3 ms·
Solitaire simulator for finding the best strategy: Current record is 8.590%
- advisedwang 7mo agoPer [1] (found via wikipedia) 35% is possible! [1] https://web.archive.org/web/20210924183919/https://www.aaai.org/ocs/index.php/ICAPS/ICAPS09/paper/viewFile/724/1090 https://web.archive.org/web/20210924183919/https://www.aaai....
- reenorap 7mo agoIsn't 8.5% low? My mom didn't use the computer much except she did play solitaire on her Windows laptop all the time. She had over a 2000 game win streak until she got dementia and stopped using the computer altogether.
- embedding-shape 7mo agoSimilarly, my half-sister's mother was almost allergic to anything technology, except for to play Solitaire, which she did every single day. I think many of the games offer configurable "difficulty" though, there are modes where it's guaranteed to be solvable for example. And most of them surely are made slightly easier by default. With a randomly shuffled real deck, wouldn't surprise me that it would be ~10%,.
- blarg1 7mo agomy brother who never really had much interest in computer games, could always get the high score in any game I was playing eg pinball, racing I'd spend ages to get a point above his score, and he'd come back and get an even higher one. He did the same thing to my dad with solitaire, my dad plays it everyday and was proud of his high score. Then my brother who was just visiting, beat my dads highscore by a huge margin, and years later my dad is still annoyed he can't beat it.
- fishtoaster 7mo agoWinning 2000 games in a row sounds statistically unlikely unless the Windows version of solitaire does something behind the scenes to make the game more winnable. https://en.wikipedia.org/wiki/Klondike_(solitaire)#Probability_of_winning https://en.wikipedia.org/wiki/Klondike_(solitaire)#Probabili...
- InitialLastName 7mo agoI suspect that the later (Win 7+) versions of Windows solitaire (and minesweeper, for that matter) did, in fact, cull the unwinnable games.
- jasperry 7mo agoI don't know of any algorithm to cull non-winnable Klondike games. Playing deal-1 instead of deal-3, and with unlimited flipping of the stock, the win chance is probably close to 50%, but that still makes 2000 in a row statistically impossible. My guess is that the poster's mom was actually playing FreeCell, in which nearly every game is winnable and people do get streaks like that.
- npinsker 7mo agoIt’s quite doable, if you don’t mind culling some winnable games too. The object isn’t to have a perfect classifier.
- chocochunks 7mo agoYou don't need an algorithm. You can just record seeds that are solvable. The current version of Klondike in MS Solitaire is winnable unless you play "Random" difficulty.
- dezgeg 7mo agoCould there be mixup with FreeCell?
- reenorap 7mo ago
- InitialLastName 7mo agoDraw 1 is much more winnable than draw 3. With perfect knowledge (or an infinite undo stack), evidently ~80% of Klondike games are winnable. With imperfect knowledge but good strategy, humans win about 11% of draw 3 games. So given they have implemented a more rudimentary strategy (first come, first serve), 8.5% doesn't seem that low.
- dacracot 7mo agoYes, it is low, but that is the challenge to fork the repo and up the winning percentage.
- stronglikedan 7mo agoLow for a human, yes. But for a computer simulation, it's a record high.
- Hackbraten 7mo agoWhat is 8.590% of what?
- joseda-hg 7mo agoChance of winning a random deck
- rotexo 7mo agoI love seeing this. Klondike with a physical deck was my “I need to do something with my hands or I’ll go crazy” release valve during the pandemic after work. I thought about trying to build something like this, but it was far outside of my technical ability. I settled for a simulator that would run toilet solitaire so I could see the distribution of cards remaining after games with a well-shuffled deck.
- dacracot 7mo agoNew PR with unconfirmed claim of 12.188%. Will review and confirm soon.