Rating + - Games Won Oppo
------ --- --- ----- --- ----
1 Rebel 8.0 Pentium 90 MHz 2473 30 -28 632 69% 2334
* 2 MChess Pro 6.0 Pentium 90 MHz 2459 37 -36 383 63% 2364
3 Genius 3.0 Pentium 90 MHz 2416 27 -27 691 62% 2330
3 Genius 4.0 Pentium 90 MHz 2416 26 -25 769 65% 2310
5 Rebel 6.0 Pentium 90 MHz 2415 31 -30 540 60% 2344
6 MChess Pro 5.0 Pentium 90 MHz 2414 27 -26 740 63% 2321
7 Rebel 7.0 Pentium 90 MHz 2411 27 -27 691 61% 2334
8 Hiarcs 4.0 Pentium 90 MHz 2394 28 -28 619 56% 2353
9 Chessmaster 5000 Pentium 90 MHz 2392 52 -48 212 67% 2268
9 Nimzo 3.0 Pentium 90 MHz 2392 28 -27 643 59% 2328
11 Hiarcs 3.0 Pentium 90 MHz 2382 30 -30 545 56% 2337
12 Genius 4.0 486/50-66 MHz 2380 29 -29 588 57% 2327
13 MChess Pro 4.0 Pentium 90 MHz 2363 29 -29 578 52% 2347
14 Genius 3.0 486/50-66 MHz 2360 24 -23 890 63% 2270
15 Fritz 3.0 Pentium 90 MHz 2357 28 -28 627 54% 2331
16 R30 v. 2.5 2356 52 -48 215 68% 2226
17 Rebel 7.0 486/50-66 MHz 2347 29 -29 573 55% 2314
18 MChess Pro 5.0 486/50-66 MHz 2343 30 -30 540 54% 2312
19 Mephisto Genius 2.0 486/50-66 MHz 2341 23 -23 917 58% 2282
20 Kallisto 1.98 Pentium 90 MHz 2335 30 -30 526 49% 2341
21 WChess 1.06 Pentium 90 MHz 2328 28 -29 600 48% 2342
22 MChess Pro 4.0 486/50-66 MHz 2319 24 -24 826 54% 2289
23 Fritz 4.0 Pentium 90 MHz 2316 46 -44 243 60% 2243
24 WChess 1.03 486/50-66 MHz 2305 27 -27 677 49% 2311
25 Chess Machine 30-32 MHz Schröder 3.1 2303 32 -30 546 68% 2172
25 Hiarcs 3.0 486/50-66 MHz 2303 25 -25 786 52% 2287
27 Rebel 6.0 486/50-66 MHz 2302 24 -23 884 55% 2268
28 Meph Genius 68 030 33 MHz 2300 47 -46 228 57% 2253
29 Ch.Machine 30 MHz King 2.0 aggr/R30 off 2296 21 -21 1153 65% 2185
30 Chessmaster 4000 486/50-66 MHz 2293 34 -33 462 66% 2173
31 Chess Genius 1.0 486/50-66 MHz 2288 23 -23 931 54% 2260
32 Rebel Decade Pentium 90 MHz 2284 29 -30 563 45% 2321
33 Chess Machine 30 MHz Schröder 3.0 2274 62 -55 168 73% 2102
34 MChess Pro 3.5 486/50-66 MHz 2273 23 -23 920 53% 2255
34 Mephisto Gideon Pro 486/50-66 MHz 2273 35 -34 421 62% 2184
36 MChess Pro 3.12 486/50-66 MHz 2265 36 -34 438 69% 2123
37 Fritz 3.0 486/50-66 MHz 2258 22 -22 975 52% 2246
38 Mephisto Lyon 68030 36 MHz 2252 29 -27 676 70% 2101
39 Chess Genius 1.0 486/33 MHz 2244 32 -32 463 50% 2244
40 Mephisto Portorose 68030 36 MHz 2238 40 -36 407 77% 2028
41 Mephisto Vancouver 68030 36 MHz 2233 37 -34 451 73% 2057
42 MChess Pro 3.12 486/33 MHz 2227 50 -48 208 60% 2153
43 Comet32 Pentium 90 MHz 2224 37 -39 349 36% 2325
44 Berlin Pro 68 020 24 MHz 2220 25 -25 791 59% 2154
45 Kallisto 1.82-1.83 486/50-66 MHz 2218 22 -23 977 43% 2265
46 Kasparov SPARC 20 MHz 2214 26 -25 747 52% 2201
47 Mephisto RISC 1 MB ARM 2 14 MHz 2202 21 -21 1132 57% 2152
47 Hiarcs Master 2.0 486/33 MHz 2202 46 -46 229 51% 2193
49 Chess Machine Schroder 512K ARM2 16MHz 2194 27 -26 703 62% 2110
49 MChess 1.1-1.71 486/33 MHz 2194 44 -40 326 74% 2015
51 Saitek RISC 2500 ARM2 14 MHz 128K 2192 22 -21 1062 56% 2148
52 Mephisto Montreux ARM 14 MHz 512K 2188 32 -30 586 73% 2013
53 Chess Machine The King 512K ARM2 16MHz 2176 32 -32 473 55% 2137
54 Mephisto Vancouver 68020 12 MHz 2160 23 -22 997 66% 2046
55 Mephisto Lyon 68020 12 MHz 2151 19 -18 1442 59% 2087
56 Socrates 3.0 486/33 MHz 2142 49 -50 203 47% 2163
57 Fritz 2.0 486/33 MHz 2133 30 -31 527 45% 2167
58 Mephisto Portorose 68020 12 MHz 2127 26 -24 896 73% 1956
59 M Chess 1.1-1.71 386/25-33 MHz 2126 36 -35 408 65% 2017
60 Fidelity Elite 68030 32 MHz (vers.9) 2122 38 -36 396 70% 1977
61 Mephisto Berlin 68 000 12 MHz 2117 24 -23 903 58% 2063
62 Mephisto Lyon 68000 12 MHz 2104 22 -21 1068 58% 2047
63 Mephisto Vancouver 68000 12 MHz 2102 22 -22 1020 56% 2060
64 Mephisto MM 4 Turbo Kit 6502 16 MHz 2092 46 -42 292 73% 1918
65 Novag Sapphire H8 10 MHz 2088 23 -23 928 49% 2096
66 Fidelity Mach IV 68020 20 MHz 2073 19 -19 1378 51% 2067
66 Mephisto Almeria 68020 12 MHz 2073 42 -40 305 62% 1985
68 Hiarcs Master 1.0 486/33 MHz 2071 48 -48 214 48% 2087
69 Mephisto Portorose 68000 12 MHz 2045 26 -25 786 63% 1953
70 Gandalf 2.1 Pentium 90 MHz 2044 47 -53 228 27% 2217
71 Mephisto Polgar 6502 10 MHz 2042 43 -41 283 59% 1979
72 Fritz 1.0 486/33 MHz 2041 48 -47 215 55% 2008
73 Fidelity Elite 68000 x 2 (vers. 5) 2036 49 -48 209 57% 1987
74 Nimzo 2.2.1 486/33 MHz 2034 46 -47 229 42% 2089
75 Zarkov 3.0 486/25-33 MHz 2030 46 -48 232 39% 2109
76 Mephisto Roma 68020 14 MHz 2029 27 -26 712 65% 1919
76 Rex Chess 2.3 386/25-33 MHz 2029 65 -62 126 59% 1964
78 Mephisto Dallas 68020 14 MHz 2022 35 -32 492 72% 1861
79 Mephisto Almeria 68000 12 MHz 2020 32 -30 534 67% 1898
79 Zarkov 2.5 386/25-33 MHz 2020 56 -53 168 61% 1941
81 Kasparov Brute Force H8 10 MHz 2019 23 -23 906 44% 2059
82 Fritz 1.0 386/25-33 MHz 2007 66 -67 113 46% 2038
83 Novag Diablo 68000 16 MHz 2005 21 -21 1103 41% 2072
84 Fidelity Mach III 68000 16 MHz 1993 14 -14 2410 52% 1980
85 Complete Chess System 486/33 MHz 1984 47 -47 221 47% 2007
86 Mephisto MM 5 6502 5 MHz 1977 19 -19 1314 48% 1990
87 Kasparov President/GK-2100 H8 10 MHz 1975 29 -30 558 47% 1998
88 Mephisto Dallas 68000 12 MHz 1972 21 -21 1158 61% 1890
89 Mephisto Roma 68000 12 MHz 1970 18 -18 1519 56% 1927
90 Mephisto Polgar 6502 5 MHz 1969 17 -17 1733 41% 2033
91 Mephisto Milano 6502 5 MHz 1965 24 -25 820 42% 2025
92 Novag Super Expert C 6502 6 MHz Sel 5 1961 18 -18 1532 46% 1989
93 Zarkov 2.6 386/25-33 MHz 1960 64 -62 125 56% 1917
94 Mephisto Academy 6502 5 MHz 1937 18 -18 1495 44% 1979
95 Chessmaster 3000 386/25-33 MHz 1932 68 -74 103 36% 2031
96 Rex Chess 2.3 386/16-20 MHz 1930 53 -53 174 49% 1936
97 Saitek Leonardo Maestro B 6502 18 MHz 1928 73 -67 105 62% 1842
98 Mephisto Amsterdam 68000 12 MHz 1926 22 -22 1020 58% 1872
99 Novag Super Expert B 6502 6 MHz sel 4 1919 29 -29 562 48% 1934
100 Mephisto Mega IV 6502 4.9 MHz 1917 18 -18 1544 44% 1956
101 Saitek Maestro D 6502 10 MHz 1915 26 -26 736 51% 1909
102 Fidelity Excel Mach IIc 68000 12 MHz 1914 23 -23 955 50% 1911
103 Novag Expert Turbo Kit 6502 16 MHz 1910 67 -63 120 61% 1829
104 Mephisto MM 4 6502 5 MHz 1903 19 -19 1288 51% 1900
105 Fidelity Travel Master H8/330 10 MHz 1897 65 -79 123 22% 2112
106 Kasparov GK-2000 H8 10 MHz 1896 29 -29 593 42% 1952
106 Mephisto Modena 6502 4 MHz 1896 28 -29 615 41% 1959
108 Psion Atari 68000 8 MHz 1882 18 -18 1487 44% 1928
109 CXG Sphinx Galaxy 6502 4 MHz 1881 17 -18 1670 38% 1971
110 Novag Ruby H8 10 MHz 1878 30 -30 545 42% 1934
110 Fidelity Excel Mach IIa 68000 12 MHz 1878 46 -47 226 47% 1898
112 Saitek Turboking II 6502 5 MHz 1868 23 -23 963 37% 1964
113 Conchess Plymate Victoria 6502 5.5 MHz 1867 26 -27 701 40% 1941
114 Fidelity Excel Club 68000 12 MHz 1857 23 -23 931 51% 1848
115 Mephisto Mega IV Brute Force 1851 36 -37 382 40% 1921
116 Novag Super Expert/Forte A 6502 5 MHz 1836 19 -20 1318 39% 1917
117 Fidelity Par Excellence 6502 5 MHz 1835 22 -22 1022 48% 1846
118 Fidelity Avant Garde 6502 5 MHz 1834 20 -20 1196 46% 1866
119 Mephisto Rebell 6502 5 MHz 1825 19 -19 1373 39% 1907
120 Saitek Stratos/Analys B 6502 6 MHz 1820 20 -21 1213 38% 1908
121 Mephisto Super Mondial 6502 4 MHz 1818 19 -20 1335 37% 1908
122 Novag Forte B 6502 5 MHz 1816 24 -24 837 44% 1860
123 Saitek Leonardo Maestro A 6502 6 MHz 1811 30 -30 547 45% 1846
124 Novag Forte A 6502 5 MHz 1809 22 -22 964 48% 1821
125 Conchess Plymate 6502 5.5 MHz 1808 15 -16 2040 40% 1881
126 Saitek Simultano/Corona C 6502 5 MHz 1806 23 -24 969 33% 1932
127 Fidelity Excellence 6502 4 MHz 1800 18 -18 1565 43% 1852
128 Novag Expert 6502 4 MHz 1788 24 -24 843 42% 1847
129 Fidelity Elegance 6502 3.6 MHz 1780 40 -40 303 54% 1751
130 Conchess Plymate 6502 4 MHz 1779 39 -38 333 55% 1747
131 Mephisto MM 2 6502 3.7 MHz 1773 49 -48 208 52% 1756
132 Saitek Turbostar 432 6502 4 MHz 1764 22 -22 1005 45% 1802
133 Fidelity Excellence 6502 3 MHz 1757 22 -23 960 42% 1815
134 Saitek Kasparov Blitz 1734 45 -47 239 37% 1825
134 Novag Super Nova 1734 37 -38 350 40% 1802
136 Novag Super Constellation 6502 4 MHz 1731 18 -18 1626 35% 1840
137 Conchess Glasgow 6502 4 MHz 1719 39 -39 327 50% 1722
138 Chessplayer 2150 Atari/Amiga 1714 58 -63 146 33% 1837
139 The Final Chesscard 6502 5 MHz 1700 57 -64 156 27% 1871
140 Novag Super VIP 1688 53 -58 174 32% 1816
141 Chessmaster 2100 Amiga 68000 1687 65 -76 120 25% 1874
142 Mephisto Europa 1682 50 -52 190 42% 1740
143 Saitek Superstar 36K 6502 2 MHz 1678 23 -24 954 31% 1818
144 Fidelity Elite A/S 6502 3.2 MHz 1677 23 -24 1023 29% 1837
145 Chess Champion 2175 Atari/Amiga 1672 56 -62 157 30% 1821
146 Conchess Glasgow 6502 2 MHz 1667 24 -25 855 34% 1781
147 Novag Quattro 6502 4 MHz 1662 30 -31 560 33% 1790
148 Fidelity Prestige 6502 4 MHz 1660 64 -66 118 44% 1702
149 Chessmaster 2000 Atari 68000 8 MHz 1654 41 -45 312 25% 1840
150 Novag Constellation 6502 3.6 MHz 1647 31 -32 504 41% 1714
151 Novag Constellation Primo 1637 57 -59 149 44% 1683
152 Novag Constellation 6502 2 MHz 1596 32 -33 479 35% 1706
153 CXG Super Enterprise 1562 36 -39 386 28% 1726
Rating : ELO rating
+/- : deviation
Games : number of games played
Won : percentage of games won
Oppo : medium opponent ELO rating
By Thoralf Karlsson,
thoralf.karlsson@mailbox.swipnet.se Chairman of the SSDF
and Göran Grottling, goran.grottling@mailbox.swipnet.se Treasurer of the SSDF
MChess Pro 6.0 Pentium 90 MHz is the only new entrant on this last list of 1996. After 383 tournament games it has got a rating of 2459, which is 45 points more than MCP 5 has on the same hardware!
In spite of the loss against Marty Hirsch's new program, Rebel 8.0 P90 is still the leader of the list with a rating of 2473. The 64 games which have been played since last list has lowered the rating with nine points. Many of the other programs have also changed a few points, up or down.
We must realize that the rating figures are not absolutely fixed. Even if as many as 600 games have been played, we can only say that the "true rating", with 95% probability, is somewhere inside a range of about sixty points (+/- 30). Due to the fact that the ratings in the top of the list doesn't differ so many points, small changes can affect the individual order of the programs rather much.
Lately there has been much discussion about the SSDF Rating list on the newsarea rgcc on the Internet. The claims have been that some opening books give misleadingly high ratings. The suggested cures have been to disqualify games played with certain openings, not to count repeated games, to play without opening books or from some (secret) starting positions. The aim of the suggested changes is that the ratings better should reflect the "true" strength, which the achievements of the chess engine is considered to be.
We admit that special preparations against other chess programs could give a higher rating, at least temporarily until new programs arrive. But based on own investigations we believe that it is a question of only 20-40 points. Winning lines against program A can be loosing lines against program B! Those who read our list should anyhow be aware of the possibility that the rating of a new entrant could be somewhat lower after a year or so due to this phenomenon.
Is there a practical way for SSDF to eliminate this (probably rather small) uncertainty of the ratings? No, we don't think so. There are several strong, both theoretical and practical, arguments against the above mentioned methods.
The best solution to this problem is that the programmers provide a better variety in their opening books and general play and that they implement learning functions. This minimizes the effect of the kind of openings which has been discussed on rgcc.
Partly in order to hasten the effects of these suggested changes, we have decided to play more games between the latest program versions and less (or none at all) against the oldest versions. Another reason for this change is a wish for a better balance between different programmers in our "testpool".
SSDF provides information based on the outcome of tournament games played between chess programs, using the opening books their programmers have adviced us to use. For those who want to study the games and the openings, we plan to create a database (probably in pgn-format) which will be available in some way on the Internet.
During the next months we will play with the new programs Genius 5.0, Hiarcs 5.0 and Junior.