Originally Posted by Squize
I don't really plan, unless I'm pitching for something and then I do fairly comprehensive Game Design Docs ( Although those are more client targeted rather than being something you can really sit down and work from ).
I find when I know what game I'm writing I know which components make it up, and any which are tricky I just ignore. It usually works out by the time I start on them my brain has worked it out for me.
One thing I always do ( Which V reminded me of ), is a to do list. I only do this near the end ( When it's worth while ) and just put all the bugs and nice to haves in there. It normally ends up being about 2 sides of A4 ( You remove one bug, but find another 3 things wrong with it ), but I find it really helpful so nothing slips off the radar ( In theory ).
Oh, and smoking lots helps. Takes your mind off what you're doing and gives you a little break, and then things just kinda click into place.
Squize.