I don't know how many people on the group have actually written
solvers (I am in regular contact with Danny A. Jones and Gary
Campbell, who have both provided material for the FreeCell FAQ
on my site). Here is a research question:
We have pretty accurate figures for the win rates for FreeCell with
zero freecells, with the standard eight columns. But what are the
rates for wider games, from 9 columns all the way up to 13? These
games can be played as variant games on NetCELL (www.freecell.com).
From the stats available there for top players
(<
http://www.freecell.com/f/c/scores.html>), I'm going to make a wild
guess at the win rates with perfect play:
9 columns 8%
10 columns 35%
11 columns 71%
12 columns 92%
13 columns 98%
Would anyone like to take a crack at analyzing some or all of these
variants, by running at least the first 32,000 standard MS deals?
The 10x0 game appears in Pretty Good Solitaire (www.goodsol.com)
under the name Vineyard, though the game's inventor, Peter Voke,
wanted two red aces to be put on the foundations at the start and the
other 50 cards dealt in ten columns of five. How would this affect
the win rate?
A new variant called Winery adds two freecells, and places two cards
in them initially, dealing the other fifty as ten columns of five. (By
the way, what would be the win rate in regular 8x4 FC if you deal four
cards initially to the freecells and the rest as eight columns of four?
This would be exactly like ForeCell, but with any card being allowed
to be placed in an empty column, as usual in FC).
Michael Keller
FreeCell FAQ
Solitaire Laboratory
Received on Wed Oct 19 2005 - 13:51:49 IST