6 of 6 people found the following review helpful
1.0 out of 5 stars
Very unimportant research., July 7 1999
By A Customer - Published on Amazon.com
This review is from: Beautiful Mates: Applying Principles of Beauty to Computer Chess Heuristics (Paperback)
This dissertation investigations problems of the mate-in-two variety, and shows that a few heuristics that correspond to the human concept of "beauty" tend to reduce solution times (10% to 33%). My disappointment arises because this is not a difficult problem; there isn't a chess program on the marketplace that can't solve mate-in-N problems in the blink of an eye. Furthermore, the advantage obtained through these heuristics is very minor; on a small sample of positions it is always possible to tune one's chess engine to obtain comparable speed-ups. My conclusion is that the author's time at the University of Sussex was not well spent.
5 of 6 people found the following review helpful
1.0 out of 5 stars
Disappointing, July 17 1999
By A Customer - Published on Amazon.com
This review is from: Beautiful Mates: Applying Principles of Beauty to Computer Chess Heuristics (Paperback)
This is a book bound version of a simple article comparing simple chess heuristics and simple + "beauty" heuristics as an attempt to emulate human chess thinking, where the "beauty" heuristics follow now outdated ideas in human 'intuition". The resulting programs are subjected to relatively simple mate in 2, 3 and the odd mate in 4 and 5 chess problems. The resulting time and nodes analysed between the two systems are probably statistically insignificant as only 24 problems are studied. An interesting read, but by no means revolutionary.
3 of 5 people found the following review helpful
4.0 out of 5 stars
Original and Interesting - when is there a follow up??, Sep 16 1999
By A Customer - Published on Amazon.com
This review is from: Beautiful Mates: Applying Principles of Beauty to Computer Chess Heuristics (Paperback)
This is an orginal book, following an interesting proposal to get Computer chess games to 'think' more like human players. It is clear the author values this approach more than current 'brute force' search, and I must admit, I am swayed by his arguments (e.g. that the Deep Blue v Kasparov match did nothing to advance computer chess) - although I could have done with more results to convince me. I think the reader from Concord, MA missed the point of the book - the author never says he is proposing a 'better' (i.e. higher ELO) chess program - just one that thinks and plays more like a human (from an AI research perspective). To conclude, if you are looking for a book about conventional (outdated?) computer chess programming then this is of little interest. However, if you are after a novel book on the quest to 'capture' what goes on in the human mind when analysing problems, and how to represent that in, for example, chess problems, then this book is a good starting point.