Author: Ralf Elvsén
Date: 15:06:14 07/15/00
Looking through old issues of "Tidskrift för Schack", a monthly publication from the Swedish Chess Federation, I found this announcement in the November 1978-issue: "Thursday 30 Nov 1978, CHESS 4.5 - the worlds strongest chess computer program - will play an exhibition game against the Swedish radio listeners." Then the history of CHESS and its achievements is given. Its rating is estimated to 2000, though which kind of rating is unclear. Then comes a short description of the alpha-beta search. CHESS is said to consider "mobility (number of attacked squares), pawn structure (passed pawns, double pawns, isolated pawns etc.), placement of the pieces (rooks on the 7:th row) and king safety (number of pieces and pawns in the immediate vicinity). In the endgame an algorithm is triggered which encourages the king to approach pawn formations and in positions where CHESS has a great advantage he goes for a king side attack." "CHESS has an opening book with 5000 positions. If a known position is encountered, CHESS will immediately play the recommended move without calculation. In the exhibition game CHESS will be given 30 s /move on average. This will allow it to examine 10 000 positions in order to choose a move based on a 4-5 ply search CHESS will run on a 6600-computer at [the location]. ... In parallel with the chess program the computer will run programs for [a lot of other stuff] and at the same time try to find matches between 3000 unemployed and a database of 25 000 free positions. Everybody will think the computer is working on 'their' problem." I'm just trying to translate... In the next issue the game is given along with the identity of the persons responsible for the various moves. [Event "Exhibition game"] [Site "?"] [Date "??.??.??"] [Round "?"] [White "CHESS 4.5 (6600 computer)"] [Black "Swedish radio listeners"] [Result "1-0"] [ECO "C43"] [PlyCount "67"] [EventDate "1978.11.30"] 1. e4 e5 2. Nf3 Nf6 3. d4 exd4 4. e5 Nd5 5. Bg5 Be7 6. Bxe7 Qxe7 7. Qxd4 Nb6 8. Bb5 a6 9. Bd3 Nc6 10. Qe4 d5 11. exd6 Qxe4+ 12. Bxe4 cxd6 13. O-O Be6 14. Ng5 h6 15. Nxe6 fxe6 16. Nd2 O-O-O 17. f4 Kc7 18. a4 Nd7 19. c4 Nd4 20. Rfe1 a5 21. Nf3 Nxf3+ 22. Bxf3 e5 23. c5 exf4 24. cxd6+ Kb6 25. Re7 g5 26. Rae1 Nc5 27. Rd1 Rhf8 28. Rd4 Rf6 29. b4 axb4 30. Rxb4+ Ka7 31. Rb5 Nxa4 32. Rbxb7+ Ka6 33. Be2+ Ka5 34. Rb5+ 1-0 "That CHESS 4.5 would win the game was no big surprise but we were very impressed by the computers strong finish of the game." ... "More than 1000 listeners tried to suggest moves."... "Impressive finish - note in particular moves 23 and 31 - none of us in the studio saw them coming so CHESS certainly played strong." ... "CHESS saw the position as equal after move 16, it saw itself up with 0.5 pawns after move 24, 1.1 pawns after move 30 and 6.3 pawns after black's move 33. In the final position CHESS gave the evaluation 1 254.4 pawns, so now we know how much a mate is worth." Among the people responsible for the various moves I recognize two: Pia Cramling (move 18), now one of the strongest female players, GM and at the moment (15 July 2000) leading the Swedish championship. Konstanty Kaizauri (move 33), at the time a strong player. There are probably more strong players I don't know about and several patzers as well. The identity of the person who suggested what apparently is considered to be black's worst mistake (22.... e5) is kept as a secret by the author of the article :) To be fair to this unknown person both Hiarcs and Crafty thinks the position is even after 22 ... e5 and prefers 23. fxe5 . After 23. c5 they think black is OK after 23. ...Nxc5 . 23. ... exf4 seems to be the first non-trivial mistake by black.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.