I'm in the process of addressing what's quite a pleasant problem to have with Shennaigans: The Musical. The game is quite quick to play and as a consequence some players who are enjoying it are playing a lot of games. Obviously that's great, but it's generating an emergent phenomenon: They want some sort of way to play multiple games as a cohesive whole with an overall winner.
The most common way to solve this issue is simply to assign a player a point for winning a round and to declare the game over when a player reaches a particular total. Looking back at other games with very short rounds that feel like self contained games I start to wonder if they've come about in this fashion. For instance I wouldn't be suprised to hear that Love Letter was initially developed with a round being the entire game and the multiple game structure was added later in development. It works well enough as a self contained game for that to be true, I can envision the dev team having a similar experience of finding players enjoying playing lots of rounds and looking for a way to build around that.
Adding multiple round scoring does create some slight complexity in terms of how we should track score. The game is presently simple to manufacture so it's desriable to avoid needing tokens or any other component to do it. On the other hand it's nice for games to be self contained units so it's not great to rely on people having a pen and paper to hand. Finally organic memory is terrible and everyone hates it.
Putting all of that together our tracking mechanism needs to come from our existing component: The cards themselves. There are 54 cards in the game but at most 10 are used at a time, so there are always 44ish that aren't in use - plenty to keep track of some light scoring. Problem solved!
The best design decisions solve several problems at one stroke, so rather than being content with that it's important to ask if there's an opportunity to do better. In this case I think there is: Players are often upset that they can play a lot of games without getting to see how a particular card (quite often their instrument, but sometimes just a mechanic they think is cool) works in practice. If we're giving players cards to track their score, is there any way that we can also use their effects?
Of course the answer is yes. The simplest implementation is to let each player substitute one of their score cards for their active card at some point during the game. The problem is that this cannot be allowed to remove the artiste or manager from play. This in turn presents a game balance issue as finding the artiste is at the core of the gameplay and if any person can prove that they're not the artiste by making a subsitution that'll fall apart.
The solution that we playtested (or I should say the most successful solution that we playtested - we tried other things that didn't go so well) was to oblige all players to *maybe* switch their card at the same time. This was done without it being clear whether they actually switched or not. A player with the manager or artiste may not make a switch, but they still shuffle their cards together and pick one so that it's not apparent to the other players that they haven't change anything. This approach worked pretty well and generated some nice feedback. In particular people appricated victory having a meaningful effect in increasing the number of card changing options they had for future rounds.
That works pretty well and I wouldn't feel bad publishing the game with that solution, but I still wonder if it'd be possible to go a step further. Looking at player engagement over time the muddling round (in which players swap cards with each other) appears to be the least enjoyable part of the game - however it's vital and attempts to remove it have typically made the game as a whole worse.
Adding the extra option to the muddling round might be a more productive way to make the score cards meaningful and to capture the benefits of letting players have a significant say in which cards see the table most often. Since it's a muddle (two cards are swapped under the table, the player has the option not to swap them) it could be performed with the player's victory pile without making it obvious that a manager or artiste card prohibits them from making an actual swap.
On the other hand this may introduce some sort of awful optimal strategy. Perhaps it's best for everyone except the artiste to swap to "catch the artiste" roles and claim a trivial win by perfect coordinated information. It bears some thinking about and at the end of the day finding the best solution to the multi-game scoring problem - that is to say a solution which solves that problem elegantly while integrating with existing mechanics and solving other issues if possible - will require the same thing it does every night.
Playtesting and lots of it. I'm confident that we've got a good method and can go with that if needs be, but I'll keep trying for a great one for as long as I see viable options.
That's it for game design this week, but I've got a small announcement to make while I'm here: This project is going to be put on hold for a couple of months. I will still be working on it and sneaking in the odd playtest whenever I happen to have a group together. The game is very nearly complete and I'm expecting to return to it from Dec 4th next year with an eye to launching Shenanigans in early 2016. The TPO blog posts will become much less frequent until then, but if you can't get enough of my writing you can still follow the articles I put out for other outlets at my ... er ... at my right here. Since you're reading this on the BGG aggregate page! Hiya
I look forward to joining you agian next year for more hijinks and musical fun