More accurately, how do I design them?

Sit down a play a game you like, look at it, ask your self this: How is this level I played designed, what makes it fun. Do this again and again with many games. Also do this for GUI design as well. Do this for features and functions as well etc. etc.

I don't have any assets, and without them I won't go too far either.

Well don't worry about that. Her you go: http://opengameart.org/ read the different licenses well though :)

Oh and no matter what people tell you then you are awesome, because you already made a game.

Game design is not coding. What you have done in this past 3 years is learning how to code, not how to design games. You have not been doing things wrong. Now that you have learned how to code, maybe now you should start reading about game design -- what makes a game a good game.

There are many approaches to the actual game development phase, some which can simply be made up, which I'll explain. You could probably find a step by step process on the internet somewhere, but a sense of discovery is more powerful than a sense of being right.

I prefer making things up as I go along; there are endless new ways of attacking a problem by simply making up the answer, as long as it works. This is a steady and fun track because it leads to discovering things I never forget.

Technically I keep a goal, a to do list that won't be complete until my game reaches a complexity that overwhelms me, and a task at hand I can focus on; this is as much of a process as I need to stay organized. I keep as many backups of actual work as necessary to feel security and keep a history of accomplishment in case I fell nostalgic.

The most common recommendation found here on gamedev, random articles, and pumping the search engine with random phrases about game programmer[ing]s is someone spent their time cloning old games, they think you should do the same because it is proven to work.

As always you're stuck making up your own mind unless you can find sombody who tells you what to do. If you like that sort of thing.

can focus on; this is as much of a process as I need to stay organized. I keep as many backups of actual work as necessary to feel security and keep a history of accomplishment in case I fell nostalgic.

This "history of accomplishment" is a very good point, i usually

throw out some of my old zips - and indeed this is some loss

(the rest of this answer i mostly did not understood, what for example is "making up answer" ?

@fir

Sorry, it'd take too long to explain everything. The concept I have of making up an answer would be when the answer that I arrive at is either self evident, derived from numerous references at one time, something instinct, or completely fictional - until it is proven valid. The opposite of making up an answer would be asking for help or basing my findings on someone else's work or something in nature without contributing to it (e.g. if opium manufacturers pulled weeds out of the ground and sold weeds, or if bottled water companies sold you water without testing its purity, they aren't contributing to the solution, it already was there).

The most common recommendation found here on gamedev, random articles, and pumping the search engine with random phrases about game programmer[ing]s is someone spent their time cloning old games, they think you should do the same because it is proven to work.

Thanks. This is EXACTLY what I'm going to do. I'm starting to make a very precise clone of Super Mario Bros. I already made first map, collision system works nice. Now i'm going to add destructible bricks, enemies. I hope everything goes alright. As of now, I'm having much more fun than I had thinking of my own game. And I really see myself doing more stuff like this. I maybe even turn this project into retro games mash-up, by combining many retro games into one title. I'm really looking forward to this, and I hope everything will go smoothly.

Everyone else, thank you for you advices, some of them are really quite good.

@fir

Sorry, it'd take too long to explain everything. The concept I have of making up an answer would be when the answer that I arrive at is either self evident, derived from numerous references at one time, something instinct, or completely fictional - until it is proven valid. The opposite of making up an answer would be asking for help or basing my findings on someone else's work or something in nature without contributing to it (e.g. if opium manufacturers pulled weeds out of the ground and sold weeds, or if bottled water companies sold you water without testing its purity, they aren't contributing to the solution, it already was there).

Still I am not sure If i do understand (maybe its because weak english) - you recommend to try to find an answer alone, instead

of searching about it in google or books?

Recently I have the opposite attitude I am searching the net

for avaliable answers and solutions not trying to make my own.

@fir

As long as you are scientific about it, then I'd approve.

I got exactly the same issues, and additionally I would add motivation slump on the top. I know how you feel.

I even pushed a bit further this crazyness: I made a very complex, really beautiful, and utterly boring game:

http://www.creatures-of-gaia.com

It was for a contest, so I tried to polish everything: I hired an artist, got some decent music, it was multiplayer and all that fluff ...it was everything except fun. And this fun factor really isn't easy. We all think our "ubber game idea" will be great and loads of fun, but when you actually play it, it can easely feel dull.

What do I want to say with this?

Being able to code a game, and being able to make it fun, are two completely different matters.

After that failure (which costed me money and I didn't earn a single dime), I needed a change of mind since I was down. I decided to make a damn simple game: a space shooter. Nothing easier right? Well, technically not, but again "content wise" it's a bit more tricky. I had to decide what were the enemy ships properties, where and when they appear... And somehow, all this arbitrary stuff collided with my analytical mind. Should I make the ships move faster? Or slower? Or more/fewer of them? And what should be the next wave? ...as silly as it sounds, I wasn't prepared to find difficulty in this. I'm also a bit perfectionist so I think it makes things tougher. In the end, I made a random generator making random looking ships with random properties appearing in random waves. It's nice and all, playable, but again quickly dull and repetitive. Here it is:

http://sss-demo.site11.com/

In other words, I have no recipy on how to make a game interesting, entertaining and fun. I even think this "fun factor" has no recipy, it's just harder than people expect. It's putting everything in the right mix and with the right twist. What I try to do now is to experiment along with "silly stuff" and if something "appears to be fun", which then, if it is, could be polished. Kind of a trial error process with tweaking on the fly and asking people what they think of it.

Good luck with it!

Being able to code a game, and being able to make it fun, are two completely different matters.

Exactly. And I come to realization the hard way.

Thanks for everything.