Here's some info I shared on a forum during the last event about coming up with ideas. I hope it's useful!

__

The creative process is one of the hardest parts of the entire compo/jam, because it dictates everything you're going to be doing for the next few days. I usually don't make ANYTHING until I have a full plan laid out of what I'm gonna make, but deciding what you're gonna make can be ultra-paralyzing if the theme is too vague, or focuses too narrowly on some mechanics.

So here is some things I do to handle the planning side of Ludum Dare:

__

Situation: Theme was just announced Don't program anything yet. Don't make any art or anything. First you need to brainstorm really hard for a while, until you feel comfortable with an idea. Nothing sucks more than tying yourself to an idea you've changed your mind on halfway through the jam. What usually happens in my experience is that you either don't finish or you don't end up liking what you've made.

What I usually do is open notepad/xed and vomit out anything I can think of. I try to focus on mechanics and gameplay first and foremost, but sometimes you can get good ideas for mechanics from interesting situations and funny shit. Generally try to think of ideas you can make in 12 hours or less. Then you can spend the rest of the time polishing things up or adding more levels/music/etc. And if something goes wrong or you need to change something, you have enough headroom to do some refactoring.

Then once you have an idea, you need to make a list of all the things you want in your game, and what you will need for it to function. If you get some new ideas later you can always add on to it, or change it. But don't get carried away with feature-creep because

Situation: Theme is good, exciting even! Sometimes the theme gives you an idea right away, or the theme is fun enough that you don't really stress a whole lot on what you want to do. Sometimes you don't need to spin anything, and can make a very fun and engaging game without any wild alterations or interpretations on the theme. This obviously doesn't happen all the time, but basically try not to overthink too much. Keep things simple unless you can't.

Situation: Theme is vague or uninteresting. These are the WOOOORST. But it happens. So what I usually do is start thinking Why does this theme suck? What would make it more fun? Can I reword it without changing it too much?

Feel free to mess around with the theme as much as you'd like, while trying to stay faithful to it. Something dumb like tWo BuTtOnS can be deconstructed or made a small part of your game; like instead of it being a controls restriction, it could mean "these two buttons are the most important of these eight buttons" or "helping grandma find her two buttons she lost" or "you are against another guy in the game to see who can spam your button the hardest".

Situation: Theme is too specific and narrow. These can be harder to deal with sometimes. Tips from more vague themes work here too, but the goal is the broaden the theme out a bit to find some new angles to work with. A theme like "You Are The Monster" could be broken down by changing the context of what 'Monster' means, or maybe change what it means in the context of your game.

Situation: You're REALLY struggling with the theme. Ignore the theme.

Seriously, you can do this! The theme - technically - is completely optional; Theme is tied to a single rating that defines how close your game fits in it. You don't have to do the theme, but it helps to have it to give you a direction. If that direction is directly into a brick wall, then you can ignore it completely and figure something else out.

__

I hope these ramblings help give you an idea on some problem solving techniques you can use. This is easily one of the hardest parts of the competition because of choice paralysis and mechanics design, so don't feel too bad if you're struggling; Everyone is!