I began the one game a month excursion last month (October.) I made 2 games that month, one was for a game jam, the game was titled Arc. We are hoping to polish this game up some more. That game went really well and is fun.
The other game is Cluster Puffs. This game took the bulk of the month and is a systems driven game. The idea behind it is that you have to kill a bunch of cute and cuddly creatures for their own good (or they will eat all their food and they will all die.) There were a myriad of issues with this, the most pervasive being ‘why is this fun.’
After getting the initial system working with spheres, I found that it was indeed a neat simulation and if you let it go their population will boom and bust. But the ease at which the player could cull the population lacked challenge or thought.
The game went through many more iterations, some for performance reasons, other for design reasons.What I finished with s the best permutation of the thing as a whole, but lacked some of the charm of the earlier systems. In the future I will be answering these design questions up front, and not hoping the answer comes to me as I work (even though the answer will of course, be wrong.)