Recently, I’ve stumbled over this page:
http://www.squidi.net
A guy called Sean Howard decided to come up with 300 new, unique game ideas. He posts one game idea every day so he will be finished in less then a year. As we speak, he reached game number 46. I like his approach. It’s a strategy I often use in my projects: go wide instead of going deep. Instead of fixing on one idea, come up with many ideas and select the best one afterwards. Ideas are like urine: the midstream urine is the best. 
However, I have come critique for the project. First, it seems like Howard is taking some short-cuts. He sometimes uses a strategy which I would call “instant game design”. Instant game design comes in many flavours, the most popular is the RPG flavor. The recipie is simple.
1. Take some random game,
2. Add a standart RPG mechanic on top
3. Voilà
Although the resulting game can be delicious, it doesn’t really show how a great cook you are, does it? Howard pulled this particular trick at least once. He also used other instant game design strategies, for example the flavor tactics where you make an otherwise real-time game turn-based and thus exchange skill for tactics. He also used the flavor sim where you take a big, familiar system (theme park, city, family, dungeon, hospital) and put the player in charge of it.
Another short cut is to introduce some kind of dark/light duality. It seems like this particular idea is source of much inspiration for many people as it appears in lots of games (Black & White, Ikaruga, Zelda: A Link to the Past, Metroid Prime 2: Echoes, etc.). Again, the games I’ve listed here are excellent and even my colleague Yu-Chung is working on a game with this mechanic. However, Howard alrady based at least 6 mechanics on this idea (Negative Space, Neg-Space Wars, Grey Matter, Flipside, Flipside Inversion, Demon World). I does seem like some kind of obsession.
My last critique is maybe the most serious one. Contrary to popular belief, the initial game idea is maybe 1% of the process called game design. The remaing 99% is laboriously derived from a series of prototypes, play testing, rediesign, more prototypes etc. Game Design is a long, iterative process. Quite often, you arrive at the bitter conclusion that the idea you had doesn’t work at all or simply isn’t as revolutionary as you thought. So although Howard is creating an impressive portfolio of intriguing ideas, he is not doing any real game design.
That being said, I am still interested in what he is going to come up with so I will visit his site frequently. I whish him all the best.
i think you brought this together pretty precisely, i just want to sum it up (again):
-game design is a long process, one shouldn’t mix it with the collection of ideas.
-it’s great to be able and to have a (as big as possible, but then necessarily categorized!) library of ideas at hand
i personally believe, that this kind of library keeping is somewhat that tools are (typical tools like a hammer, a screwdriver, etc.). any time you have a problem (in search for a solution to a gamedesign situation in which you need anything / something “more” or “better”) you can try to find it from scratch. or simply use a tool to solve the problem. this probably is a poor example, because you don’t fix a leaking pipe with sticking a hammer into the leak, but by using the hammer to fix the pipe. but typically gamedesign problems don’t need a small chunk of detail to solve them, but a complete change in some mid-level range. i.e. you have a jump’n'run baseconcept, and you’re not satisfied with your enemies. so you want to drop some of the enemie types and add new ones. and those should of course fit as good as possible to the main gamemechanic of jump’n'run.
here it is probably pretty clear that a library of ideas will help a lot.
i’ve got a not to small library of scrap-books, papers and a tiddlywiki full of ideas myself. but all texts there tend to be very loose. a very strange style needing to connect a lot of thoughts together and thus most often only understandable by myself. (and that also only after some thinking.)
what do you think: is it worth to take the time and write everything up in electronic form and try to write it detailed and understandable for other gamedesigners?
Yes, most defiantly. By sharing the ideas and actually devoting the time to bring them to a presentable form, you automatically develop the ideas further… or expose their weaknesses, which might be just as useful.