Quest 2: Game Idea Validator · Create Games with AI

Quest 2: Game Idea Validator

I played a bit with the Game Idea Validator, and it really depends on how well you describe the three parameters: the idea, the target audience, and the hook. It provides very valuable feedback, including market research, competitive and similar game examples, and suggested next steps.

Here are brief summaries for two game ideas:

Stardust Drift - Reality Check

Verdict: Build the prototype, not the pitch.

Strong concept (fun ceiling ~7.5/10), proven by Osmos and Gravity Ghost — pure-skill gravity games work. But it's not a two-week money-maker; it's a "make something small and genuinely good, then grow it" project, best as a $5–7 premium Steam/itch title, not mobile F2P. Cut full n-body physics and moving planets (they cause chaos and a shifting trajectory line) and cut procedural generation. Build one thing first: a 3-day test of whether steering-by-gravity actually feels good in your hands. If day three grabs you, commit the real six-to-eight weeks. If it's boring, you've learned the most important thing cheaply.

-----

Overpressure (working title) — Reality Check

Verdict: Build the prototype, not the game. Strong concept (fun ceiling ~8/10, but execution-gated), proven on both ends — "build carefully, then survive the consequences" is the reliable tower-defense engine, and the realistic-plant audience clearly exists and pays: Nucleares has over 1,000 reviews at 84% positive, with Chernobyl: Simulator, Rasvyat and a dozen others behind it. But your real edge is that prep→pressure structure, not the "real physics" — every one of those games is a first-person sim or an idle builder, so nobody's doing your tight, top-down crisis-puzzle loop. It's not a two-week money-maker, though: web is a wishlist funnel, not a business model, and the money lives in a $10–20 premium Steam release for the Zachtronics/Factorio crowd (the full game is realistically 2–6 months).

Chase legibility over accuracy — "real-flavored," not real-accurate — because this genre dies on opaque sims that make players feel cheated rather than challenged. Make the pressure phase pausable so it's a thinking test, not a twitch one; trim the number of systems and scenarios; and timebox the physics ruthlessly so it doesn't eat all fourteen days. Build one thing first: a 1–2 week slice that answers whether prepping a plant and then defending it against a pressure surge actually feels fun — one scenario, a simplified-but-faithful sim, the full loop, one survivable crisis. If the jolt is there on the last day, commit the real months and aim it at Steam. If it's flat, you've learned the most important thing in two weeks instead of six.

0 comments

Quest 2: Game Idea Validator