I want to make a puzzle game, but I don't know how to start. I have observed that each puzzle game has a core mechanic that enables opportunity for numerous puzzles. I'm not sure how to think of ideas. I was wondering how to come up with ideas for my game. I do not want any of your ideas because I want the game to be original.
Those released games you play, they are finished games, obviously, and have gone through countless iterations. The gameplay mechanics you're enjoying might not even have been a part of the original idea. They can be happy accidents, or at the very least, iterated to "perfection". Most likely a combination. Judging ones own work against finished products is never easy, you should have seen those games when they were in their first stages: might not have been that impressive.
What I'm saying, great "ideas" take work, they don't magically appear from some void. You start with something simple, and you build from there. Think of a few core gameplay mechanics, experiment, and see what you come up with.
I don't know what kind of game you want to do, but if I wanted to do a game with say, physics and puzzles, I'd start simple. I'd make a player I could move. I'd facilitate a way he could manipulate objects. Maybe he could pick them up, push them, throw them, run into. Or perhaps some sort of gravity gun like in Half-Life. Maybe something completely different. Then I'd throw some obstacles into the mix, and some other moving parts, that Braid trick, where you rewind time, that was great, huh?
If you can't come up with some ideas, take some ideas from other places, experiment, be inspired, and maybe out of that something "magic" and original happens!
Starting development on a new game isn’t writing the story, making assets, or even the game engine. You want to make a game, you need gameplay mechanics, that’s where you should start. If you can get your mechanics to work, if they are fun, even if your graphics are all stick figures and blank textures, then you know this game is worth more effort. Now you’re ready to make it look nice. But how to get here? The answer is of course prototyping mechanics, and it’s the most important phase of development.
A prototype should be more than just drawings on paper — although paper prototyping might be a stage in your game development — I’d strongly advice you should have made several playable prototypes before starting on anything more concrete. It should be very simple and small of course, but a playable game nonetheless: rudimentary controls and mechanics so you can get a fairly realistic impression of what the finished gameplay might be like. You need to have enough in your prototype for it to be of value.
The whole point of a prototype is to find out if an idea works in practice. Since we’re talking about video games, we might say that the point is to find out if the idea is fun in practice. But the prototype should also give you insights into game design, I’m not talking about visuals or story here, but game design in terms of game rules and mechanics. In fact, prototyping is game design. You want to test out as many ideas and mechanics as possible, to find several that works great, but each prototype should have a narrow focus. Be very specific about what you are actually testing.
When making a prototype, don’t worry much about anything, because you’re not going to keep anything except for the core mechanics, which will all be redesigned and iterated. You don’t need to think about how to structure your code, your architecture, you use whatever language or engine or whatever tools that gets the job done in the shortest amount of time possible.
In addition to the advice recommended by Eric and Jeff, you may be able to benefit from paper prototyping. Once you have come up with several ideas for game mechanics, you could try using paper prototyping to flesh them out quickly. This could also be useful for identifying bad game mechanic ideas without wasting any time actually programming these game mechanics.
I've noticed several partipants of the Ludum Dare use paper prototyping in their efforts to create a game in 48 hours. They've said good things about it, and it seemed to be pretty beneficial for those who made puzzle games.
When I get stuck thinking of an original idea, I try to just play a bunch of games similar to what I'm going for and usually an idea stems from that. I would just play a bunch of puzzle games and see what you really like and what you don't. Be careful not to base your idea too much off of one game though. And if you're still having trouble, try to write code to mimic some other game to see how they could have went about enabling their core mechanic. That way you kind of learn what may have went through their head when writing the game. Or you will think of a new idea as you are coding and maybe decide to do that instead. Either way you'll learn something. Just start coding and the ideas will come. Good luck!
You can try combining two different gameplay mechanics into one to create a brand new game (e.g. Gunpoint combines stealth and puzzle gameplay mechanics).
If you can not think of any good ideas or what gameplay mechanics to combine, you could use this random mechanic mixer which will combine different gameplay mechanics. At the bottom of the page is a description of all the different gameplay mechanics.
Lastly, just like everyone else said, play games to get ideas, and prototype your ideas to see if they are fun.
It's very hard to come up with a revolutionary new game mechanic. Next to impossible.
What is interesting is the ability to piece different existing pieces together in some kind of remix.
Refer to Jasper Jules' Family Tree (http://www.jesperjuul.net/text/swapadjacent/index_image006.jpg) of match 3 games. Each one challenges an underlying rule/premise for the game. What if we held time constant? What if the Match 3 was circular?
What makes Match 3 really interesting and popular (Candy Crush is currently the #1 game on Facebook) is that the basic mechanics are very familiar, but it's an iterative twist.