And that’s the primary reason why every playable alpha and early access game I’ve ever played has waited until the end of the development process to get their graphics looking good… assuming they have the budget to do so. You don’t need high end models when you’re still developing your game mechanics, game loops, and other primary content. Once the stuff is in place that will keep players playing, and talking about your game to others, then you’ll know how much resources you have left for the fluff, both in terms of funding and your players’ available computer resources.
It’s also why Star Citizen, despite having squandered over half a billion dollars and counting, may look like better than modern $50 million game graphically, yet is continually outperformed by tiny indi-games, both in mature game loops, and in game stability. The latter focus on the “difficult but cheap” gameplay, while Star Citizen continues to focus on the “easy but expensive” graphics, leaving more important the gameplay to languish in limbo. And then they inevitably redo the graphics repeatedly due to the changing demands of gameplay, new ideas that are incompatible with the current models, the fact that what’s considered “AAA graphics” changes with time and technology, or mostly due to Chris Roberts changing his mind about what he wants something to look like again.
