Sometimes you’ve got a game that’s a fair way through testing and further refinements aren’t bringing any significant improvement. At this point you may be approaching the best game possible within the limitations set by the major decisions that you’ve made. This is a good place to be, but that doesn’t mean that it’s okay to rest on your laurels – it’s time to slash and burn.
The most important part of this process is to save all of your work so far and squirrel it away somewhere safe that’s completely seperate to your development environment. It might be that the best game you can reach with small changes is the best game you’re going to make and you’ll want it back when all is said and done.
That taken care of it’s time to utterly gut what you’ve created so far. The goal is to explore some big changes and see if they’ve got the potential to eclipse the best you can manage within the decisions you’ve made so far. Of course you already made comparisons like this when the decisions happened first time around, but mountains look different from the top and it’s almost certain that what you have doesn’t entirely match what you thought you’d have when you started.
Break something utterly fundemental to the game and rewrite the components that depend on it to work without it. You might wind up deleting 50% (or more!) of what you’ve written, don’t worry about replacing all of it, just replace enough of it to generate the minimum needed for a game. Give it a go, it won’t outperform your more finely tuned effort, but it will benefit from some of the refinements and should give you a much more informed view of the decisions you made earlier in the process and their impacts.
It will feel like a huge step backwards, but it can help significantly. A game designer in this position is like a blind person trying to climb to the highest possible point. You can tell which nearby steps are uphill and downhill and you can walk uphill until you reach the top of a hill – but you can’t see how high that hill is before you get there and you can’t see if there are other nearby hills. The only option is to pick a few directions and run a way along them to see if you can find anything nearby that might be better. In artificial intelligence we talked about this in terms of avoiding local maxima – that is areas that are as good as it gets locally, but are insufficient compared to the overall solution space.
Artiste is presently undergoing this sort of kicking, each playtest is now only producing a marginal improvement so I’m trying out some major changes. It might well be that this doesn’t lead anywhere and I restore from my backups sometime down the line, but it’ll be a while before we have art so it doesn’t hurt to do some exploration.
Presently the game’s core structure is that you each use a power, shuffle the cards around, use another power and have a final discussion to work out what’s going on. The most exciting points of the game are the first round (where people are getting to grips with what’s in thsi game) and the final discussion (which is the payoff for the rest of the game). So what happens if we delete every other part of the game and rewrite the cards so that they don’t need it?
While we’re at it, the rules people get wrong most often is when to look at their cards and muddling. What if people get to look at their cards at the most natural time (When they’re dealt) and the state (look / don’t look) changes only once? What if muddling is deleted from the game as a core concept? Or even deleted entirely?
Alright, some of this might be going a bit far, but you get the idea. I’ve knocked together a quick version in which players get one action each, that they do in any order and their cards go facedown never to be looked at again once they’ve used their one action. I’m not sure how it’ll play, but even if it fails spectactularly I’m likely to learn some lessons about other directions to experiment in when looking at alternatives to the present version of the game.
They say ‘kill your babies’ and I think it’s good advice, but back them up first, just in case.