The Week in Chess

Tuesday, July 30, 2013

Owl Chess Engines Rating List Purpose

The Ivanhoe 1945a x64 gauntlet matches was released last July 28, 2013 and it generated some comments which questioned the credibility of the Owl Chess Engines Rating List.

For the record, the Owl Chess Engines Rating List was created to satisfy curiosity and for fun. It was never intended to be a clean and pure rating list site like CCRL, CEGT or IPON. The choice of chess engines, software tools and rating list conditions were driven by the purpose defined above. You will see that all engines available for testing includes the many "controversial" clones which were denied entry with the pure rating list sites. The choice of time control of 1 minute + 1 second was guided by what limited resources were available and the personal need to see quickly the results of the random things that needs volumes of data to see a good picture.

With the choice of fast time control of 1 minute + 1 second comes with it the inherent unreliability of data which includes the presence of losses due time forfeits. I can live with 5% losses by time forfeit because these are random occurrence that are generated by the combined interaction of the CPU, Operating System, Chess GUI Software (Arena), Chess Engine, very tight time control and other things such as BIOS, RAM, motherboard, etc. in which I have no control over. Excessive losses due to time forfeits above 5% of total games are already a chess engine time management algorithm problem and normally would not be included in the chess engines rating list when detected early. All the accompanying tournament games results are available for download without altering the losses by time forfeits which maybe useful for anybody who are interested in this statistical data. Creating a chess engines tournament is already very resource consuming, sanitizing time forfeits and the related tedious balancing of scores will be a nightmare.

For those that are looking for reliable chess engines rating list site with longer time control without seeing time forfeits, my personalized site may not be appropriate and I highly recommend CCRL, CEGT and IPON.


Bouquet 1.7b x64 vs. Ivanhoe 1945a x64 rematch

I have conducted a rematch between Bouquet 1.7b x64 and Ivanhoe 1945a x64 due to the comments made by Velmarin. In the Bouquet 1.7b run, where Bouquet 1.7b was eventually ranked very high at #8 in the rating list, I made a disclaimer that it failed to win a single match in my AMD Quad core which is the primary chess engines tournament computer with a minimum of 4 simultaneous matches running.

The computers used where Bouquet 1.7b scored favorably were all AMD dual cores. But there were some rematches made because of irregular scores. I traced the dual core computer that has the irregular scores and found out that it was the powerful of the 3 dual cores with 4 simultaneous tournaments running at a time while the other 2 has 3 tournaments.

In the Ivanhoe 1945a run, Bouquet 1.7b lost to Ivanhoe with a wide margin. I did not bother to look at the details of the result in the haste to publish the Ivanhoe run right after it was finished and the many favors I have done for Bouquet 1.7b. When the comments of Velmarin and Dmitri Gusev came out regarding the losses by time forfeit, I checked on the details of the tournament and found that the time forfeits were more than the 5% of the personal limit I set.  I traced the AMD dual core computer that caused, it was again the powerful one with 4 tournaments running.

After some reflection, I noticed a similarity of the problem I encountered with Stockfish which was published "Stockfish 3 Killed While Sleeping on Threads". The findings was that Stockfish get worse with more tournaments running simultaneously and the resolution was enabling the "Use Sleeping Threads" option in the UCI, which was also pointed out by Marco Costalba. This was independently verified and confirmed by Dmitri that the threads problem exist.

It is Velmarin's turn to check on the problem by running it in a computer with many processess/threads running to verify his Bouquet engine's reliability in computers with heavy processing load. Hopefully, Velmarin can reproduce the problem I identified.

Upon reconsideration, I decided to do a rematch between Bouquet 1.7b and Ivanhoe 1945a to clear some garbage.

These are the results of the rematch, the first score will be used and will appear in the next rating list issue.

Bouquet 1.7b x64 vs. Ivanhoe 1945a – Rematch
CPU: AMD Dual Core – 5000 bogomips, 4GB
OS: Linux/Windows 7
GUI: Arena 3.0
Time control: 1 minute + 1 second







Losses by






Time Forfeit
Description Score% Win Loss Draw Total Bouquet Ivanhoe
Bouquet-Ivanhoe Re-match 1 62.00 33 9 58 100 0 1
Bouquet-Ivanhoe Re-match 2 61.50 31 8 61 100 0 2
Bouquet-Ivanhoe Re-match 3 62.50 36 11 53 100 1 2
Total 62.00 100 28 172 300 1 5

Download the results of the above tournament games here.
.

Saturday, July 27, 2013

Ivanhoe 1945a x64 - Gauntlet Matches 100 Rounds

Ivanhoe 1945a is an Ippolit clone by +Dmitri Gusev  which was released last June 21, 2013.

It scored 45.6% with 470 wins, 646 losses and 884 draws against the top 20 strongest chess engines in the 100 rounds gauntlet tournament. Ivanhoe 1945a beat 7 strong engines including Rybka 4.1, Robbolito 0.21Q and Bouquet, but lost to the 13 stronger chess engines. It also beat Ivanhoe 46h by insignificant margin but was not able to displace because of higher overall ELO rating. 

Ivanhoe 1945a was derived from Ivanhoe 999946h, the last active version circulating in the Internet which was released last December 11, 2011. There were half-hearted releases in between but never became popular probably because there was no executable that came with it. The author/s assumed that somebody will compile it for them. Ivanhoes stagnated and was the weakest in the Ippolit family being overshadowed by Fire, Robbolito and even Igorrit. It is a crippled open source chess engine that will never compile unless some missing codes were inserted.

Ivanhoe 1945a is a corrected version of Ivanhoe 999946h where some missing source codes were added that makes compilation easier. Any interested person that wants to study Ivanhoe is wise to start with Ivanhoe 1945a.
.
Rank Engine ELO Raw Games Score% Points Win Loss Draw Chg
1 Houdini 3 Pro x64 3170 190 100 79.0 79.0 68 10 22 3
2 Robodini 1.1 x64 3090 112 100 69.0 69.0 60 22 18 2
3 Stockfish 130525 x64 3131 73 100 65.5 65.5 49 18 33 1
4 Strelka 5.5 x64 3037 54 100 62.5 62.5 45 20 35 1
5 Komodo 5.1r2 x64 3082 51 100 62.0 62.0 45 21 34 1
6 Critter 1.6a x64 3085 18 100 57.5 57.5 35 20 45 0
7 Fire 2.2 xTreme x64 3052 12 100 57.5 57.5 26 11 63 1
8 Gull 2.1 x64 3041 9 100 56.0 56.0 33 21 46 0
9 Firenzina 2.3.1 xTreme x64 3056 1 100 55.5 55.5 24 13 63 1
10 RyanFish 1 x64 3069 0 100 54.5 54.5 32 23 45 2
11 ComStock 3 x64 3068 -8 100 53.0 53.0 32 26 42 1
12 PanChess 00.400 x64 3053 -25 100 50.5 50.5 16 15 69 -1
13 Tactico Power 2011 x64 3026 -27 100 50.0 50.0 17 17 66 0
14 Ivanhoe 1945a x64 3030 -29 2000 45.6 912.0 470 646 884 3030
15 Ivanhoe 46h x64 3038 -34 100 49.0 49.0 15 17 68 0
16 Igorrit 0.086v9 x64 3059 -38 100 48.0 48.0 21 25 54 -2
17 Rybka 4.1 x64 3019 -50 100 47.0 47.0 36 42 22 -1
18 Vitruvius 1.11C x64 3012 -51 100 46.5 46.5 23 30 47 -1
19 RobboLito 0.21Q x64 3048 -56 100 45.0 45.0 16 26 58 -2
20 LEOpard 0.7c x64 3029 -62 100 44.5 44.5 25 36 39 1
21 Bouquet 1.7b x64 3060 -139 100 35.5 35.5 28 57 15 -4
.
Download computer chess engines tournament games here.

Owl Computer Chess Engines Rating List #59

The 59th Owl Computer Chess Engines Rating List released, 07/27/2013.  Changes in the ELO rating was caused by the gauntlet matches of Ivanhoe 1945a which is nowhere to be found in the list because of failure to displace the  reigning best version Ivanhoe 46h. It could be located in the All Versions rating list #59.




View the full rating list here.
 
 

Sunday, July 21, 2013

Komodo 5.1r2 x64 - Strongest Chess Engines Gauntlet Tournament, 100 rounds

Komodo 5.1r2 x64 is a UCI chess engine by +Don Dailey and +Larry Kaufman released last July 11, 2013.

After the long wait, great expectations, the deadly gauntlet fireworks, and the settling of the smoke, the reality emerged. Komodo 5.1r2 scored with 1135 wins, 739 losses and 826 draws earning 3081 ELO rating points, 27 points gain from previous version and #5 in the Top 100 Strongest Chess Engines List.

The formidable top 4 opponents were still left standing at their top positions like Houdini, Stockfish, Robodini and Critter. Surprise tormentor of Komodo is Igorrit 0.086v9, the abandoned and long forgotten Ippolit clone. Firenzina 2.3.1 drew while the rest 21 strongest chess engines including Rybka 4.1 were demolished.

Here is the gauntlet tournament result:

Rank Engine ELO Games Score% Points Win Loss Draw Chg
1 Houdini 3 Pro x64 3167 100 61 61 41 19 40 -2
2 Stockfish 130525 x64 3130 100 59 59 49 31 20 -1
3 Robodini 1.1 x64 3088 100 55 55 38 28 34 0
4 Igorrit 0.086v9 x64 3061 100 53 53 42 36 22 1
5 Critter 1.6a x64 3085 100 51 51 29 27 44 0
6 Firenzina 2.3.1 xTreme x64 3055 100 50 50 34 34 32 0
7 Komodo 5.1r2 x64 3081 2700 57.33 1548 1135 739 826 3081
8 Strelka 5.5 x64 3036 100 49.5 49.5 30 31 39 1
9 Fire 2.2 xTreme x64 3051 100 49 49 28 30 42 0
10 Bouquet 1.7b x64 3064 100 49 49 26 28 46 1
11 Vitruvius 1.11C x64 3013 100 47 47 32 38 30 1
12 PanChess 00.400 x64 3054 100 47 47 28 34 38 0
13 Gull 2.1 x64 3041 100 45.5 45.5 21 30 49 1
14 RyanFish 1 x64 3067 100 43 43 25 39 36 -5
15 Ippolit 0.080b3 x64 2974 100 43 43 27 41 32 1
16 LEOpard 0.7c x64 3028 100 42 42 19 35 46 -2
17 Rybka 4.1 x64 3020 100 40 40 19 39 42 1
18 ComStock 3 x64 3067 100 39.5 39.5 22 43 35 -3
19 Saros 3.3b x64 2980 100 40 40 32 52 16 2
20 Hannibal 1.3 x64 2895 100 38.5 38.5 29 52 19 1
21 RobboLito 0.21Q x64 3050 100 37 37 19 45 36 -2
22 Tactico Power 2011 x64 3026 100 36 36 12 40 48 -1
23 Shredder 12 x64 2913 100 38.5 38.5 33 56 11 2
24 Ivanhoe 46h x64 3038 100 33 33 26 60 14 -2
25 Black Mamba 1.4 x64 2915 100 28.5 28.5 22 65 13 -3
26 Naum 4.2 x64 2865 100 25.5 25.5 17 66 17 1
27 Akkad 0.52b x64 2956 100 23.5 23.5 11 64 25 -2
28 Sting SF 3 x64 2964 100 28 28 28 72 0 -3
.
Download the computer chess engines tournament games here.

Chessdom News