Fractured Soul: postmortem
Multiverse Factory was a team originally coordinated by Gummbum. It wound up including DR0ID, myself, Lisa Sau (who had to drop out), nitrofurano, JDruid, ldle, and kiddo. Somehow I got chosen to be the director/producer, so here's my postmortem.
I was uncertain how well we'd work as a team, since most of us had never worked together before. Given that, I thought we did a great job. Our game, Fractured Soul, is complete, it runs smoothly, and (for me anyway) it's kind of fun.
What went right:
- Dedication: Gummbum put in so much work over the course of the week, and DR0ID and I were also working really hard. nitrofurano was also a very valuable and dedicated resource, especially toward the end.
- Working together: Even though we're in a bunch of different countries, and we were communicating only over IRC, I think we made a good, cohesive team.
- Engine: It was a big help to have Gummbum's gummworld2 library to work with. He was able to adapt it to suit our needs and generally make everything work.
- Mechanics: We put a lot of thought into this, and it shows. I'm impressed at how smooth our platforming action and collision detection is, compared to your average pyweek platformer. This is due partly to gummworld2's handy timing algorithm. I had something pretty demanding in mind for the mechanics: I wanted your motion restricted to a grid without feeling like it's restricted to a grid. Along with ldle and Gummbum, I eventually got it working to my satisfaction.
- Music/sound: All done by JDruid. The atmosphere of the game is perfect and the audio is compelling.
- Menu integration: I love it when you can jump right into a game. I'm very pleased that Gummbum was able to eliminate all menus.
- Level editor: DR0ID's tiledtmxloader library (integrated into gummworld2) made it possible to design levels completely in Tiled, something I didn't even realize was possible when we started. This made it possible to have extensive...
- Level design: We wound up with a lot of content for a pyweek game. While lots of entries struggled to have 9 levels, we wound up with 26 (30 if you count the "menu" levels). They're small, but they required a lot of care to design.
- Practice game: We spent about two weeks before the competition on a practice game. This let us sort out many logistical issues ahead of time.
- Work management. Having someone on point to lead, coordinate, and make decisions really helped. It let the respective specialists focus effectively.
- Credits. DR0ID is handy with special effects, and our credits screen is a nice touch.
What went wrong:
- Artwork. We started with two artists. Lisa had to drop out completely because of a new job, and kiddo couldn't contribute anything until Friday due to technical issues. Kiddo designed the main character and drew all its frames, which is great, but we didn't get a chance to refine it, and the walk cycle is a little off. He's got so much talent, we would definitely have loved to get some more artwork by him into the game.
- Theme. Nine Times was definitely our least favorite theme. I insisted that we not adapt a different game idea to fit the theme, and so we had a very weak idea starting the week, despite a lot of preparation. It turned out okay, but it could have been much better with a theme we actually liked. One place this shows is the story, which I didn't really get in place until the last day.
- Work management. We had no efficient way of tracking/assigning/claiming high level tasks. We made do with the wiki, which was disorganized and inconsistent. I was asking "what are you working on now?" a lot in IRC. A dashboard to track current tasks would have helped. And it would be cool to look back and see how long we spent on certain aspects, and if those could be economized in the future. Maybe we should see if there's something lightweight and useful available, perhaps a Google app.
Overall I was very excited to be working in a team like this. This experience was new for me, and it was quite positive. I'm looking forward to the feedback on the game. Thanks to all who try it out!
(log in to comment)
Comments
It makes me really want to try working on a larger team too. Thanks for the post-mortem.
Pivotal Tracker to be a lightweight and useful way of tracking tasks (we didn't end up using it this Pyweek though).
I have found
davec on 2011/04/19 18:25:
It's a great game - especially considering it was created in a week :-).