Designing Solitaires

Lately I’ve been addicted to the solitaire games from Zachtronics and the Global Game Jam offered an opportunity to try our hands at creating our own game for the genre. Here, I try to collect some lessons I’ve learned about solitaire design in the meantime.

Decisions

A great threat of solitaire games and perhaps a stigma on the genre is that playing these games don’t require much thought from the player and they exist as a last resort of a bored office worker to rush the clock forward. This view is probably somewhat informed by the randomness of Klondike (the variant popularized by the classic game shipped with Windows), the game has a fairly low solvability rate of around 52%. But solvability in itself might matter less than the ability to make decisions. Thanks to not knowing what cards lie behind your stacks, classic Klondike involves a lot of guess-work and shuffling through your deck, with the gameplay often consisting of the player just automatically placing whatever they can on stacks, essentially boiling it down to being a game of pattern-recognition instead of decision-making.

In opposition there is FreeCell, which features four free cells where you can store a single card each and most importantly, all the cards are laid out face up, so you understand the entire deal which allows you to formulate a plan of order in which you can try going about unlocking certain cards and stacks. This variant has a very high solvability rate of 99.999%, but it doesn’t mean the game is easier, in fact FreeCell feels more like a game where your - often tough - choices and skills can grant you wins instead of the random deal determining whether you can act the part of a successful card placer bot.

Freedom and bubble struggle

My first reaction to pondering these issues was that the problem must be about the lack of freedom in Klondike as you barely have choices while in FreeCell any card can be placed in the free cells at any moment. So if we just add mechanics that open up the possibilities, we get better solitaires. Granted, there is some truth to this notion but it’s not entirely the right avenue to good designs.

Oblivious to this latter realization, we brainstormed a few mechanics, starting from a group of free cells that would automatically fall back on your stacks when filled (like a bubble that you blow until it bursts), then we had other ideas that involved ways to be able to place cards on the back of each stacks and have them “bubble-up” while you tried sorting the cards into sets (disclaimer: the theme for the jam was “bubble”, we aren’t just naturally obsessed with bubbles). After a bit of playtesting the issue was glaring in our face: if we allow the player to do moves that neither help with solving the game, nor push the game’s state toward a stuck ending, a solitaire can easily become boring and mechanical as you plop cards around without much forethought, neither getting punished nor rewarded in the process.

Stuck or stack

So this is pretty much it. If you want to design a solitaire game that is fun (at least according to our tastes), the moves you allow for the player have to be meaningful steps towards order (the usual goal of solitaire games) one way or another. Any move that just keeps the chaos at its current level is a waste of time and usually not something the player can reason about, hence they will easily just keep doing those things until they realize they aren’t having fun.

In other words, it’s best to avoid moves that are reversible, at every placement you should be committing to a decision that furter collapses the possibilities ahead of you towards winning or losing. While in our game there is a bit that doesn’t obey this rule (if you have an empty space you can shuffle your bubble stack around), I still think we managed to step in the right direction and hopefully if we design more solitaires in the future they will be even better in this regard.

Undoable

Speaking of reversibility, in the digital realm the universal reverse move is the undo functionality for any turn-based game. Interestingly, the Zachtronics games for example do not offer any undo whatsoever (apart from the epic Fortune’s Foundation, which offers a single undo). In the real world, undoing moves is also limited to a handful at max as it’s difficult to remember them (using a brain like mine at least).

In terms of development, once you implement a single undo in a game it is easy to allow any amount, but I believe that undo in a solitaire actively hurts the experience because if you are able to go back freely, you are more likely to progress by pure trial and error instead of thinking ahead and calculating lines you could take. Hence, we made it so that you can only undo the last step.

Solvabubbles

That said, I also enjoy the experience of trying to figure out the solution of a given deal instead of leaving it hanging on a stuck state (a great solitaire that implements this is Pirate Solitaire, another Zachtronics-inspired gem), so we allow the player to restart the current deal from the beginning as an addition to the single undo. While there is currently no mathematical method or solver to estimate solvability for popiteru deals, I have yet to find a deal that I couldn’t solve in practice, the deals can greatly vary in difficulty though (there has been a few deals I had to spend an embarrasingly long time to sort out).

Additionally, since I don’t like the idea of trying to solve puzzles that might not even be possible, I have fantasized about only dealing solvable games either by reversing the rules of the game and generating moves backwards semi-randomly or implementing a brute-force solver that would reshuffle whatever deal it determines to be unsolvable, but this fantasy was cut because we didn’t have time left from the 48 hours we had for the jam. Luckily, the game still turned out to be very solvable which also means the rules are viable to be played with physical cards as well (just replace the tokens with 8 different ranks from a standard deck and match each rank’s 4 suit together).

While we bring the next update with the ability to share and play specific deals, if you think you have found an unsolvable deal I’d appreciate if you’d drop us a screenshot of the beginning (before any moves have been made).

Happy poppin!