Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

What an awesome project. I would love to mess with random stuff like whirlpool strength and see what it does to the game. Efforts like this to make the decompiler output intelligible e.g. meaningful variable names make this much more approachable for a technical person like me without much of the niche platform-specific reverse engineering skills. In fact there are countless games I'd love to dive into like this.


Train a ML system on a range of parameters (whirlpool strength) until you have a decent port of the game to a neural network and/or tree-based algo. Then try to optimize the game based on people’s enjoyment.


I think it would be much easier to use change the parameters directly in the original program. Just have to use a gradient-free optimizer.


Yours is a better technical approach to the vague dream I had described.

Am I being downvoted because I don't have an optimal solution to some problem?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: