Pick something where you wouldn’t be sure exactly how to vote on it with 50/50 odds. Ideally something which is either easily Googleable, or has a canonical data source - then include that in the market description!!!

Ideally no one has to ask any clarifying questions on your market as you’ve already provided everything necessary for them to play the game. And lastly, make sure to 2x check the end date is correct!

The wiki is better for this

Not intending to receive bounty just contributing. My first question had a clear outcome, yes/no, in the news, easy. But it seems like a very likely no, and so it's boring and unpopular. I've watched since, and see that writing a good question that isn't boring or unpopular and isn't immediately bought down to a swift NO in hour one, is very hard in itself. Before any considerations you're asking for here

Ambiguity: Try to think of realistic edge case scenarios that would make it resolve vaguely

Subjectivity: Maybe think about the kind of question you would want to bid on. If you look at someone else's market and it seems too personal/subjective to them, a lot of people would be less inclined to bid on it.

Controversy: Have clear resolution criteria. Maybe include that it will resolve N/A if something unexpected happens that makes it fall out of the normal criteria.