Scoping something you can actually finish

Immersive projects fail on scope with almost total reliability. The concept is a whole animated world; the deliverable that would have impressed anyone is one object, beautifully lit, that responds well to being looked at. Reviewers and employers read finish, not ambition, and an unfinished world reads worse than a small thing done properly.

Decide three things before you open an editor: the single idea, the performance budget, and what a visitor with reduced motion and a three-year-old Android sees. If any of those are undecided, you are not scoping, you are hoping. Write them at the top of the README and hold yourself to them.