Author: Russell Reagan
Date: 00:47:12 01/13/05
Go up one level in this thread
On January 12, 2005 at 21:09:16, chandler yergin wrote: >How do you now the "Best Move" until you have calculated them ALL? > >Hmmm? You do have to examine every move, but you don't have to examine every move *completely* to know which move is the best move. Let me ask you a different question. Who has the most money: Bill Gates, Tiger Woods, or me? I can get the "best answer" in two different ways. The simple way, brute force: I could examine each person's finances *completely*. I could drive to the library and do research all day and get a close estimate of how much money Bill Gates and Tiger Woods have. The next day I could drive to my bank and wait in line to see exactly how much money I have. After I know how much money each of us has, I can compare the numbers and determine who has the most money. The smarter way, lazy evaluation: I get on the internet and see that Bill Gates has many billions of dollars. I see that Tiger Woods has many millions of dollars, but less than one billion dollars. I don't know exactly how much I have, but I know it is less than one thousand dollars. Now I compare these, and I see that Bill Gates must have the most money. Using the smart way, it doesn't matter if I know exactly how much money each of those people have, as long as I have upper and lower bounds. Bill Gates has at least one billion dollars. I have less than one billion dollars. Therefore he has more money. In the same way, the computer program doesn't have to know exactly how good or bad each move is. If it knows that move A wins at least one pawn, and that move B loses at least one rook, then it can stop examining move B. Move A might actually win three pawns, and move B might actually lose two rooks and a queen, but we don't need to know that to know that move A is better.
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.