September tests

Superleague top 32 KO 40/4'

Logo

Homepage

Tournament Name

Authors

Tester Group

Country

superchessengine.com

Top 32 KO 40/4'

Daniel Charles E.Ly

Superleague

Engine testing condition

Cpu

GUI

Opening Book

Memory

total

Memory

hash

EGTB

(3-4-5??)

EGTB

cache

Time

Tournament

type

Round

Total games

Centrino 1.86 Ghz

Fritz9

Perfect9.ctg

1 Gb

64Mo

3-4-5

16 Mo

4mn+1s

>5mn

see

more conditions

119

more conditions :	Top 30 free uci engines (golden + silver + bronze) + 2 from (reserve + promotion) 3 games + 10 games for 1/2 finals & final
Chess Engines :	Top 32 free uci engines (top 30) + 2 from reserve+promotion
date:	26-09-06
Pgn	Top32ko.pgn
Tester Homepage:	superchessengine.com
Results:	Top 32 Ko 40/4' 1/2 finals and final (10 games eng vs eng) Winner is Rybka 1.0 beta 32 bits with 7.5/10 vs Spike 1.2
Tester comments :	1st round : no surprise, the 16 strongest UCI free were too strong and managed to qualify for the next round . Deep Frenzee 1.1 was in good shape against Naum 2.0, but Naum in the tie break managed to save his place. 2nd Round : 2 Big surprise Naum 2.0 this time lost the tie break vs Pharaon 3.51 (Maybe Naum 2.0 was too tired after the 1st round :D ) lol SlowChess that took Glaurung 1.2.1 out after a big fight 3-2 No problem for Rybka, Ruffian,GB Fruit, Fruit 2.1, Spike & Toga 1/4 finals Now there are only very strong opponents .The big match was Fruit vs Gambit Fruit . GB Fruit was the winner 2-1 No problem for Rybka 1.0, Toga & Spike 1/2 finals Rybka, Toga, Spike , GB Fruit .... wahou !!! the semi finals were very very interesting Rybka 1.0 vs GB Fruit : GB Fruit was doing very good , won even some games vs Rybka 1.0. The fight was very close . after 9 games 4.5-4.5 , But Rybka like a Killer won the last game and advanced to the final round Spike 1.2 vs Toga : Rivalry between Spike and toga is big . Spike was very good against Toga (Did Spike 1.2 authors tuned Spike against toga ??) . Spike even took 2 games with black color !!! Spike won 6-4 and will meet Rybka for the big final 40/4' KO Final Rybka vs spike: It was a Rybka 1.0 beta show. Spike 1.2 fought like a warrior. However, it looked like Spike found his master 7.5 - 2.5 !!!!

Opening Book with Fritz GUI

Logo

Homepage

Tournament Name

Authors

Tester Group

Country

Opening Books

Daniel Charles E.Ly

Superleague

Engine testing condition

Cpu

GUI

Opening Book

Memory

total

Memory

hash

EGTB

(3-4-5??)

EGTB

cache

Time

Tournament

type

Round

Total games

Centrino 1.86 Ghz

Fritz9

1 Gb

64Mo

3mn +1s

>4 mn

Gauntlet

300

more Testing conditions :

5 books .ctg + a no book (Norm Pollock idea :D )

Takker tour II , Fritz 9 book, The Turk book, Paradigm, gen18, + secret=no book

for all book optimize (defaults) if I don't talk about the setting

Takker tour II+ Fritz9 = settings = optimization (defaults)
The Turk book settings (Djorge Vidanovic) = min games = 1, up to moves 60 moves, learning to 75%
Paradigm settings (Dr Wael Deeb) = min games = 2
Gen 18 (Norm Pollock) = min games = 8

Chess Engines :

Toga II 1.2.1

date:

21-09-06

Pgn (link to your homepage or here)

booktournament.rar

Tester Homepage:

superchessengine.com

Results:

Rating with Elostat

Rk	LOGO/Book	Author	Mo	W	D	L	% Wh	%Bl	Total %	Pts W	Pts B	Pts /100	elo
1	TAKKER TOUR	Takker/W. Fuller	235	37	52	11	68	58	63	34	29	63	+78
2	Fritz 9 book (commercial)			33	52	15	58	60	59	29	30	59	+53
3		Djordje Vidanovic	26	30	49	21	60	49	54.5	30	24.5	54.5	+27
4		Dr wael Deeb	263	11	25	14	47	49	48	23.5	24.5	48	-11
5		Norm Pollock	265	22	51	27	51	44	47.5	25.5	22	47.5	-13
6	No Book			6	44	50	37	19	28	18.5	9.5	28	-136

    Program                          Elo    +   -   Games   Score   Av.Op.  Draws

  1 Takker Tourbook book           :   78   58  42   100    63.0 %    -14   52.0 %
  2 Fritz 9 book                   :   54   62  41   100    59.0 %     -9   52.0 %
  3 Turk book 1                    :   27   66  42   100    54.5 %     -4   49.0 %
  4 Paradigm 1.0                   :  -10   41  69   100    48.0 %      3   50.0 %
  5 Gen 18                         :  -13   40  68   100    47.5 %      4   51.0 %
  6 Secret book                    : -136   53  52   100    28.0 %     29   44.0 %

Games        :    300 (finished)

White Wins   :     86 (28.7 %)
Black Wins   :     65 (21.7 %)
Draws        :    149 (49.7 %)
Unfinished   :   1198

White Perf.  : 53.5 %
Black Perf.  : 46.5 %

ECO A =     28 Games ( 9.3 %)
ECO B =    199 Games (66.3 %)
ECO C =     19 Games ( 6.3 %)
ECO D =     20 Games ( 6.7 %)
ECO E =     34 Games (11.3 %)

Individual statistics:

1 Takker Tourbook book      :   78  100 (+ 37,= 52,- 11), 63.0 %

Fritz 9 book                  :  20 (+  5,=  9,-  6), 47.5 %
Turk book 1                   :  20 (+  7,= 11,-  2), 62.5 %
Gen 18                        :  20 (+  6,= 12,-  2), 60.0 %
Secret book                   :  20 (+ 10,= 10,-  0), 75.0 %
Paradigm 1.0                  :  20 (+  9,= 10,-  1), 70.0 %

2 Fritz 9 book              :   54  100 (+ 33,= 52,- 15), 59.0 %

Turk book 1                   :  20 (+  5,=  9,-  6), 47.5 %
Gen 18                        :  20 (+  6,= 14,-  0), 65.0 %
Takker Tourbook book          :  20 (+  6,=  9,-  5), 52.5 %
Secret book                   :  20 (+  9,= 10,-  1), 70.0 %
Paradigm 1.0                  :  20 (+  7,= 10,-  3), 60.0 %

3 Turk book 1               :   27  100 (+ 30,= 49,- 21), 54.5 %

Fritz 9 book                  :  20 (+  6,=  9,-  5), 52.5 %
Gen 18                        :  20 (+  7,=  7,-  6), 52.5 %
Takker Tourbook book          :  20 (+  2,= 11,-  7), 37.5 %
Secret book                   :  20 (+  9,= 10,-  1), 70.0 %
Paradigm 1.0                  :  20 (+  6,= 12,-  2), 60.0 %

4 Paradigm 1.0              :  -10  100 (+ 23,= 50,- 27), 48.0 %

Fritz 9 book                  :  20 (+  3,= 10,-  7), 40.0 %
Turk book 1                   :  20 (+  2,= 12,-  6), 40.0 %
Gen 18                        :  20 (+  6,= 11,-  3), 57.5 %
Takker Tourbook book          :  20 (+  1,= 10,-  9), 30.0 %
Secret book                   :  20 (+ 11,=  7,-  2), 72.5 %

5 Gen 18                    :  -13  100 (+ 22,= 51,- 27), 47.5 %

Fritz 9 book                  :  20 (+  0,= 14,-  6), 35.0 %
Turk book 1                   :  20 (+  6,=  7,-  7), 47.5 %
Takker Tourbook book          :  20 (+  2,= 12,-  6), 40.0 %
Secret book                   :  20 (+ 11,=  7,-  2), 72.5 %
Paradigm 1.0                  :  20 (+  3,= 11,-  6), 42.5 %

6 Secret book               : -136  100 (+  6,= 44,- 50), 28.0 %

Fritz 9 book                  :  20 (+  1,= 10,-  9), 30.0 %
Turk book 1                   :  20 (+  1,= 10,-  9), 30.0 %
Gen 18                        :  20 (+  2,=  7,- 11), 27.5 %
Takker Tourbook book          :  20 (+  0,= 10,- 10), 25.0 %
Paradigm 1.0                  :  20 (+  2,=  7,- 11), 27.5 %

Tester comments :

Takker tourbook won again the book tournament .

this is the 3 rd time for Takker Tourbook (1 with Vigi Varkley+ 2 with superchessengine :D ) and good news , this book is free :D

Fritz 9 tournament book is a good book , It finished 2nd ... However , F9 book is not free .It scored well with the 2 colors .

The Turk Book is a excellent book for online tournament , it was doing good against F9 book (52.5%)

Paradigm was neither bad or good with each color ... It could be a solid book for long time control

Gen 18 was doing good, just 0.5 pts behind paradigm !!! No super killer move, it looks like solid for an opening book, even if gen18 with white side is slightly better than with black side

A simple remark with "No Opening book" don't expect to win against a strong opening book :D

Thanks to all opening books authors who spent time to hand tuned the book and to offer free opening book to chess engine fans.

Patzer 3.80

Logo

Homepage

Tournament Name

Authors

Tester Group

Country

Patzer 3.80

Daniel Charles E.Ly

Superleague

Engine testing condition

Cpu

GUI

Opening Book

Memory

total

Memory

hash

EGTB

(3-4-5??)

EGTB

cache

Time

Tournament

type

Round

Total games

Centrino 1.86 Ghz

Fritz9

Perfect9.ctg

1 Gb

64Mo

3-4-5

16 Mo

3 mn or

180s

Gauntlet

144

more Testing conditions :

Patzer 3.80 played against almost all free uci engines 2 times

Chess Engines :

Patzer vs all free uci engines

date:

17-09-06

Pgn (link to your homepage or here)

Patzergauntlet.pgn

Tester Homepage:

superchessengine.com

Results:

Engines	perf	games	+	=	-	%	pts/144
PatzerUCI/FRC	2431	144	+ 69	= 28	- 47	57.6%	83 pts

Patzer 3.80 individual  stats :

Adam 3.0                      :   2 (+  2,=  0,-  0), 100.0 %
Abrok 5.0                     :   2 (+  0,=  1,-  1), 25.0 %
Alex 1.77                     :   2 (+  2,=  0,-  0), 100.0 %
AliChess 4.06                 :   2 (+  0,=  0,-  2),  0.0 %
Amyan 1.597                   :   2 (+  0,=  1,-  1), 25.0 %
AnMon5.60                     :   2 (+  1,=  0,-  1), 50.0 %
Anechka 0.08                  :   2 (+  1,=  1,-  0), 75.0 %
Arasan 9.3                    :   2 (+  0,=  1,-  1), 25.0 %
Arion 1.7                     :   2 (+  1,=  0,-  1), 50.0 %
Aristarch 4.50                :   2 (+  0,=  2,-  0), 50.0 %
Atlas 2.20                    :   2 (+  2,=  0,-  0), 100.0 %
Ayito 0.2.994                 :   2 (+  2,=  0,-  0), 100.0 %
BibiChess 0.5                 :   2 (+  2,=  0,-  0), 100.0 %
BigLion 2.23w                 :   2 (+  2,=  0,-  0), 100.0 %
Booot 4.10.5                  :   2 (+  0,=  0,-  2),  0.0 %
Cassandre                     :   2 (+  2,=  0,-  0), 100.0 %
ChessAlex 1.3                 :   2 (+  2,=  0,-  0), 100.0 %
Chispa 4.0.3                  :   2 (+  1,=  0,-  1), 50.0 %
Clueless V1.2.2               :   2 (+  2,=  0,-  0), 100.0 %
Colossus 2006f                :   2 (+  1,=  1,-  0), 75.0 %
Deep Frenzee 1.1              :   2 (+  0,=  1,-  1), 25.0 %
Dragon 4.6                    :   2 (+  0,=  0,-  2),  0.0 %
Eagle 0.7.5                   :   2 (+  1,=  0,-  1), 50.0 %
Feuerstein 0.4.4.2            :   2 (+  1,=  1,-  0), 75.0 %
FireFly v2.0.3                :   2 (+  2,=  0,-  0), 100.0 %
FreeChess v 2.15c             :   2 (+  2,=  0,-  0), 100.0 %
Fruit 2.1                     :   2 (+  0,=  0,-  2),  0.0 %
Gaia 3.5                      :   2 (+  0,=  1,-  1), 25.0 %
Gambit Fruit 1.0 Beta 4bx     :   2 (+  0,=  0,-  2),  0.0 %
Gibbon 2.01b                  :   2 (+  1,=  1,-  0), 75.0 %
Glaurung 1.2.1                :   2 (+  0,=  0,-  2),  0.0 %
Hamsters 0.0.6                :   2 (+  1,=  1,-  0), 75.0 %
Hermann 1.7                   :   2 (+  2,=  0,-  0), 100.0 %
Homer 1.1 beta3               :   2 (+  2,=  0,-  0), 100.0 %
Ktulu 4.2                     :   2 (+  0,=  1,-  1), 25.0 %
Leila_042u                    :   2 (+  0,=  1,-  1), 25.0 %
Lime_v48                      :   2 (+  2,=  0,-  0), 100.0 %
List 512                      :   2 (+  0,=  1,-  1), 25.0 %
LittleThought v0.96           :   2 (+  2,=  0,-  0), 100.0 %
smash                         :   2 (+  2,=  0,-  0), 100.0 %
Madeleine 0.2                 :   2 (+  1,=  0,-  1), 50.0 %
Marvin 1.3.0                  :   2 (+  2,=  0,-  0), 100.0 %
Muse 0.899b                   :   2 (+  1,=  0,-  1), 50.0 %
NSVChess 0.14                 :   2 (+  2,=  0,-  0), 100.0 %
Nejmet_3.07                   :   2 (+  1,=  0,-  1), 50.0 %
Pepito v1.59                  :   2 (+  0,=  0,-  2),  0.0 %
Philemon C                    :   2 (+  2,=  0,-  0), 100.0 %
Piranha 0.5                   :   2 (+  2,=  0,-  0), 100.0 %
Pooky 2.7                     :   2 (+  2,=  0,-  0), 100.0 %
Queen 3.09                    :   2 (+  2,=  0,-  0), 100.0 %
Roce 0.0345                   :   2 (+  2,=  0,-  0), 100.0 %
Ruffian 1.0.5                 :   2 (+  0,=  1,-  1), 25.0 %
Rybka 1.0 Beta 32-bit         :   2 (+  0,=  0,-  2),  0.0 %
SOS 5 for Arena               :   2 (+  0,=  0,-  2),  0.0 %
Simontacchi 1.8-Decepticon    :   2 (+  2,=  0,-  0), 100.0 %
Sjeng 12.13                   :   2 (+  0,=  1,-  1), 25.0 %
SlowChess Blitz WV2.1         :   2 (+  0,=  1,-  1), 25.0 %
Snitch 1.6.2                  :   2 (+  1,=  1,-  0), 75.0 %
Spike 1.2 Turin               :   2 (+  0,=  2,-  0), 50.0 %
T.rex 1.9 beta                :   2 (+  2,=  0,-  0), 100.0 %
Tao 5.6                       :   2 (+  0,=  1,-  1), 25.0 %
Terra 3.4                     :   2 (+  1,=  0,-  1), 50.0 %
Toga II 1.2.1                 :   2 (+  0,=  1,-  1), 25.0 %
Tornado 1.0 Mainz             :   2 (+  2,=  0,-  0), 100.0 %
Ufim 8.01                     :   2 (+  0,=  1,-  1), 25.0 %
WJChess v1.63                 :   2 (+  0,=  2,-  0), 50.0 %
Yace Paderborn                :   2 (+  0,=  0,-  2),  0.0 %
Zappa 1.1                     :   2 (+  0,=  1,-  1), 25.0 %
Naum 2.0                      :   2 (+  0,=  0,-  2),  0.0 %
AICE 0.98                     :   2 (+  1,=  0,-  1), 50.0 %
Pharaon 3.5.1                 :   2 (+  1,=  1,-  0), 75.0 %
Twisted Logic 0.065e37        :   2 (+  1,=  0,-  1), 50.0 %

83 pts/ 144   57.6%

ECO A = 4 Games ( 2.8 %)
ECO B = 44 Games (30.6 %)
ECO C = 14 Games ( 9.7 %)
ECO D = 35 Games (24.3 %)
ECO E = 47 Games (32.6 %)

Patzer 3.80 = 2431 elo with bayelo 2439 elo with elostat

(with bayeselo based on Aristarch 4.50=2600 elo)

Tester comments :

Patzer 3.80 on single processor is a good chess engine .

It didn't score well against top 10 uci engines (Rybka, Toga , Fruit, Glaurung...). However, versus middle strength engine it was doing a good job and managed even to have equal score with strong engine like Anmon, Aristarch, Spike or take a 75% score vs Colossus,Pharaon .

Patzer score 57% versus all others free uci engines .

Approximatively 2431 elo for blitz

(Deep) Patzer 3.80 on multi Cpu is probably 50 elo better .

I think Patzer 3.80 is a good medium strengh chess engine . (probably 35 th to 45 th best free uci engines )

Thanks to Roland Pfister for this good chess engine.

GOLEN LEAGUE V9

Logo

Homepage

Tournament Name

Authors

Tester Group

Country

superchessengine.com

Golden leaguev9

Daniel Charles E.Ly

Superleague

Engine testing condition

Cpu

GUI

Opening Book

Memory

total

Memory

hash

EGTB

(3-4-5??)

EGTB

cache

Time

Tournament

type

Round

Total games

Centrino 1.86 Ghz

Fritz9

Perfect9.ctg

1 Gb

64Mo

3-4-5

16 Mo

2mn+1s

>180s

Round

Robin

110

more conditions :

Top 11 free uci engine play each others with each color

Chess Engines :

Top 11 free uci engines ( see golden league list ) + Pharaon 3.5.1

date:

14-09-06

Pgn

Golenleaguev9.pgn

Tester Homepage:

superchessengine.com

Results:

    Program                     Elo    +   -   Games   Score   Av.Op.  Draws

  1 Rybka 1.0 Beta 32-bit      : 2818  145 140    20    65.0 %   2710   30.0 %
  2 Gambit Fruit 1.0           : 2818  145 119    20    65.0 %   2710   40.0 %
  3 Toga II 1.2.1              : 2801  149 126    20    62.5 %   2712   35.0 %
  4 Fruit 2.1                  : 2784  153 133    20    60.0 %   2714   30.0 %
  5 Spike 1.2 Turin            : 2736  167 116    20    52.5 %   2719   35.0 %
  6 Pharaon 3.5.1              : 2704  132 167    20    47.5 %   2722   25.0 %
  7 Naum 2.0                   : 2704  116 167    20    47.5 %   2722   35.0 %
  8 Glaurung 1.2.1             : 2689   80 162    20    45.0 %   2723   60.0 %
  9 SlowChess Blitz WV2.1      : 2640  146 149    20    37.5 %   2728   25.0 %
 10 Ruffian 1.0.5              : 2640   88 149    20    37.5 %   2728   55.0 %
 11 List 512                   : 2586  123 138    20    30.0 %   2734   40.0 %

crosstable

performance elo with elostats .

Rank	1-2	3	4	5	6-7	8	9-10	11
Engines	Rybka GB Fruit	TogaII	Fruit2.1	Spike 1.2	Pharaon 3.5.1 Naum 2.0	Glaurung	Slowchess Ruffian	List 5.12
Pts/20	13 pts	12.5pts	12pts	10.5pts	9.5pts	9pts	7.5pts	6pts
Perf elo	2818	2801	2784	2736	2704	2689	2640	2586

Games        :    110 (finished)

White Wins   :     32 (29.1 %)
Black Wins   :     37 (33.6 %)
Draws        :     41 (37.3 %)
Unfinished   :      0

White Perf.  : 47.7 %
Black Perf.  : 52.3 %

ECO A =      1 Games ( 0.9 %)
ECO B =     41 Games (37.3 %)
ECO C =     11 Games (10.0 %)
ECO D =     22 Games (20.0 %)
ECO E =     35 Games (31.8 %)

look under the .pgn for individual stats engines per engines.

Tester comments :

Only 2 big surprise on this double round robin for golden league

Gambit Fruit for the first time did a better score than Toga II (Gambit Fruit is very good for Blitz and Toga is a lot better with longer time control ). average Toga is 60-80 pts better than GB fruit at longer time control

second surprise is Pharaon 3.5.1 that perform at 2704 elo !!!

for other engines they almost play at their high and normal level .Only exception is List that play a bad tournament (50-60 elo below his real standard )