I found some footage of the first prototype for Space Otter Charlie, and I was shocked at how many ideas made it into the final game. I cut together a little video comparing the prototype to the final game, as I thought some folks might be interested to see how the game evolved.

It was a fun look back for me, and it made me reflect on a few things along the way.

- If it's fun when it's ugly, it'll be even better with art - I start my prototyping with controls and mechanics. I don't want to underplay the role of great art in a game, but I find that making it fun without worrying about art style keeps the prototype fast and focused.

- Polish takes time, and it's worth it - Even though a lot of features got into the final game, a ton of time was spent smoothing out the rough edges. Player movement is the heart of this game, and we spent a lot of time dialing that in. The prototype captured the general feel I was going for, but from there, we did a lot to tweak acceleration curves and collision response, add sliding and acceleration, and provide feedback through animations and sounds. It was being tweaked for at least the first year of development. You can really tell the difference from prototype to final, and getting that feel right was worth all that time.

- Capture the process - I always mean to hang on to old builds, video clips and concept art, but in the heat of development, it often gets lost or tucked away somewhere. This time around I spent a little more effort archiving the process, and I'm glad I did. A couple of reasons:

When you hit plateaus in development, where it feels like you aren't making a lot of progress, having a reference point for how far you've come is helpful and inspiring.
When you are stuck for ideas, looking back at prototypes helps you find ideas that you might have forgotten about along the way.

- Vertical Slices help define the game - Vertical slices are often talked about as a milestone along the road to completion. I like to prototype a vertical slice - meaning a prototype that captures enough of the game features to tell how they feel in concert with each other. Features don't need to be complex, but I find it useful to stub in stuff like objectives, rewards and upgrades if those are going to be central to the game. The initial prototype for the game took about a month, and even though most of the code and assets got thrown out, the time was probably the most impactful month on the whole project.