This week I've been focusing on a problem that I've come to realise is a specific instance of a more general problem. What's more, it's a problem that a lot of games don't have great solutions too, so it's potentially a really interesting area to explore.
My specific problem is that I have a SatW prototype in which the players are at a masquerade ball, trying to invite people to go back to their place once the ball is over. The central challenge is identifying who will synergise well with their existing guests, while preventing their opponents from doing the same. In service to this they can do things like swapping character's masks over or even replacing them with various sorts of booby trap character and trying to trick their opponents into inviting them.
The problem is that it is hard to set a level of information for players to receive that provides an interesting play experience. The reason for this is that the ideal amount of information to give out depends upon how much will be changed before they can utilise it, they need enough that they can overcome some level of obfuscation, but not so much as to make it pointless. This in turn depends on the player count, which might vary from game to game.
Initially I thought of this as an information problem, to be solved by better understanding the flow of information through the game, but there's another way to look at it in terms of a general problem.
Interactivity is desirable, in most games you want your actions to directly or indirectly influence other players.
A variable player count is desirable, gaming groups aren't consistent and it's nice to be able to support a slightly variable group.
These desires do not play well together. Fundamentally if a player can be affected by another player, the amount that they are affected between turns depends on the player count. This leaves the extent of external effects outside of the designer's control, which in turn means that games aren't the tuned experiences that we as designers would hope to provide.
I'm sure you've all experienced games that claim to cover a range of players, but simply do not play well with certain counts. I think that this problem is often responsible.
Of course there's no need to reinvent the wheel and this issue is as old as competitive games, so it's good to look at other solutions (whether they were successful or not) and see how they apply in the context of my own work. Often, but not always, a solution from one area can be adapted for another.
There are plenty of ways to avoid the problem altogether: Make a game for fixed player counts, use simultaneous play, design a coop game using a shared resources dodge and so on. For now let's focus on ways to tackle the problem rather than avoiding it (though that's an approach worth leaving on the shelf for now).
The most direct solution is to reduce the impact of each interaction. This could be brute forced by having effects like "Your opponent loses points equal to the number of players divided by four", but there are more integrated ways to achieve it. Take for instance something like changing the number of available spaces in Viticulture. This has the indirect result of changing "Placing a piece also blocks 1/10th of actions your opponent might take" to blocking a smaller proportion if more blocks are going to be made. However the result is implicit such that the players don't have to perform a calculation based on player count each move.
Another solution is to provide opportunities within other players turns. A game like Cosmic Encounter could be very slow if each turn you had an opportunity to claim one base, while your opponent's had an opportunity to remove one of your bases. There would be many more opportunities for removal than placement. Allowing players to ally with players in other engagements off turn and progress towards their goals in this manner mitigates the extent to which a large player count can undo more of their progress before they can act again.
It's also possible to reduce the quantity of interactions that impact on a player off turn by making them targeted or reducing their desirability. Consider something like Britannia, in a three player making a move to deny another player points is more impactful, as you've made a large gain against one player and given the third party a small advantage over you (compared to making a move purely for your own benefit rather than a denial move). In a five player game it's less worthwhile because there are more third parties to profit from your conflict.
Bringing these general solutions back to my specific problem opens up new ways for me to run my masquerade ball. Reducing the impact of each interaction could be achieved by having more characters in play based on the number of players so that switching two of them hides proportionally less information. Off turn opportunities would involve generating ways for players to peek at cards during their opponents turn, ideally as a consequence of their own actions (a cost they pay for more effective actions?) to further increase interactivity while reducing the problem. Making effects more targeted would involve finding ways to make players more aware of what their opponents are trying to do so that swaps are balanced around misleading a specific player for a specific reason rather than to create information entropy. Again such a change can only help the game.
In terms of game design philosophy I hope this illustrates the value of considering your games problems are an example of broader patterns and looking to existing solutions. It's possible to get too close to something and consider only solutions from the domain that you're in, which can lead to missing out on some great ways to improve your games.