Chaos, and breaking the game

[vegeta897]

continuing discussion of a tangential comment

I think most people who hear the pitch for this project understand and expect the inherent chaos involved.

This project can turn into a more sanitized (but still very interesting) collaboration where users contribute levels and "modules" used by the levels that are isolated from each other to avoid breaking other content.

I would like to see if we can make the experience work without doing that. Can breaking the game still be fun and interesting?

Example: A new contribution makes a change to how the jump behaves, and this makes level XYZ impossible to beat

• Could the player still have a good experience if they understood why they can't beat the level? Not even the specifics, but given the context of the project, their imagination could do the work here.
• Could we give the player certain abilities (cheats?) that allow them to overcome breaking changes, in an emergent gameplay fashion?
• How often would it be fun for players to "solve" a broken level this way? How do we avoid (or do we even need to avoid) players "solving" levels that aren't even broken?

@iznaut followed up on my comment and pointed out that the game should allow levels that drastically change the game, even to the point where it isn't even a platformer. I was thinking too small! Such a level should be able to exist without destroying all the platformer levels. The amount that the game breaks itself needs to be such that the player can understand and deal with it.

For less drastic changes, I proposed one way to approach designing them that is potentially more interesting than just maintaining your own copy of the game's rules:

If I was creating a level where I wanted changed physics, I'd make a new item to give the player, and/or change the properties of the world itself, to achieve the same effect. This solves problems like config versioning/compatibility and lets others borrow the items I create for their own levels, and keeps just a little bit of chaos because the items can extend defaults rather than overwrite them. e.g. "These shoes make you run really fast" but "really fast" means 4x whatever the default is, not some predefined value. This also allows the player to understand why the physics are changed, instead of hiding it in a config file they can't see.

I'd really like to know how others feel about embracing chaos as a design.

[lazerwalker]

I'm not sure these are competing design goals! I think "a fundamental mechanical change breaks how a level works" is potentially a fun and desirable bit of chaos, whereas "our level structure within the project leads to git merge conflicts that everyone has to spend their time fixing" isn't interesting or valuable chaos. Consciously minimizing process/production chaos allows us to spend our time maximizing design chaos.

[vegeta897]

Oh, absolutely. It's important to make contribution as frictionless as possible.