Actually, Graham., I believe there are specific behaviors that are small and able to be replicated* that separate those who don't get stuff done from those who do get stuff done. Maybe not everyone uses the exact same ones, but they certainly exist in a wide variety of endeavors. For example, professional basketball players often practice very small skills with great deliberation (like shoulder-tucking while shooting), whereas amateur players just play games without much self-awareness, so they never (or only haphazardly) improve their vital skills and behaviors.
Working every day is certainly a helpful skill, but it's not very specific -- what do you do every day? How do you start? Do you first review your to-do list, then disable your Internet connection? Or perhaps something else? (I'm asking these questions rhetorically -- you might do things in an entirely different fashion.)
*I'm not saying they're easy to do. However, spending time breaking them down into very small, concrete actions helps.
Usually most projects go like this for me.
1. Initial concept, the idea, basic gameplay.
2. Create prototype to test initial concept and gameplay design ideas.
3. Refactor and adjust design ideas, until mechanics are found fun and correct.
4. Depending on the size of the project, make more prototypes to test particular gameplay mechanics/ideas addiions, throw away prototype and start coding the game correctly, or just go forward with the often less than ideal prototype code base. (Works okay for smallish games.)
5. At this point your core game mechanics are ideally implemented, and locked down. Your game design direction, aesthetic, etc is clear. Now you inflate the design to account for content, variety, menu systems, and so forth.
6. Break the new enlarged design down into chunks which can be implemented, tested, and tweaked. Todo lists often help here I find.
7. Repeat until chunks are done. Chunks may get added, removed, or changed during this process, though ideally not to much.
8. Enter the final stages of content tweaking, polishing, and QA.
9. Repeat 8 about 5-10 more times, until you've just had enough and say the game is good enough. (You could polish a game forever, eventually you just have to decide its done.)
10. Final pass of QA/User testing, minor adjustments.
11. Ship.
Most all of this is just about working though. The main trick really is just to know where your end destination is more or less, and what you need to accomplish to reach it. This is where lots of pre-production/prototyping work comes into play. Really there is a whole field devoted to project management, and production as this can become quite elaborate when dealing with deadlines, budges, teams, and so forth.
Of course thats also just for developing, if your aiming for commercial success, you probably should have marketing, fan base/community building, PR, and so forth though the entire or a good chunk of development process as well, and after it.
A lot of times games become difficult to work on once you hit step 7 or so. A lot of it can change from creative work, to just a lot of labor. How you continue to motivate yourself to do the labor, and eventually finish, is just often very dependent upon the individual.
Developer
