Almost no one writes their own FPS entirely from scratch in the past decade outside of a handful truly enormous industry players, much as no one really writes the entirety of a web app themselves.

You choose one of the many existing game development frameworks and rapidly prototype a concept, if it works you keep iterating on it etc etc. This works in games as well as any other major software endeavour.

The matter of having the talent to produce a new idea that’s actually fun to play and marketing it successfully is arguably the hard bit here, rather than the technical implementation details. I’d make the argument today that the technical barriers to entry for game development are almost _too_ low now, which is why partly why Steam is filled with so much “indie” junk.

Asset stores allow you to have placeholder mechanics like you used to have placeholder art. So placeholder camera, controller, collision detection, and a placeholder level. Then you get to rig what you want to test on top of that, or just sell it as is which is called an "asset flip".

You could do that or a frramework like Unity takes care of lot of stuff for you.. weekend maybe a bit extreme, but you can get something running and functional in very short time

It depends on the twist.

You can download Unity, purchase and install for example UFPS that will handle most FPS functionality for you (includes basic gun etc model) and you can use Unity networking for LAN connectivity.

Let's say the twist is a Battle Royale game, you just make a Unity terrain and a basic sphere that becomes smaller.

After that coding the extra bits such as HP, HP loss due to sphere etc isn't a big deal.

That's one way, doing it via a mod also works, but you really need knowledge of the product you're modding (although the same is true of e.g. Unity, if its your first week with it it won't be that easy).