| Hi HN, Everyone talks about “provably fair” games, raffles, and loot boxes — but in practice, most implementations still require players to trust the server. Think about it: the server commits a seed, the player sees a hash, and after the game, the result is revealed. Technically, the house could precompute outcomes and cherry-pick them — even if it’s unlikely. How can you make randomness truly verifiable without relying on a slow blockchain? Is there a way to combine player input, server input, and public entropy to make it fully auditable? And crucially, can this be done without killing real-time performance in a game? I’m curious what people in production are doing — clever commit-reveal models, public beacon sources, hybrid systems?
I’d love to see practical approaches that actually remove trust entirely, not just add a “proof” that still requires faith. Let’s talk solutions. |