Chess Computer Ratings
Udo Sprute
Here it is !
( Please note, that the Elo points aren't USCF points.
For USCF ratings add 200 or less. )
Udo
---------------------------------------------------------
Computerrangliste nach Werner Stowasser
(Stand: Dezember 1992)
Ausgewertet wurden:
38.710 Partien "Computer gegen Computer"
6.044 Partien "Mensch gegen Computer"
Erklaerung der Abkuerzungen:
GR = Grossrechner
PC = PC-Programm
Nr. Schachcomputer Elo
----------------------------------------------------------------------
1. Deep Thought II (GR) 2.520
2. Deep Thought I (GR) 2.470
3. Cray Blitz (GR) 2.435
4. Hitech (GR) 2.400
5. Mephisto Vancouver 68030 2.322
6. Mephisto RISC 1 MB ARM2 2.319
7. Mephisto Lyon 68030 2.310
8. Saitek Kasparow RISC 2500 512 KB ARM2 2.309
9. Gideon ChessMachine 512 KB ARM2 (PC) 2.291
10. The King ChessMachine 512 KB ARM2 (PC) 2.289
11. M Chess 1.66 AT 80486 (PC) 2.285
12. Mephisto Portorose 68030 2.281
13. BeBe (GR) 2.260
14. Fidelity Elite 68040 2.255
15. Mephisto Vancouver 68020 2.236
16. Fidelity Elite 68030 2.235
17. M Chess 1.41 AT 80486 (PC) 2.228
18. M Chess 1.66 AT 80386 (PC) 2.224
19. Fidelity Elite 68020 2.220
20. Mephisto Lyon 68020 2.210
21. Belle (GR) 2.200
22. Mephisto Portorose 68020 2.189
23. Fidelity Mach IV 68020 2.174
24. M Chess 1.41 AT 80386 (PC) 2.164
25. Fidelity Elite Premiere 68000 2.162
M Chess 1.66 AT 80286 (PC) 2.162
Mephisto Vancouver 68000 2.162
28. Fritz AT 80486 (PC) 2.158
29. Mephisto Berlin 68000 2.155
30. Fidelity Designer Master 68020 2.149
31. Mephisto Almeria 68020 2.145
32. Mephisto Lyon 68000 2.133
33. Fidelity Elite 68000 X 2 2.123
34. Chessmaster 3000 AT 80486 (PC) 2.121
35. Fidelity Elite 68000 2.108
36. Mephisto Portorose 68000 2.090
37. M Chess 1.41 AT 80286 (PC) 2.087
38. Fidelity Mach III 68000 2.079
39. Novag Diablo/Scorpio 68000 2.077
40. Mephisto Roma 68020 2.072
41. Rexchess V. 2.3 AT 80386 (PC) 2.068
42. Fidelity Designer Master 68000 2.067
Peri Epsilon 68000 2.067
44. Mephisto Dallas 68020 2.066
45. Fritz AT 80386 (PC) 2.065
46. Zarkov V. 2.5 AT 80386 (PC) 2.064
47. Mephisto Almeria 68000 2.062
48. Saitek Kasparow GK 2000 H8 - 10 MHz 2.060
49. Chessmaster 3000 AT 80386 (PC) 2.023
50. Mephisto Milano 6502 - 5 MHz 2.020
51. Mephisto Roma 68000 2.020
52. Novag Super Expert/Forte C 6502 - 6 MHz 2.018
53. Mephisto Polgar 6502 - 5 MHz 2.015
54. Mephisto Dallas 68000 2.014
55. Mephisto MM V 6502 - 5 MHz 2.011
56. Mephisto Roma II 68000 2.000
57. Fidelity Mach II 68000 1.993
58. Mephisto Academy 6502 - 5 MHz 1.976
59. Mephisto Amsterdam 68000 1.974
60. Saitek Leonardo M D 6502 - 10 MHz 1.972
61. Fidelity Travel Master H8 - 10 MHz 1.968
62. Novag Super Expert/Forte B 6502 - 6 MHz 1.966
63. Fritz AT 80286 (PC) 1.959
64. Fidelity Prestige Avantgarde 6502 - 8 MHz + GME 1.955
65. Mephisto Mega IV 6502 - 5 MHz 1.954
66. Mephisto MM IV 6502 - 5 MHz 1.940
67. Fidelity Excel 68000 1.927
68. Psion Chess Atari 68000 (PC) 1.925
69. Fidelity Designer 2100 6502 - 6 MHz 1.922
70. Saitek Turboking 6502 - 5 MHz 1.919
71. Saitek Travel Champion H8 - 8 MHz 1.912
72. Chessmaster 3000 AT 80286 (PC) 1.910
73. Fidelity Elite Avantgarde 6502 - 5 MHz 1.905
74. Saitek Leonardo M C 6502 - 8 MHz 1.902
75. CXG Sphinx Galaxy/Dominator 6502 - 4 MHz 1.900
76. Novag Super Expert/Forte A 6502 - 5 MHz 1.898
77. Fidelity Par Excellence 6502 - 5 MHz 1.895
Fidelity Phantom 6502 - 5 MHz 1.895
79. Peri Delta 6502 - 5 MHz 1.891
80. Conchess Plymate Victoria 6502 - 5,5 MHz 1.885
81. Novag Expert/Forte B 6502 - 5 MHz 1.880
82. Novag Expert/Forte A 6502 - 5 MHz 1.872
83. Mephisto College 6502 - 5 MHz 1.868
84. Fidelity Kishon Chesster 6502 - 3,6 MHz 1.863
85. Saitek Leonardo M B 6502 - 6 MHz 1.862
86. Mephisto Monte Carlo/Supermondial 6502 - 4 MHz 1.860
87. Mephisto Rebell 6502 - 5 MHz 1.851
88. Saitek Corona 6502 - 6 MHz 1.850
89. Fidelity Excel Display 6502 - 3 MHz 1.848
90. Saitek Leonardo M A 6502 - 6 MHz 1.846
91. Fidelity Excellence 6502 - 4 MHz 1.845
92. Novag Constellation Expert 6502 - 4 MHz 1.844
Saitek Stratos 6502 - 6 MHz 1.844
94. Saitek Advanced Trainer H8 - 7 MHz 1.837
95. Conchess Plymate 6502 - 5,5 MHz 1.835
Fidelity Elegance 6502 - 3,6 MHz 1.835
97. Peri Gamma 6502 - 3 MHz 1.834
98. Saitek Simultano 6502 - 5 MHz 1.832
99. Saitek Renaissance 6502 - 4 MHz 1.816
100. Fidelity Excellence 6502 - 3 MHz 1.810
101. Mephisto MM II 6502 - 3,7 MHz 1.809
102. Conchess Plymate 6502 - 4 MHz 1.806
103. Saitek Turbostar 432 6502 - 4 MHz 1.803
104. Mephisto Modena 6502 - 4 MHz 1.800
Novag Superconstellation 6502 - 4 MHz 1.800
106. Fidelity Elite Glasgow 6502 - 4 MHz 1.797
107. Novag Super Nova - Single Chip 1.790
108. Mephisto Glasgow 68000 1.785
109. Chessplayer 2150 Atari/Amiga 68000 (PC) 1.784
110. Fidelity Elite Budapest 6502 - 3,6 MHz 1.772
111. Saitek Kasparov Blitz/Prisma - Single Chip 1.760
112. Chessmaster 2100 Amiga 68000 (PC) 1.751
113. Fidelity Elite A/S 6502 - 3,2 MHz 1.750
114. Novag Constellation Quattro 6502 - 4 MHz 1.748
115. Fidelity Prestige 6502 - 4 MHz 1.743
116. Novag Super Vip - Single Chip 1.742
117. Mephisto B & P 6502 - 3,7 MHz 1.740
118. Mephisto Europa 6502 - 3 MHz 1.736
119. Conchess Glasgow 6502 - 4 MHz 1.735
120. The Final Chesscard 6502 - 5 MHz (PC) 1.733
121. Novag Constellation 6502 - 3,6 MHz 1.732
122. Chess Champion 2175 Atari/Amiga 68000 (PC) 1.712
123. Fidelity Elite 6502 - 4 MHz 1.710
124. Novag Constellation Primo - Single Chip 1.703
125. Chessmaster 2000 Atari 68000 (PC) 1.702
126. Check-Check AT 80386 (PC) 1.695
127. Fidelity Playmate-S 6502 - 3,2 MHz 1.690
128. Conchess Glasgow 6502 - 2 MHz 1.685
129. Novag Constellation 6502 - 2 MHz 1.677
130. Mephisto Mondial 6502 - 2 MHz 1.675
131. Fidelity Sensory 12 6502 - 3 MHz 1.670
132. Novag Amigo - Single Chip 1.650
133. Fidelity Sensory Super 9 6502 - 2,5 MHz 1.648
134. Mephisto Schachschule - Single Chip 1.644
135. Saitek Cavalier - Single Chip 1.642
136. Saitek Superstar 6502 - 2 MHz 1.630
137. Mephisto Excalibur II 68000 1.629
138. Fidelity Sensory 9 6502 - 2 MHz 1.620
139. Mephisto Excalibur I 68000 1.608
140. CXG Sphinx Professor - Single Chip 1.595
141. CXG Super Enterprise - Single Chip 1.585
142. Steinitz MGS III 6502 - 2,5 MHz 1.584
143. Bogol 5.0 ASB 6502 - 3 MHz 1.575
144. Mephisto Mirage 1806 - 8 MHz 1.562
145. Morphy MGS III 6502 - 2,5 MHz 1.542
146. Chess 2001 6502 - 4 MHz 1.540
147. Mephisto III S ESB 6000 1806 - 6,1 MHz 1.534
148. Fidelity Champion 6502 - 2 MHz 1.530
149. Saitek Turbo S 6502 - 3 MHz 1.514
150. Sandy MGS II 6502 - 2 MHz 1.510
151. SciSys Mark V 1806 - 2,5 MHz 1.495
152. Sargon ARB 2.0 / MGS II 1806 - 2 MHz 1.480
153. Destini MM 6502 2 MHz 1.470
154. Mephisto III 1806 - 3,5 MHz 1.464
155. Novag Savant Robot 6502 - 2 MHz 1.462
156. Regence TSB 4 6502 - 2,5 MHz 1.460
157. Novag Solo - Single Chip 1.443
158. CXG Enterprise - Single Chip 1.423
159. Milton 6502 - 2 MHz 1.420
160. Saitek Turbo - Single Chip 1.416
161. Sci Sys President Chess 1806 - 2 MHz 1.414
162. Mephisto II - 1806 - 3,5 MHz 1.412
163. Fidelity Sensory Voice - Single Chip 1.394
164. Fidelity Poppy - Single Chip 1.380
SciSys Mark III Super System - Single Chip 1.380
166. Peri Beta - Single Chip 1.376
167. Novag Savant Royal Z80 1.375
168. Novag Savant II Z80 1.366
169. Fidelity Voice Z80 1.359
170. Mephisto Mobil - Single Chip 1.355
171. Chess King Master - Single Chip 1.324
172. Fidelity CC 7 Z80 1.311
173. Boris Diplomat - Single Chip 1.300
Fidelity CC 10 Z80 1.300
175. Mephisto Junior - Single Chip 1.294
176. Fidelity Sensory 6 - Single Chip 1.275
177. Novag Savant I Z80 1.273
178. Mephisto Mini - Single Chip 1.268
179. Boris 2.5 - Single Chip 1.267
180. Mephisto I - Single Chip 1.252
181. CXG Sphinx Granada - Single Chip 1.173
182. SciSys Mark II - Single Chip 1.090
183. CXG Sphinx Chesscard - Single Chip 1.087
184. SciSys Mark I - Single Chip 985
CUFFELL TIMOTHY MICH
>184. SciSys Mark I - Single Chip 985
Is this for real? I though knowing all the rule was worthabout 950 points.
Did this machine actually sell in quantity?
--
-Tim Cuffel Finger for PGP 2.1
CIA, FBI, terrorism, assassination, cryptography, RSA, cocaine, DEA, Swiss bank
account.
Peter Stockhausen
Deep Thought II played (and against whom, with what result) to achieve that
2520 FIDE rating?
Peter
--
----------------------------------------------------------------
Peter_St...@mindlink.bc.ca Vancouver, Canada
----------------------------------------------------------------
joev dubach
In article <C0KA9...@hermes.hrz.uni-bielefeld.de> spr...@math10.mathematik.uni-bielefeld.de (Udo Sprute) writes:
>184. SciSys Mark I - Single Chip 985
Is this for real? I though knowing all the rule was worthabout 950 points.
Did this machine actually sell in quantity?
Well, the "950 conjecture" is false. For a counter-example, check out
the computer program Kon on ICS:
"Kon is a computer program, part of an artificial intelligence
project. Its author is Jay Scott, sc...@cs.uiuc.edu, at the
University of Illinois at Urbana-Champaign (Jay on ICS).
The goal of the project is not to produce a strong chess program,
but rather one that learns to improve its own play. Kon is still
in a very early stage of development. It is running now partly as
a reliability test--can it stay up for a long time without crashing?"
Kon has ratings, and records, of
Standard : 483 Blitz : 517
Wins : 21 Wins : 20
Losses : 79 Losses: 191
Draws : 4 Draws : 7
Total : 104 Total : 218
as of the last time I checked. Also, lots of those wins were against
slime, an expert who likes to manufacture artificially low ratings
(currently, he's at 184 standard and 284 blitz.)
The rating system has no natural floor; it is theoretically possible
for someone to have any rating, even a negative rating. Take this
with a grain of salt, and realize that ratings are only a rough
measure of strength.
Joev
--
_____________________________________________________
DDDDDDDDD JJJJJJ | Joev Dubach
D J J | dub...@husc.harvard.edu
D J J | 249 North Mail Center
D D J J | Harvard University
DDDDD * JJJJJJ * | Cambridge, MA 02138
an...@vax.oxford.ac.uk
> In article <1993Jan9.0...@ucsu.Colorado.EDU> cuf...@spot.Colorado.EDU (CUFFELL TIMOTHY MICH) writes:
>
> In article <C0KA9...@hermes.hrz.uni-bielefeld.de> spr...@math10.mathematik.uni-bielefeld.de (Udo Sprute) writes:
> >184. SciSys Mark I - Single Chip 985
>
> Is this for real? I though knowing all the rule was worthabout 950 points.
> Did this machine actually sell in quantity?
>
> Well, the "950 conjecture" is false. For a counter-example, check out
> the computer program Kon on ICS:
...
>
> Kon has ratings, and records, of
> Standard : 483
...
> The rating system has no natural floor; it is theoretically possible
> for someone to have any rating, even a negative rating. Take this
> with a grain of salt, and realize that ratings are only a rough
> measure of strength.
But I though it was agreed that the ratings on ICS were very low, compared with
USCF or FIDE etc. I presume the conjecture about the knowing all rules rating
was meant to mean a FIDE rating or something similar.
--
James Annan.
[---------------------------------------------------------]
[ Major and Lamont have failed to achieve zero inflation. ]
[ Now they're going for zero employment. ]
[---------------------------------------------------------]
CUFFELL TIMOTHY MICH
>In article <DUBACH1.93...@husc10.harvard.edu>, dub...@husc10.harvard.edu (joev dubach) writes:
>> In article <1993Jan9.0...@ucsu.Colorado.EDU> cuf...@spot.Colorado.EDU (CUFFELL TIMOTHY MICH) writes:
>>
>> In article <C0KA9...@hermes.hrz.uni-bielefeld.de> spr...@math10.mathematik.uni-bielefeld.de (Udo Sprute) writes:
>> >184. SciSys Mark I - Single Chip 985
>>
>> Is this for real? I though knowing all the rule was worthabout 950 points.
>> Did this machine actually sell in quantity?
>>
>> Well, the "950 conjecture" is false. For a counter-example, check out
>> the computer program Kon on ICS:
>
>>
>> Standard : 483
ICS is an isolated system. It may even have proprietary Elo rules. In any
case, I was think USCF/FIDE type ratings.
>
>...
>
>> The rating system has no natural floor; it is theoretically possible
>> for someone to have any rating, even a negative rating. Take this
>> with a grain of salt, and realize that ratings are only a rough
>> measure of strength.
>
I know how ratings work. I heard (read?) somewhere that the "rough measure
of strength" of a player just taught the rules is around 950, as compare with
the general pool of chess players. This is not to imply that 950 has any
mystical qualities, anymore than 3000 does at the other end of the scale.
I just thought that a computer playing at 983 or whatever was inconcivable
poor and as unmarketable as a random move generator.
--
-Tim Cuffel Finger for PGP 2.1
Two wrongs don't make a right, but three lefts do.