In the last game jam I worked on, the ideation phase for me looked a lot like this
1. think about some games that I like and try to "borrow" concepts from them that I think work well in a game jam.
2. setup some somewhat arbitrary constraints for the game (maybe the game has to be first-person or no real-time physics or whatever)
3. do some chaotic unstructured text-dump brainstorming in a text file. I prefer simple plain text editors like Sublime Text or Vim because I'm easy to distract with more featureful word processors
4. when ideas from #3 sound right, really roughly sketch out what I think that might look like on paper. This is less for art planning and more just so I can visually diagram things and make an idea more concrete.
5. start prototyping the simplest version of what I was sketching up
I do #1 and #2 because I think I work better when I don't have the entire design space of all possible games to think about
I do #3 and #4 more as just something to get me started that doesn't require any complicated tools or coding or anything. I've heard some people like to make actual paper prototypes but that feels like too much work for me. At this point, all I'm really trying to do is get myself excited enough about an idea to start prototyping
#5 is where the real magic happens so you should be trying to get here as fast as possible. Don't spend too much time on the previous steps. Just use them to build momentum to get to the prototyping phase. 100% agree with what "No Time To Play" said above