Gravity Assist is my first Ludum Dare game – my first participation in any game jam, in fact. Players launch a particle angry birds-style through a map of gravity fields, repulsor fields and others, trying to tailor their trajectory to maximise their score. The game keeps the trails of previous attempts, guiding your next attempt and getting prettier as you explore the space.
What Went Right
Unity and CFXR: Using Unity was absolutely the right choice. While everything in a 48 hour game is necessarily hacky, and my physics code is no exception, Unity allowed me to worry about the game more than the physics, and to quickly put together graphics which, I think, look pretty good. CFXR allowed me to quickly iterate through sound design too.
Concept and Scope: The concept ended up working out really well, which I feel was probably as much luck as anything else. It’s simple, lends itself to player choice while being satisfying and is intuitive enough to allow depth without swamping the player with controls and options.
Replayability and Reward: Once I realised that the strength of the concept was the player’s ability to try many solutions and compare their success (on their own terms, not necessarily just score), I focused on making it easy and rewarding to do so. The trails help this hugely, allowing players to even treat it as a drawing game if they like, while pressing space to restart quickly encourages iteration and “just one more try”. The comments back this up, fortunately! I also wanted to ensure players got satisfying feedback for their actions, so the objects in the game light up and make sounds when activated.
What Went Wrong
Unity: While Unity is great, I spent far, far more time than I wanted trying to get GUI to work in a nice way. Even putting the tutorial text in was painful, and if I had a better GUI solution I would have probably been able to double the amount of levels.
Difficulty and Physics: It’s not an easy game, in some senses. It’s forgiving in that you decide when to advance, but some levels are frustrating and could have done with more testing. Partly this is down to the hacky physics, which don’t react intuitively in some cases. For example it’s impossible to get a particle to orbit an attractor.
Feedback: Similarly, many comments have mentioned things like better indication of launch angle, and previous power. These are things I would have liked to add but ran out of time. When the screen fills with tracks, it’s all but impossible to judge your next shot!
What Would I Add?
The concept has shown itself to be a fertile one, and I have a list of ideas for how I would expand the game. Decelerators, switches, moving elements, ricochet plates and loads more would all suit the game perfectly, hopefully without adding too much complexity to overwhelm the player.
Overall, I’m very pleased with my first entry. As I say, I suspect hitting on this concept was more luck than anything else, but I do feel I capitalised on it pretty well given the time. It’s been really rewarding to see comments confirming the addictiveness and aesthetic choices. I’ve played some great games by others as well, so basically the experience has been fantastic and I’ll see you in April!
Check out my blog at gameoflyje.wordpress.com or tweet me @six_ways!