A) taking on projects larger than you should

B) not coding to milestones

C) not providing yourself with feedback along the way. That's why I always right my visual stuff first, so I can constantly see how new code is affecting my game.

I think it happens because people just get bored of working at the same project without visible results. You can spend forever coding say the classes used in the game and never be able to put anything on the screen. Its kind of discouraging to people that that much time yields no visible results. People just end up losing interest because developing a game alone can be quite a long hard task.

What exactly are we talking about? A lack of interest in programming (in which case you shouldn't program), or just that you lose interest in the projects you're programming before they're finished? (In which case you should do as said above, set milestones, only take on reasonable projects, make sure you can see some progress)

I'm kinda same as you Mad_Koder
When I begin a new project, (and god knows that did happen quite a few times so far ;) ) I feel full of enthusiasm and motivation.

What I feel prevented me from completing my projects (I havent completed one so far... :/) is I always wanted to start programming things that I like (mainly graphics, and I guess many game programmers like programming game because they like programming graphics stuff) So I always used to start with "ok, so my game will by several layers, lets begin with the main charachter layer...". I think you know what I mean.

As suggested ItsDoh, I've understood that being unaware of what you're unable to do with your current skills and knowledge is a main cause for failure. Usually I start coding what I can code and quickly end up realizing that I'll never have what it takes to "complete" my game (And when I say complete, I mean what I *thought* will be needed to complite, which might in fact be 1% of what will actually be needed to complete)

On the other hand, the countrary happens sometime, mainly because one does not have a good understanding of what a game is. For example one might think of a great adventure game, but what's in his mind is in fact just a landscape and a character, Thats not a game, though thats what one might head for. And once that's done, well, there nowhere to go next, and the project is given up. I think (but thats really IMHO) that many (didnt say all) screenshots posted in the IOTD end up this way, just being a game embryo that will never become a game (I'm not saying the demo isnt great !!!)

Ah... there would be plenty of things to say about that, but I'm done for now, cuz I started making my MMORPG two weeks ago and, well, you know... ;)

Set more goals, do at least 10 minutes per day, and make sure it's something workable and interesting before you start -- this is where a lengthy design process comes in handy.

It is very easy to burn out when you are starting out -- just selectively restrict yourself to the fields you are capable in and work on that. Before you know it, your area of knowledge will expand.

I've programmed since 1979. I've pumped out seemingly endless lines of code on marathon 16-18 hour days for 3-4 weeks at a time. I must say that even as a professional architect and developer, you go through funks.

In my experience, this is often caused by not being able to see much progress in the hours of work you put in. Do as others have suggested; take your big project and break it down into smaller pieces. My advice is to look into these areas; iterative development, agile methods and test driven development.

Once you see progress; then you'll be back into the swing of things.