I have two small but significant pieces of advice. First, know exactly what you want to do. If you don't have a relatively clear focus of what you want, you'll have a difficult time seeing it through. That said, you don't have to know all of the ins and outs and have everything set in stone. Your project is going to naturally warp and shape and evolve; just know the general concepts and what you want it to be.

My second piece of advice, is to never stop. I'm not sure if you're looking at doing the coding, the art, both, etc, but don't let yourself lose momentum. Life is going to get in the way, and it will test you and even break you at times. Those who are willing to stand with determination deserve the rewards for their own completion.

One last segment: I said to do what is passionate to you. I've worked on different projects, some were my own and others belonged to other people. The ones I lacked passion for got kicked aside. One of my favorite things when I played Warcraft, was at the beginning of the game the trolls had a quest titled "Your place in the World." I wanted to carve a name for myself in that virtual world. I wanted to be known. I channel that energy every time I start working on my own project. I'm carving my name in the world. I wish you the best of luck in carving a name for yourself!

My advice is start, it doesn't matter where just start; it's the most difficult thing to do.

My second piece of advice is to ask for criticism, you will often do things in a way that could be improved on.

I got a little piece of advice, basically when you have set a task for yourself try and cut it into individual smaller tasks it helps keep track of how much you've done. I know this doesnt help that much but still good luck :D

I have often found that consciously keeping focus on the project helps. When things are getting old and you feel that you somehow have lost interest... it eventually happens. I try to look back and see what made me start in the first place. This game mechanic or this kind of technique is what made me want to start this project. Almost always I have found it worth while to keep at it. Cut back on the feature list and scale it down until you are back at your original idea. I guess a meatier way of saying "start small"... ;)