CHAOS was one of the leading programs since it first appeared at ACM 1973 until the mid 80s. CHAOS, which stands facetiously for Chess Heuristics And Other Stuff, has participated in twelve ACM tournaments[1] and four world championships[2] , runner up in 1974 behind Kaissa, defeating favorite Chess 4.0. In 1980 at the 3rd WCCC, CHAOS was close to becoming champion but lost the playoff against Belle. Its last tournament was the 16th AMC 1985 when it lost two games in a very strong field.
CHAOS was written in Fortran and required in excess of 3,000,000 words of memory to execute, using most of it for storing the tree. It examines only about 50 nodes/sec (1985 70 nodes/sec [5] ) or about 10,000 per move. The reason why CHAOS evaluated so slowly — their programmers did a better job of evaluation! Because every minimax decision is made based on a single evaluation, it seemed plausible that the more accurate this number is, the better the quality of tree searching. However, that comes at the cost of missing deep combinations from brute force search, but CHAOS's strength/nodes-evaluated ratio was impressive. The program carried out a selective search with iterative widening, a bit different from the others. Its book contains about 10,000 lines.
Quotes
Alex Bell
Alex Bell on the CHAOS - Chess 4.0 game 1974 [6] :
CHAOS played 16 NxP!! - a move which has been acclaimed as the finest ever made by a computer. White evaluates that his domination of open lines is compensation for a piece. This judgement is absolutely correct. Of course the piece is not sacrificed entirely and play continues in a very similar fashion to MASTER's sacrifice of a bishop to TELL.
rq2k2r/3n1ppp/p2bpnb1/8/Np1N4/1B3PP1/PP2Q2P/R1BR2K1 w kq - 1 16
Duel on the Chessboard
Quote from Computer vs. computer: Duel on the Chessboard[7] on ACM 1979:
The biggest and most powerful computers do that very well. In one second, they can examine thousands of possible moves. The problem is, they stop to consider lousy moves that a human player wouldn't waste a fraction of a second on. On the other side of the fence are the slower but "smarter" computer programs. They can't think about zillions of chess moves, so they need a lot of information about chess plugged into them. CHAOS is one of these latter, pumped with chess information from John J. O’Keefe, one of Michigan's top players.
Table of Contents
History
CHAOS started in the late 1960s at RCA Systems Programming division in Cinnaminson, NJ with Ira Ruben, Fred Swartz, Victor Berman, Joe Winograd, and William Toikka as first authors. In January 1972, Sperry Univac (1986 Unisys) officially took over the RCA Computer Systems Division and eventually moved everything to Blue Bell, Pennsylvania. Later the program has been affiliated with the Computing Center of the University of Michigan where most of its programmers were staff members [4] .Knowledge vs Search
CHAOS was written in Fortran and required in excess of 3,000,000 words of memory to execute, using most of it for storing the tree. It examines only about 50 nodes/sec (1985 70 nodes/sec [5] ) or about 10,000 per move. The reason why CHAOS evaluated so slowly — their programmers did a better job of evaluation! Because every minimax decision is made based on a single evaluation, it seemed plausible that the more accurate this number is, the better the quality of tree searching. However, that comes at the cost of missing deep combinations from brute force search, but CHAOS's strength/nodes-evaluated ratio was impressive. The program carried out a selective search with iterative widening, a bit different from the others. Its book contains about 10,000 lines.Quotes
Alex Bell
Alex Bell on the CHAOS - Chess 4.0 game 1974 [6] :Duel on the Chessboard
Quote from Computer vs. computer: Duel on the Chessboard [7] on ACM 1979:Authors
Photos & Games
CHAOS Team
From left, Fred Swartz (rear), Vic Berman (front), Bill Toikka, Ira Ruben (seated), Joe Winograd [12]
WCCC 1974
WCCC 1980
See also
Forum Posts
External Links
Chess Program
Misc
References
Ben Mittman, Monroe Newborn (1980). Computer chess at ACM 79: the tournament and the man vs. man and machine match. Communications of the ACM, Vol. 23, Issue 1, pdf from The Computer History Museum » ACM 1976
What links here?
Up one level