A Postmortem on Murder Mitosis
I felt like writing up my very first postmortem. This was not only my first Ludum Dare, but also my first game project that was actually released in some form.
We went into this as a team of two programmers with no artist on board and surprisingly little preparation. My team partner had never even worked with the Unity engine before this jam. Nevertheless, we stocked up on junk food and caffeine and just went for it. The next 72 hours were the most productive we had been in years...
What we ended up with Murder Mitosis. A fun little multiplayer arena game about strategic positioning and board control in a small petri dish, that gets more and more crowded as the game progresses.
https://www.youtube.com/watch?v=lmKzk9eatf8
What Went Right:
I am very happy with how well we managed to incorporate the theme. From the moment it was announced, we knew that we not only wanted something that resembled the theme visually, but we wanted the theme to be intrinsically linked with the game's mechanics. Styling the game around microbes in a petri dish was just an added bonus.
The game ended up being about tactical movement and trying to find space (and food) in a small arena. Little creatures are out to kill you and your only weapon is to convert them to your color and make them hunt your enemies instead. Every time you do this, however, these creatures multiply. This makes the arena more crowded and more deadly as the game progresses. The goal of the game is to trap your enemies, thus cutting them off from their food and starving them.
We had no artist on board, so the visuals had to be something simple. The creatures are just small colored bubbles and the players are simple boxes with cute pixel faces on them. I added in some shader magic to create pressure effects and an organic wobble to make everything seem more alive.
From then on I took over art duties and programming the AI. I quickly threw together the energy bars, the menu elements and a logo. I added cute little faces to the creatures that would change expression whenever they collided with each other. Very simple stuff really. I am quite pleased with how surprisingly pretty the game turned out in the end.
The AI on the other hand took a little longer to build. From the moment we had decided that we would do a multiplayer game, we knew we would have to have AI players in there, or most people just wouldn't be able to even play the game.
I ended up implementing a simple mesh navigation system. AI-Nodes around the arena are running subroutines to detect the amount of food and the number of enemy creatures in the vicinity. From that data a quality score is calculated. Each AI player simply tries to get to a position with the most food and as little enemies as possible nearby---without touching any enemies.
What Went Wrong:
Developing the AI ate up way too much of my time! That is not to say that it wasn't absolutely necessary to have good AI players. Instead, the problem was that I had it in my mind that I could solve that AI problem using local information only, instead of coding an underlying graph system. For some reason I thought that a graph and a simple navigation algorithm would take too long to build. So rather than going for the obvious solution I started tinkering. Needless to say, if I had just gone to work coding the graph solution right away, I would have saved half a day of work.
Another part that I am not happy with is the lack of options in regards to control schemes. You can customize your player character, but you select your input method from the button that you press when joining. While it's a minor issue it keeps bugging me personally. I guess that's just one of those features you have to eventually cut when doing a jam.
One thing that we should have definitely spent more time on was teaching players the game. When we first uploaded an early version of our game, people had trouble figuring out anything. Luckily, we still had enough time to redo the tutorial text in the menu screen before the jam ended. However, we still see some early confusion when players play the game for the first time, and it can take a moment until they catch on.
All these issues are really just symptoms of our biggest problem: Not having time to let anyone but ourselves playtest the game. This was mostly due to neither of us leaving the house. We had completely shut ourselves off from the outside world for the full 72 hours, which meant that we had noone in reach to play with.
Which brings me to the actual biggest problem of all: It was early sunday morning when we had suddenly run out of caffeine! With no time to leave the house, we decided to just order some stuff. Yes, there are people who will bring you caffeine. At least they claim that. We paid online, but noone ever showed up... :coffee::rage:
Overall:
Participating in the Ludum Dare Jam was an amazing experience. A great demonstration of what can be achieved in little time, by just neglecting sleep and basic hygiene...
It was great fun, if you are interested in what we'll be turning this game into after the jam, follow us on Twitter @HastilyAssemble or on YouTube.
If you haven't played the game yet. Do so now and tell us what you think. We really appreciate your feedback and any creative ideas we can steal.
Thank you all for a great time! Now off to play more games.
Play Murder Mitosis: https://ldjam.com/events/ludum-dare/38/murder-mitosis