Is it possible to write a script to tell if a given flower garden game is able to be won?
lcc123
REGISTERED
lcc123
REGISTERED
lcc123
REGISTERED
Comments
I just look at "high scores." If they show the game is not being won, it probably can't be won.
@barzonymous, someone has to be first!
And someone finally was. T'was not me.
@lcc123 That game, in your post, is winnable. The only way to prove that a game is able to be won, or not, is for someone to win it.
@barzonymous, sometimes they can, if you have more perseverance than the previous players.
The problem with the high scores that I see is that it would seem that most people go for quick times and then give up after a few minutes of attempts (at least on flower garden game). So for most games I have seen--while they have been played by multiple people, they have have not been solved. Nevertheless, if enough time/effort is spent, it would seem that most can actually be solved. Here is a good discussion around the methodology of a program that was written to test this on the original windows 32000 spider solitare games:
https://www.tranzoa.net/~alex/plspider.htm
The programming of the greenfelt website and multiple games seems fairly complex, and I figured if they were able to get it coded correctly and the game rules correct, it wouldn't be too much harder to write a simulation program that could play the game say 100 or 1000 times (or more for difficult games) and see how many tries it took for the computer to solve the games (if it was actually solvable) and then maybe put a "rating" at the top of the screen with the level of difficulty in a range of say 0-100%. With 100 being unsolved and the lower the rating the fewer tries it took the computer to complete. That way the people that are looking for a challenge could play those harder games and those looking for a more likely win could skip to others.