Design has been on my mind a lot lately. Liferaft has recently been in a kind of growing-pains stage which is a point in the development where Mike and I fumble around and try to learn how to develop video games. That’s actually a lot of it for us. Each new game we do is usually a whole new challenge with a lot of different problems that we’ve never solved before. The pain comes in the form of arguments between Mike and I, general frustration towards not knowing the best way to do something and all the shitty feelings that come with that.

Right now we’re working on the different levels or “areas” for Liferaft. We’ve never really done that before, not in this scope, and it’s scary. We fleshed out “level 1″ with a fair amount of detail here but that painted us into a serious corner given that we haven’t used our extremely precious friend-playtest-kleenexes yet. It took a very serious conversation and a bit of collapse on my part to realize we need to scrap that entire level and go back to the drawing board with a different approach. It was difficult, but those are the kinds of moments that we must concede to in order to make the best game we possibly can. It also helps with sanity.

Back at the drawing board Mike pointed to the Valve approach towards level design. It’s hard to argue with Valve when it comes to level design. Essentially they ward against putting in any detail into the level before it’s been well tested thoroughly and gone through whatever incredible gauntlet they have over there. The tough part about that for us though, is the gauntlet. We don’t have a gauntlet to stock with dozens of testers nor the time to build or even manage one. Nonetheless, the nugget we can take away from their process is to simply make quick drafts, undetailed “debug-mode” platforms on a black background. That’s how we’re designing levels now and it’s already starting to be a relief. What was once a gigantic task to create the all of episode 1 bit by bit is now something that we can visualize by playing these “skeleton levels.” Even if they’re bad, it’s still something.

Beyond that, there was also some trouble with the manner in which we thought about the levels.


Let’s think about levels as structures for gameplay and reduce them down to their core values in order to better teach the player how to play the game. We can’t just expect people to understand everything about our game, we don’t want them not knowing about an ability they could always use only to find out from a friend and have a worse experience.


Let’s think about levels as a physical structure and a real building. Goss grew up in isolation for decades, let’s make it less like a playground and more like a prison with a flood of new and exciting experiences all at once to drive home that feeling of freedom when she finally breaks out.

So those are basically our thesis statements for the argument and it came down to a matter of stamina really. It was a very hard fought dispute from both sides and in the end I think we made the right decision to go with Mike’s idea of focusing on usability as opposed to drama.

While the graph on the right is more interesting, it doesn’t mean it’s a good starting point. By having a clear-cut stairstepped graph to go from and be confident with we can then have more freedom to make modifications as we edit the levels so that the graph on the left begins to look like the graph on the right. But I guess we’ll have to see. Like I said, we don’t really know what we’re doing.

I guess part of me relies on my art background to rationalize this direction. When drawing it’s always best to start with a loose structural sketch of whatever it is that you want to draw. Using the simplest shapes to organize the form into something that may eventually tell a story is always a great starting point when drawing/painting. Between these two methods proposed above, I think it’s clear now that I have taken a step back that the more organized form would be Mike’s way. Sure it’s less romantic, but there’s nothing romantic about perspective drawings or boxy figures. They’re purely functional, you add the story/drama/detail later. Ok, well there’s something sexy about gesture drawings…

Hmmm, what is the game equivalent of a gesture drawing? Cactus? He is the Zen master.