There are thousands of "How to start" tutorials, questions and answers, but I miss "How to continue" types. I am looking for games good for learning more advanced features/mechanics/etc. A bit of explanation: I found myself advancing in my big project way too slowly and I attribute this to lack of experience in more difficult/bigger projects then side scrollers. I hope few intermediate games could actually improve the development speed in long run.

Please note that I am more interested in programming then designing itself. I am especially looking for 3D/3D applicable games.

What I imagine as a game good for learning?

• does not require many assets (I am not an artist after all)
• quick to make and finish
• requires implementation of intermediate/advanced features
• good examples available - exemplary games by others showing "how to properly make this"

tl;dr: I made side scroller, 2d multliplayer and a raytracer - where could I continue now?

EDIT: I am not looking for beginner games like pong or akranoid but rather more difficult ones.

Thanks, those are nice beginner tips, but as mentioned earlier. I am not looking for for beginner tips. I already made games like space imapact, multiplayer pexeso and a raytracer. I doubt I could learn much from games like Pong or Tetris now...Or does quantity matter?

Perhaps instead of full games you could write tech demos for things such as particle engines, ragdoll physics simulators, skeletal animation systems, etc. Then you can concentrate purely on the technical aspects without having to cover ground you already know.

Create a rouge like MMO in the same style as ZZT .

You have all the elements of a game, with out the graphics - and 2D "adventure" worlds are easy to create.

Edit: attempting to go "3D" before learning the basics will not end well.

Perhaps instead of full games you could write tech demos for things such as particle engines, ragdoll physics simulators, skeletal animation systems, etc. Then you can concentrate purely on the technical aspects without having to cover ground you already know.

That is great idea, but it is tied two big problems:

1.st there are no lists of these things(obviously), so after the things you said I would be still a bit lost

2. if it is not a part of a game I could easily make something that is not reusable thus wrong

Create a rouge like MMO in the same style as ZZT .

It is spelled rogue.

Taking a different angle, I would personally suggest that there aren't any "must write" games that every intermediate-to-advanced programmer should write. It's pretty standard to recommend such an approach to beginners because writing simpler, well-understood games they can more easily grasp the basic concepts that provide the foundation for all games, and because by tackling these games in order of increasing complexity they can avoid being overwhelmed by a larger scale project.

If you've already done this you should know all the foundational basics. You know how to write a game loop, how to handle input, how to draw multiple items to the screen, etc. It's time to move on to bigger things and get some experience with large scale projects. This is the point where you should be attempting some personal project that you want to work on, but which may possibly be of no interest what-so-ever to others. As LennyLen suggests, you can experiment with small portions that you find technically challenging by isolating them as small tech-demos rather than full games, but there's no easy short-cut or stepping stone to help you learn how to handle a large project in it's entirety -- you just need to actually work on a large project.

Not the suggestion that you're after, but I hope that helps! :)

The problem of my "big" project is how slow it advance - I already refactored about 80% code I wrote. So basicly I spent 80% time by writing worthless code I wont need in the end. But because I lack experience in 3D I cannot know in advance what I will need and what I will not. I know it is not related to the original question, but any chance how to improve this?