Design-side Level Creation for Streamline (and the Delights Thereof)

Level creation always brings with it a host of challenges (and delights). Bursts of creative productivity and lulls of head-smashing lack thereof seem to happen with every map or level I’ve worked on. Streamline, in particular, represents one of the few, precious games that I’ve been part of the development team for since day one. Working on a project from conception always helps with buy-in and ownership—not that that was ever the remotest issue with Streamline (the outbursts of laughter during each and every playtest can attest to that). But ideating on an already existing foundation of content and content plans is never as compelling as being a part of that foundation yourself. Since I am part of that foundation, I will speak to the particular challenges (and delights) of design-side level creation for one level in Streamline in particular: GameCon.


GameCon was selected as a level theme super early on. Given that we were spinning up a new game and PAX East (which we’re local to) was on the horizon, it was pretty likely we were going to have some kind of booth presence. The earliest stages of content planning, levels-wise, typically involve some manner of spec-pitching with theme ideas, image boards, gameplay concepts, and research. Since many of us have been going to PAX East for several years and it’s nearly always been at the same venue—and we were itching for excuses to include indoor maps—the research part of the pitch was remarkably simple. If we were going to be showing off a GameCon map at a booth inside the venue PAX East has nearly always been at, my goal from the outset was to make the map immediately recognizable to anybody who sees it on the show floor.

PAX East

It helps that there are a handful of super defining features of that building.

This presents a very real challenge. I am a big advocate of self-imposing restrictions to help make creative decisions versus the potentially paralyzing “anything goes” technique. If you don’t understand what I mean, think about writing a paper on “anything you want” versus writing about your favorite food; having a tighter focus often makes a creative endeavor simpler. However, everything in moderation. In an action game built around frenetic parkour traversal and “running lines” as we often call it in the studio, immediately restricting the environment around the structure of an existing building might be going too far in that direction. When I first brought the concept spec-work into Unreal (the game engine we’re using) to begin crude layout blocking, I immediately realized that the defining features of the show floor—notably the skywalk and the multi-pronged support pillars—did little more than get in the way.


The single biggest challenge throughout the blocking process turned out to be just that: trying to shoehorn compelling traversal routes through and around these identifying structures. I think anybody that has been to PAX East will immediately recognize the level in Streamline (our environment artist, Steven, will be writing more about this next week), but with just a few weeks left before the event, it remains to be seen whether the gameplay and routing is up to the lofty bar the art department continues to set. With a deadline so close to PAX East, how could gameplay remain a relatively unknown factor? Shouldn’t the map have been playtested many times even before the layout moves from the design department and into art land? Absolutely!

But, bugs.

Perfect. Ship it!

Perfect. Ship it!

And this one made it into the Ogles bucket! Allow me to explain. Bugs permeate the whole game development process. There are big ones, requiring significant portions of an engineer’s time to fix, and there are simpler ones that are far less scary. But they are everywhere, and rarely is one heinous enough to take up the time of our CTO, Dan Ogles (he’s roughly the engineering equivalent of a final boss). One bug had more or less blocked the majority of playtests in GameCon, and it continued to block playtests in the level well after the point at which the map swapped departments from Design to Art. But the show must go on! With a relatively small team and a slew of maps we’re targeting to include at launch—not to mention the sanity of our team after playtesting our other level entirely too much—we need more than one map.


You might be wondering where the “delight” part in the title fits in with all of this. Admittedly, this post has largely been about the challenges around designing GameCon as I recall them, trepidatiously allowing those synapses to fire again. Honestly, part of it is just like the evening news: negative stuff sticks out more readily than the positive stuff. But, the truth is, delight permeates throughout; it’s less ebb-and-flow and more constant. There’s an old game dev meme floating around that likens designers (specifically, level designers) to children in a sandbox. As self-deprecating as that may initially seem, it’s a good comparison to the work we do. We’ve got all these tools at our disposal to make player-facing content and, oftentimes, the sky’s the limit. We play with what we’ve got until the fun finds itself. So work is like play, and if that’s too saccharine sweet for you, I don’t blame you: I used the whole “delight” thing to more or less tie the article together.


Cool Guy Josh

But, since I am being sincere, I’ll relate the one GameCon story in particular that fills me with joy whenever I think about it. A common challenge I’ve had since discovering the fun of bouncers (basically, quick and satisfying vertical traversal) in our previous level is how to replicate that gameplay in other maps without it feeling contrived. One of the earliest learning moments from the first, very buggy playtest of GameCon was “the verticality problem.” The cavernous nature of the real-world venue, when laced and lined with catwalks and traversal apparatuses, was just too vertical for Streamline.

GameCon saw at least two significant iterations where I removed catwalk elements and flattened down the verticality. There are now (roughly) three height levels players can traverse between, the highest of which is unreachable from ground level without transitioning via the second floor. This was all well and good for the front portion of the map, but the escalators enabling this transition between levels don’t exist in the back portion of the real-world venue. Again, perhaps being a bit too restrictive on myself in order to maintain my primary design goal, I wracked my brain throughout development to find a fitting solution to this problem. This became another big, personal challenge that really soured me on the map as I reworked that area several times.

Then, as it so often does, inspiration struck after hours of abusing Google image search for a vertical traversal solution:

Go for the head!

Go for the head!

That is an ECA-sponsored Rock’em-Sock’em inflatable Joust booth at PAX East, circa 2013. Is it contrived? Maybe! But it was a real thing in roughly the same area of the same venue for the same convention. And it totally fits into Streamline’s trappings! And of course you can bounce on an inflatable object! It’s a video game! So, after much dismay there was much rejoicing. I spoke with Steven about it and got him on board. Before long we had the asset in the game and it felt great. It blows open the back portion of the map; so much so, in fact, that subsequent playtests have brought us to begin tinkering with a launcher in the front portion of the map. As of this writing, it’s still being developed, but if it comes out the way I’m imagining, I’m sure it’ll also be a hit.

So, if you are going to PAX East this year, be sure to check out Streamline. And if a certain GameCon map makes it in (oh God, I hope it does), when you first Kong-vault onto that bouncer and soar through the rafters, remember the moral of this story: designing compelling levels is challenging (but it is also a delight).

Josh Hufton
Senior Game Designer

Leave a Reply

Your email address will not be published. Required fields are marked *