Scroll, and watch the pipeline: real player voices, your brief going in, a room of forty convening around it.
A room starts from what you already have: the brief, the store page, the trailer, the price you are considering, and the decisions you are trying to make. No panels to recruit, no surveys to field, no six-week wait.
Underneath every room sits what players already did and said, in their own communities, in their own language. Explore the layers. No source stands alone, and every one carries its limits openly.
Hundreds of thousands of real player voices across 29 markets, each one classified into the concern it raises. Low-substance noise is filtered before anything else happens; what remains is people explaining, in their own words, why a game worked or did not.
The explorer is a sample of the layers, not the full catalog. Coverage gaps and bias remain part of the model's limits, and those limits carry into every answer.
You do not describe your audience and hope. The room's composition is derived from the market's real concerns: how often a concern appears, times how often it kills a recommendation, decides who sits down. The weights are data. The faces and voices are the simulation, and they say so.
Enthusiasm leads the room because it leads the real conversation. A room of critics was a bug we fixed with data.
One market's composition, rounded to seats. It changes per market, per genre, as the data changes.
This is the wall most simulation products do not build. The voices give you the texture: what this market will argue about, in its own words, with receipts. The score comes from somewhere harder: how real players received comparable games in that market. A simulation cannot inflate it, because a simulation never touches it.
What they will say, what they will ask, what breaks trust. Words, with every line bound to a real voice.
The score, from real market outcomes, tested by hiding the answer and predicting it.