Some time ago, the Chatty DM had a great post on the various projects all Geeks and Nerds work on and then abandon half-finished. With both game and software design this is an issue that has often cropped up in my life, so I thought I would have a post dedicated to the "design" portion of game design.
Here are some suggestions for conquering Projectitis:
1.) Set yourself a daily requirement for work, but keep it small and set a specific (and minor) penalty for breaking it.
In my case I demand one line of text per day. Failure to write a line of text per day involves me putting a dollar in a jar. The jar goes to the Salvation Army come Christmas time. This helps ensure I get SOMETHING done. A single line may seem useless, but over a 6 month busy spell it adds up a small chapter completed. This ensures you get over those bad cases of writers block that usually kill a projects momentum.
2.) Accept &*%$#y work.
I love to have elegant solutions that are just perfect. In Piecemeal and Adventuring Party! I have many such mechanics. I also have more than a few solutions that are still in their "awkward teen years". Many of the good and elegant solutions I do have started being awkward and cumbersome, taking many many years of re-writes and testing. Sometimes they even got worse before they got better. But its often easier to stare at a bad solution and fix whats wrong than it is to look at a blank page and create something that is pure win from nothingness.
3.) Don't be afraid of publishing things that aren't perfect
Welcome to the electronic age, there is no real cost to publishing before you are "done". Voltaire once said the enemy of the good is the perfect. This is because if you wait until its "just right" you'll lose inspiration before then. The praise, evaluations and scorns of others can be a powerful motivator to keep working.
4.) Anything worth doing is worth doing better
Following on point 3, you should never consider anything you are working on as "done". It can always be improved even if you can't yet fathom how. Perhaps it currently is perfect, but as you change other "less perfect" portions of the system, new avenues for improvement become available.
5.) don't be afraid of an overhaul, but sit and think on it for awhile
Every now and then you will need to fundamentally overhaul your project. Every now and then you will think you need to fundamentally overhaul your project, begin the process and realised you created even more and unsolvable problems in this transition. Always put in one more "phase" or release with your current version before a major overhaul, that will give you time to think this through further.
6.) Above all, always remember that you can easily save days of planning with a few short months of extra work.