 840 43% 2699 32%
- Junior 11.1a 2644 19 19 960 44% 2690 36%
- Spark 0.3 2642 19 19 880 43% 2689 42%
- Hannibal 1.0a 2636 18 18 1080 40% 2704 36%
- Bright 0.5c 2634 19 19 960 41% 2700 35%
- Onno 1.2.70 2634 16 17 1200 40% 2704 40%
22 Umko 1.1 x64 2633 17 17 1200 35% 2745 34% NEW + 23
- Onno 1.1.1 2629 15 15 1520 40% 2694 41%
23 Loop M1-T x64 2628 19 19 960 36% 2735 36%
24 Crafty 23.4 JA x64 2626 20 20 904 34% 2739 32% NEW + 28
- Loop 2007 x64 2623 13 13 2126 37% 2718 34%
25 Jonny 4.00 2616 14 14 1864 36% 2719 31% NEW
- Fruit 05/11/03 2614 12 12 2240 38% 2700 41%
26 Equinox 0.87t x64 2613 20 20 904 33% 2740 32% NEW + 13
- Loop 13.6 2613 15 15 1520 38% 2694 39%
- Critter 0.52b 2612 18 18 1040 38% 2697 37%
- Twisted Logic 20100131x x64 2611 18 18 1120 35% 2716 32%
- Umko 1.0 x64 2610 17 17 1200 34% 2725 37% (ponder not possible)
- Glaurung 2.2 JA 2608 17 18 1080 37% 2698 36%
- Loop 2007 2608 17 17 1200 36% 2705 35%
27 Ktulu 9.03 2607 14 14 1760 37% 2702 31%
28 SmarThink 1.20 x64 2601 13 13 2323 34% 2719 33%
- Equinox 0.83 x64 2600 18 18 1160 32% 2733 32%
- Crafty 23.3 JA x64 2598 17 18 1200 33% 2725 34%
- SmarThink 1.20 2594 12 12 2240 35% 2700 37%
- Equinox 0.83 2578 19 19 960 33% 2702 34%
- Twisted Logic 20100131x 2573 15 15 1600 32% 2705 30%
29 Spike 1.2 Turin 2572 16 16 1480 31% 2706 34%
- Cipollino 3.25 x64 2568 20 20 960 28% 2737 30%
- Crafty 23.3 JA 2564 19 20 960 31% 2703 32%
30 BugChess2 1.7 x64 2559 18 18 1200 27% 2732 30%
- Crafty 23.2 JA x64 2554 18 19 1120 28% 2718 30%
31 Scorpio 2.6 JA x64 2554 18 18 1200 27% 2727 31%
32 Chronos 1.99 x64 2552 18 19 1120 27% 2718 33% (ponder not possible)
33 Tornado 4.25 x64 2551 23 24 744 28% 2719 25% NEW + 71
- Crafty 23.3 JA x64 NP 2548 20 21 960 25% 2738 30% (test without ponder)
34 Daydreamer 1.75 JA x64 2521 19 19 1120 24% 2719 30%
- Tornado 3.6.7 x64 2480 23 24 840 19% 2726 24%
35 Zarkov 6.44 2475 27 27 624 20% 2713 25% NEW
36 Gaviota 0.80 x64 2370 31 33 624 12% 2718 15% NEW
Frank
-
wsogpr
Re: Houdini 1.5 x64 impressions ...
Hi Robert,
I agree that it would not really be possible to improve the material weightings used by Houdini 1.5 for use in computer play (nor would I have any interest in so doing). When analysing positions from human games, however, things are a little different. If one analyses several positions from standard openings involving various material imbalances it is clear that the positions are playable in human play but R4 or Houndini 1.5 may give one side or another a clear advantage. Programs such as R4 with the material values changed to those I have suggested or Komodo (which I believe uses the more human piece values) give assessments more in line with human praxis. The main practical problem that this presents is that one can easily miss tactics that lead to these type of material imbalances that would be favourable in a human game because they never get into Houdini's PV.
I hope that I have explained this clearly!
I agree that it would not really be possible to improve the material weightings used by Houdini 1.5 for use in computer play (nor would I have any interest in so doing). When analysing positions from human games, however, things are a little different. If one analyses several positions from standard openings involving various material imbalances it is clear that the positions are playable in human play but R4 or Houndini 1.5 may give one side or another a clear advantage. Programs such as R4 with the material values changed to those I have suggested or Komodo (which I believe uses the more human piece values) give assessments more in line with human praxis. The main practical problem that this presents is that one can easily miss tactics that lead to these type of material imbalances that would be favourable in a human game because they never get into Houdini's PV.
I hope that I have explained this clearly!