Brainstorming on Ideas
Once you have decided what kind of game you want to make—which is usually easy since we all know what we like—it's time to think up the game. This is where you're totally on your own. There's no way to come up with good game ideas consistently.
Basically, you have to think of a game that you would like to make and develop it into something that sounds cool, is doable, and that other people will like as well.
Of course, you can get help by using other games as models or starting points. Don't copy another product exactly, but loosely following successful products is fine. Also, read a lot of science fiction books and game magazines, see what is selling, and watch a lot of movies for cool story ideas, game ideas, or just visual motivation.
What I usually do is sit with a friend (or by myself) and just throw out ideas until something sounds cool. Then I develop the idea until it sounds plausible or it falls apart. This can be very frustrating. You may overthink all your ideas and throw your hands up after two or three hours. Don't despair—this is a good thing. If a game idea survives the night and into the next day and you still like it, chances are that you might have something.
I want to say something that's very important, so listen carefully—don't bite off more than you can chew! I've received thousands of e-mails from newbie game programmers who want to create something at the level of DOOM or Quake for their first game. It's simply not going to happen. You'll be lucky if you can finish an Asteroids clone in three to six months, so don't get crazy. Set a reasonable goal. Try to think up something you can do by yourself, because in the end you'll be working by yourself—people flake out. Again, try to keep your first game ideas simple.