The weekend community challenge - a postmortem

Alright, since everyone seems to blog about their experience with the latest weekend community challenge I feel kinda pressed to do the same.
No, in all seriousness, it was an awesome experience and the first of that kind for me.
It's really something to come up with a game in 72 hours for a given theme, it forces you to prioritze your ideas if you want to deliver on time.

Here's an overview of the stages the project went through.

1. Initial brainstorming

I basically brainstormed ideas that would make up a cool game. Every thought went in a document without any weighting.
I think this process took 30-45 minitues. At the end there was a plain old text document filled with short lines describing a certain feature.
e.g:
implement sprite system, tentacles made up of segments, shoot cannonballs at tentacles, tentacle destroys different part of ship, floatable treasure chests let you repair ship, ...

2. Prioritization (is that a word?)

Next step was to choose in which order the ideas were to be implemented. Since there was a 72 hour timeline, it was clear that there had to be a functional product throughout the entire course of development. Implementing the sprite manager was certainly going to be the #1 since without it there wouldn't have been any graphics. I went for 2d because I'm not that good at arts and building 3d models would have definetely screwed over the schedule.
So I went for an ordering that would pretty fast give gameplay to test. Actually I got a little distracted and ended up in fine tweaking somewhere in the beginning but got back on track later.

Notice that at that point there was no idea that was thrown out right away. The not-so-important where just put at the end of the list.

3. Development

Starting out with a 1024x1024 texture I just drew some crude shapes that could be used for testing. Then I implemented the sprite manager (basically I cheated because it was already half done from another project of mine). After implementing the first gameplay logic it became clear that the first vision of the game was basically unplayable. Initially you had to click a cannon and then click on the tentacle it should shoot at. However, with the moving tentacles and the small cannons it was pretty hard to aim and it just didn't work. I think that was sometime around day one, so fortunately I had plenty of time to change it. I settled for a much simpler control scheme where you just point and click.
Next I implemented the repair treasure chest.

With the deadline closing in and having barely implemented 2 gameplay elements from my list I decided to scrap the remaining features and just focus on polishing what I already had.

4. Tweaking

Got distracted again and spent way too much time touching up the lovely programmer art. Playtesting revealed that the game was way too hard so I tweaked it a bit. Unfortunately it was already late and I had other stuff to do, so that was going to be the release version.

Conclusion:

I had much fun with this challenge. I'll definetely join the next one too. The only thing that needs to change is playtesting earlier. I'm not really satisfied with the game as it is now. It is pretty repetitive and there is no real tactic to winning. You just have to shoot and shoot and shoot and...

If you would like to give it a spin: http://www.tab-away.de/kraken

I'd love to read your feedback on both the game and this article.

Cheers.