4k Flash Game Post Mortems
Gompers Post Mortem
Matt Slaybaugh
http://acetio.com
http://skeptictank.net/games
I learned of the game through the MochiAds forum a few weeks before the deadline. I enjoyed UrbanSquall’s “Battalion: Nemesis” which motivated me to check out the rules of the contest, and the 4K restriction was compelling to me. The best design often comes from strict parameters.
I explored a few different ideas, and the ones that seemed to have the most potential (because of the size constraints) were the ones that were similar to the Atari 2600 games I played as a kid: space invaders, missile command, asteroids, etc. - infinite levels - no real way to win, only delay the inevitable death - fight off endless enemies that get faster each round
One idea I made into a game in school in 2004 was Gladiator which had a central character fighting off animals in an arena. The gimmick was that instead of infinite ammunition, as in many shooting games, this character had a fixed number of arrows to shoot, and then had to gather them up if he was to use them again. This idea of strict constraint seemed to match the idea of limiting the filesize of the game itself and that’s the idea I chose.
I was going to use the contest as an excuse to learn more about ActionScript 3.0, since I’m woefully behind in moving past AS2.0, but there wasn’t much time and I knew I could do it in AS2, so that’s what I used.
The first iteration was a cave painting that came to life: primitive drawings of bison and a hunter with a spear moved across a picture of a cave wall. It looked cool, but I had to simplify the graphics in order to make the 4K, and after that the graphics looked clumsy rather than just primitive.
I wanted to simplify the code as well, and I spent time coming up with formulas for everything that were all based on A] the level number and B] the characters health points which began at 10 and maxed at 10. i.e.: - Num arrows = Level - Num animals = Level - Animal HP = Level - Animal width = Level + 20 - Arrow speed = Distance/health-7 - Character speed = Distance/(30-health) - etc.
A lot of the time spent on the game was finding the right constants for these formulas. When the game was too hard I would make the animals stronger, when the game was too easy I made them weaker or the character stronger. The difficulty came in balancing the game over time. Since the difficulty of the levels was algorithmic, as opposed to explicitly designed, often the difficulty at the beginning would not match a few levels in. As it is now, you can play several levels before it gets too hard, and you have several levels of very challenging at the end before the animals overwhelm you.
After playtesting hundreds of times the game felt really boring and I knew I needed some obstacles or a maze or something. Everything I tried put me over the 4K limit and I decided that the fallen corpses would make good obstacles, since they’re already on the screen, and that would also put the player in control over the emergent maze, rather than making that random. I knew the player would have to learn that the dead enemies made things slower, and that’s why I added the seaweed in the center, in order to introduce the concept that there were multiple types of terrain.
The final theme and graphics were the last part of the game I came up with, and it was also the part that seemed to confuse the judges the most.
The gladiator/animal theme looked bad with the simplified graphics, so I began looking at the old classics for inspiration and for a while used the images from Berserk in place of the hunter and animals. Those gradually evolved into the ones used in the game. The main character looked like a ghost (the name of the game was “Ghonders” for a while ['ghost' + 'goners']) or maybe a mermaid and the dead bad guys looked like mushrooms or maybe boats when flipped over. I wanted the colors to really pop, and look good, and settled on the blue background and the yellow character, and it was then that the mermaid vs. boats idea came up.
I’m happy with how it turned out. There are things I would change now, but not many, and there are always things to change in hindsight. I’m surprised the judges seemed hung up on the graphics, since it really doesn’t matter. I doubt they would have the same reaction if presented with Pac-Man for the first time (”giant yellow head flees ghosts, eats pills”). It also seems they didn’t play beyond the first few levels. That tells me the game was too easy in the beginning. I wanted the difficulty to ramp up gradually, but with such a simple interface, it was just too simple in the beginning. And the winners surprised me. Gompers is a real game, while the winning ones (at least the dungeon one) seem more like programming exercises. But that’s just my opinion.
Other games I’ve made have been judged by the “masses” via comments in game portal sites and most of the feedback I’ve received on my games have been players at Kongregate or NewGrounds or other sites like those. They tend to be harsh critics and also seem to have rather simple tastes. This contest seems like there is a higher bar for creativity and innovation, which I appreciate.
