|
|
#1
08-17-2007, 03:12 PM
|
|||
|
|||
Checkers Has Been Solved
|
|
#3
08-17-2007, 04:04 PM
|
|||
|
|||
|
Re: Checkers Has Been Solved
[ QUOTE ]
Wow. I didn't know they were close to doing that. Kind of sad in a way as well as interesting. [/ QUOTE ] Heh. It's funny that you responded this way. I was shocked that it hadn't been done already.
|
|
#5
08-17-2007, 04:33 PM
|
|||
|
|||
|
Re: Checkers Has Been Solved
finally.
|
|
#6
08-17-2007, 06:41 PM
|
|||
|
|||
|
Re: Checkers Has Been Solved
This result was posted earlier in other forums. That article is wrong in several places.
The endgame database is of 10 or fewer pieces, not fewer than 10. These positions were solved, not played out. There aren't 19 opening moves, obviously. There were 19 openings solved out of more than 100 3-move openings played in tournament checkers (where players have to play a game from each side of a 3 move opening). Tournament checkers is not solved. The article was correct in that not all positions were searched. Several other newspaper articles falsely claimed that they were, and Wikipedia repeated this error. However, the reason given was that positions were equivalent, which is absurd. This reduces the size of an endgame database very slightly, but it doesn't drop the number of positions from about 10^21 to 10^14, as there aren't 10^7 symmetries. It's simply not necessary to search through all positions to establish the result.
|
 |
|
|
Linear Mode
