I began developing games as a hobby in 2012. It was just for the fun of creating prototypes with different ideas. Here is some of the lessons I learned making them.
A musical shoot’em up
The first prototype I developed in 2012 was a small shooter where you had to avoid musical notes on a sheet music or shoot them. Musical notes produced sounds when they appeared, so the level was created by music. I thought it was a cool idea if I knew how to compose music. I used Flixel with Flex SDK to create this prototype.
3 lessons I learned:
- “Shooting” is not fun, “destroying objects” is.
- A level design is like a melody. It needs rhythm.
- Adding debugging commands at the beginning of the development is really important (reset lives, reset stage, next stage …).
The maze games
Fast forward into the beginning of 2014, I started creating a maze game. Why ? When I heard about Unity3D and how easy it was to do a 3D game, I wanted to do a procedurally generated survival horror game with a FPS view and enemies who emitted lights (to see them in the corners). Because I knew it was very complex, I began with a prototype with random objects to collect:
The random objects happened to be rings because it was available in the starter assets pack of Unity3D. After implementing the Xbox 360 controller compatibility and online scores, I decided to move to the manor prototype. I bought a manor asset in a Unity asset sale to see how it was made and to prototype my survival horror game. The manor assets was made for Unity Pro and in 2014 it was still very expensive ($1500 per platform). So I used my free month to learn how it worked (rendering, shading, lighting …) and then I returned to Unity Free.
3 lessons I learned:
- Making a first person view realistic is hard. You have to make small camera movements that give the feeling that you are walking. If you do not, the player thinks he is flying. If you move too little or too much the camera, it is annoying.
- Making a simple corridor takes time. I just planned to do “corridors between rooms” in my design document. When I started adding them I was given the choice of width, height, color, depth, texture, angle … Spending time designing different corridor was not expected.
- Playing a modern FPS game on a computer is really difficult for non-gamers. The controls are not intuitive at all. Classic Doom controls are more intuitive because you do not have to use the mouse and you cannot look at your feet or at the sky by mistake.
Unreal Adventure game
In March 2014, Epic released Unreal Engine 4. For $20, I could use an engine for professionals without limitations. I subscribed and tested it for a month. At first, I could not download it. The servers were down. Then, I could not install it. It needed a patch. Finally, when it worked, I was happy. I tweeted “I tried
#unrealengine4 Blueprints and it seems really powerful. I am impressed.”. I could do everything I hoped … for ten minutes before it crashed. And there was no hot reload. It felts like I just spent my time relaunching the editor. So I returned to Unity3d working on the manor prototype until Epic fixes the engine.
At the end of 2014, the UE4 bugs I had seemed to be resolved so I bought it again. It worked. I made several prototypes. One of them was an adventure-exploration game. I’ve wanted to do a game where you had to discover the story through exploration like Bioshock without the shooting. But it was hard to do because I was not as motivated as before.
3 lessons I learned:
- Being able to do what you want is sometimes more complicated than working under constraints.
- Having a coherent world requires that each group of nearby objects (buildings, vegetation, etc.) possible is cohesive.
- A long game requires very long play tests so it is especially hard to do when it is on your free time.
To motivate me again, in February of 2015, I’ve decided to make a small game in 3 months on Android that I could show to everyone I talk to to improve it. This game turned out to be “Too Many Cubes“.