General guideline for developing a game?

I work full time as a music composer and at a younger age had a real fascination with game design, however ended up choosing i different path.

Recently in my free time (not much) I started learning UE4, C++, Maya, as this has always been a dream. I would love to know if anyone could give me a general outline

of what order to go with in creating a game. Im sure the story is #1, but now what? Is it asset creation? Level design?



A lot of devs suggest, start by just making simple games. Games that have all been done before. If that makes sense to you, do it… If it doesn’t, try and at least make your own style of simple game. As its important to follow the form (for learning), but not follow the herd etc. Or at least not for too long, as that won’t help your progress either…

Regarding story… In games, gameplay is usually #1. Story is important, but its also a liability when you haven’t made a game before. For example a GDD is something that’s often encouraged. But it has questionable value when you haven’t assembled a game yet. Why? Because its hard to know what works, or what’s realistic, or what is likely to work as actual gameplay. Plus, the real killer, how much time will it all take?

Overall, if you can stay away for now at least, from designing Characters (or other tricky 3D like Vehicles), AI, Multiplayer and C++ AKA ‘The 4 horseman of the Indie apocalypse’ imho… :stuck_out_tongue: … Then, I believe, you’ll progress faster. In other words, get the basics down. Get the environment or get the core gameplay nailed first . Then, over time keep iterating on that and working on more advanced features. Otherwise you may find yourself going down rabbit holes, you never crawl out of. If that happens you’ll develop new skills, but you may never actually assemble a game…

Thanks for the advice! Yeah I totally understand making a very detailed large scaled game is unrealistic for one person with the skill and knowledge I have. My goal is to create the first level, however long that takes, and maybe take it from there if it actually turns out alright.