Homegrown solo variants

Question

This afternoon, I played Ascension for the first time - and I played it solo, by simply dealing two decks and playing both players' turns. I had fun, and ultimately Ascension seems like a game where playing for more than one player isn't too much a conflict of interest - you can only do so much to influence the other player.

It led me to a broader idea, and I'm bringing it here as a question.

When a game doesn't have a predefined solitaire rule set, what's a good approach for designing a solo variant?

The three biggest categories I can think of are:

• design an AI player who follows rules, and execute them on the AI's turn (see Race for the Galaxy)
• design challenges which the solo player must overcome (see Merchant of Venus)
• play as more than one player (as I did for Ascension today)

What's a good set of criteria for a custom solo variant to a board or card game?

Answer 1

Any approach to designing solo variants should take into account this impressive body of work on BGG. This user has created 46 (to date) very robust solo versions of popular'highly ranked games.

Answer 2

Designing AI for the game is a very good way. These are also useful if you want to play a game that is best played by for example four persons, but you are only two or three. A good example of this is The Robots for Power Grid. Their randomized solution is very good for replayability and for creating several different AIs. Designing AI for a boardgame is also a very good way to get into the games core and generally a very good exercise in game design. (I use exercises in that at the University for game design students). Remember that it may be a challenge to get these AIs easy to use, and that ease of use something I would recommend to put at top of the priority in this case.

Designing new content, suited for the solo player, maybe with some other backdrop, as in High Frontier Interstellar is another solution that may be fun.

Some games may be suited for solo variants just by changing turn sequences or adjusting them to facilitate events that is needed for the game to progress, typically for games where the winner is determined based on victory points, and you end up trying to get solo high score.

Another way of doing it is to add a redesign that facilitate collaborative game play. This can then be played by one player, playing all players, as a positive bi product.

Finally your creativity sets the limit :)

Answer 3

Playings as more than one player, in my opinion, is the worst of the options you mention since it's prone to be affected by the fact that you know what the other player has and wants to do. It's a good way to learn a game, but no very entertaining.

Designing challenges, may be suited for just a bunch of games. Most often, the game is a linear sequence where you have to achieve something or a few things, may be balance those few things or focus on a big one, so setting challenges will take the dummy player out of the game since it's timing would off most of the time. Plus you could adapt to what the dummy player has to do in order to be ahead and take advantage.

Finally, designing an AI, Jon mentioned, is a great way to get to know the game, I wouldn't do it immediately, but after a few plays simulating a second player, may be even a third player. Yes three players could be too much, but remember, the idea is not to win the game, but to discover how to create an interesting and challenging opponent. I have done this a few times with more or less success and when you pass the boredom of forcing yourself to try things for each player even when you know it's not a good play, you end up with a good set of elements to "build" a player.

You may try first with games that you know well and if your fiancee usually plays in one way, you can start replicating that, and adding some alternatives here and there.