With June over, I am writing this post to say my 1GAM experiment was a failure. There are many reasons why, but I also believe the 1GAM concept is only effective for certain game types and likely not the kind of games I prefer to invest my time in.
Can You Really Find Fun In One Month?
The heading presents the heart of the problem: It’s not that we couldn’t develop the necessary parts of the game in a single month, but instead that I could not settle on core gameplay I found to be fun. Part of this is because the core gameplay could not be adequately tested without AI, and when you iterate gameplay you must also update the AI.
The Vision: Fun, 2D space combat.
What We Tried: AI ships flying in formation and patrolling. They break formation when the player gets within a certain range and engage the player. We tried this in small quantity, and while it was cool, it wasn’t fun.
Most of the month was actually spent on the AI. We messed around with different times to break formation, different behaviors like strafing and flying backward, to even changing the whole feel of the game from tight controls to momentum-heavy ‘drifting’ and we’re still not certain what we need to do to make the game feel fun. The only conclusion we came to was that ‘drifting’ felt a lot more fun than tight controls, so we plan on keeping that.
Ideas Worth Trying
Floating Asteroids That Damage Ships
Technically, this was supposed to be part of the original implementation for June, but we ended up spending so much time trying to figure out the AI that we never got around to this. Will adding asteroids as a dynamic make combat more fun? I think so.
Bouncing Collisions and Shields
This is actually more work than it sounds since we are not using the Unity physics system, but instead precisely controlled movement with artificial (controllable) drifting. We didn’t want to use Unity physics because we wanted to be able to better tune the gameplay ‘feel’ and so when we add shields and the ability to bounce off objects, this should add more liveliness to gameplay.
Even if they’re AI-controlled, the idea of being a lone space fighter against a bunch of enemies is pretty unappealing unless you have some really badass capabilities. We’ve developed our game with Factions in mind and adding AI wingmen that follow the player and assist in engaging enemies shouldn’t be difficult, although it will require a refactor of our AI combat code that currently only looks to engage one enemy target.
We use gut feeling to measure fun. Fun is different depending on who you talk to, but as a developer I know that if I don’t find the game to be fun then I will not have the desire to continue working on it. Therefore, the game needs to be fun to us first.
When we get around to trying the above, the main question we will ask each time is not “Does it feel fun now?” but instead, “Does this feel more fun than before?” Fun in games often is not a spontaneous discovery, but rather an iterated process. This brings me back to my main point – 1GAM is an interesting experiment, but due to my personal standard of never developing an unfun game, I don’t think it’s for me. Games with simple AI requirements or don’t require any AI at all are much better suited for 1GAM, as the amount of experimentation required is much lower. Perhaps if we extend the time to two months (1G2M), we will have ample time for fleshing out fun in games where AI is at the forefront? It’s worth a try.