How do games actually get pitched and greenlit?

Through a staged process in which money is released against evidence, rather than a single meeting where someone says yes. Each stage exists to reduce the risk of the next, and most projects die at one of them.

The stages, roughly:

Concept. A short document — premise, genre, audience, comparable titles, the hook in a sentence. At this stage the pitch is about whether this is worth exploring, not whether it will be made.

Prototype. A rough playable answering the riskiest question: is the core loop fun? Nothing else matters yet, and prototypes are deliberately ugly.

Vertical slice. The critical artefact. A small portion of the game built to final quality — art, audio, mechanics, interface — demonstrating what the finished product will feel like. It is expensive, it is what most pitches actually turn on, and it is frequently misread: a polished slice does not prove the rest can be built at that quality within budget.

Production, once funded, followed by alpha, beta and certification.

What a publisher is actually assessing:

The team's track record, which frequently matters more than the idea. Shipping anything at all is a meaningful signal.

Whether the scope is credible against the budget and timeline, and experienced publishers discount stated timelines heavily.

The market position — who buys this, what it competes with, and whether the genre is growing.

The business model, and how revenue continues after launch.

The hook, summarised quickly. A pitch that requires five minutes of explanation before it becomes interesting usually fails.

The other routes to funding: self-funding, platform holders paying for exclusivity or subscription inclusion, publicly funded schemes and tax relief, crowdfunding, early access revenue funding later development, and private investment.

Why projects are cancelled after funding: the prototype does not prove fun, costs escalate, the market shifts, a competing release lands first, a platform deal collapses, or corporate priorities change — and cancellation is frequently unrelated to the work's quality.