I decide that I might use this upcoming PyWeek to play around with Python Ogre.  I have some experience using Ogre, Bullet, and a few of the other libraries in C++, so I don't think that this is going to be too difficult.  I expect it be a lot more fun than using PyGame since I use it so often and there is little left for me to learn about it.

After quite a while of trying to come up with an idea that fits "nine times" I finally have something.  Basically it will be a platformer with nine different endings -- so it has to be played a minimum of nine times to see them all.

There will be thirteen levels.  The first four levels will each have three branching paths and the remaining nine levels will be shorter and more linear leading to the ending.  Each path will also have some different story I hope -- if I have enough time to do more than just make different scenarios.

Here is a diagram illustrating how the level branching will be:

I also decided to not use Python Ogre but instead use PyGame, since I don't really feel like spending a lot of time making 3D graphics.

I think I'm wasting too much time drawing tile graphics.  Not much work has gone into the actual game yet and there is a lot left to do.  Maybe I should put graphics aside for a little while and focus on getting my game engine in place.

It's strange how it feels like I've only done as much as I would have in only a few hours if this was a Ludum Dare.  Maybe I should pick up the pace a bit.