I'm primarily concerned with deals that require longer move solutions.
As such, I've only encountered a few of the shorter move solutions. When
I ran my solver against the first 500 deals, the shortest were 28
atomic-moves for deals #58, #224, and #378. It's possible that
converting their solutions to meta-moves would come close to your 21
move criteria.
Here are move results for your four problems using: the
Woods-Callan-Ettlinger solver in FcPro26, Shlomi Fish's FreeCell Solver
2.8.6 (-opt -to 01ABCDE), Dr. Tom Holroyd's PatSolve 3.0 (-E -M64), and
my basic solver -- BFS-020.
I'm not sure, but the W-C-E solver in FcPro26 probably uses meta-moves
instead of atomic-moves.
Deal FcPro FCS Pat BFS
------- ----- --- --- ---
#002755 77 526 30 24 (might be 21 meta-moves or less.)
#004010 77 192 42 26 "
#005290 226 137 43 25 "
#010156 106 137 61 42
Note: Typically, the output from FCS has to be massaged before it's
suitable for entry into MS-FreeCell or FcPro26. FcPro26 has an internal
implementation of FCS that works okay, but the FCS output from the
solver is not suitable for input. I wrote a converter that takes the
output from FCS and removes the auto-moves. However, some of the
remaining moves are interpreted as meta-moves by other solvers. The FCS
move counts above are after removing auto-moves.
If you ever get bored, try solving Michael Keller's favorite deal #80488
(I believe). My longest solution (so far) goes to deal #57148 ... with
#201616 a close second.
Happy Solving, Danny A. Jones
WKRfresno_at_aol.com wrote:
> Havn't looked above 32000, but here are a few short games that may be of
> interest. Single card moves only.
> 10156, 4010, 5290, 2755. How many moves do the various solvers require?
>
> Does anyone know of a game in MS32K that can be solved in fewer than
> 21 moves
> (using single- and multiple-card moves)?
>
> Bill Raymond
Received on Wed Sep 24 2003 - 17:36:30 IDT