I have an 8x4 MS Freecell solver that does a recursive search for
shorter solutions to a deal. However, the solutions lack any
resemblance to how a real player might have ordered the same moves.
I've tried a couple of modest ideas to post-process solutions and make
them more 'human'.
The first idea was to take the moves to home cells and perform them
ASAP. I then ordered the remaining moves based on moving cards with
largest face values first whenever possible. This approach was of
limited success.
The next idea still involved moves to home cells being performed ASAP,
but I then moved cards to empty columns and free cells before moving
them to an occupied column. Again, moves were ordered by card face
values in each grouping. This approach was also of limited success.
At this point, I realized that I wasn't going to get off easy on
'humanizing' my solutions. So, I decided that the best approach would
be to solicit constructive suggestions from others.
One way to get started would be to take a solution and discuss its
shortcomings. Suggestions for other ways are welcome.
Anyone interested in participating in this discussion?
Received on Tue Oct 11 2005 - 11:38:39 IST