Rogue Agents: Bootstrapping Crazy Ideas Within A Larger Project

Before attempting any of the spy-themed puzzles in Schell Games' recent VR title I Expect You To Die (IEYTD), players experience something else entirely: A coaster-esque intro credits sequence that choreographs everything from cars to coffins -- an immersive take on the iconic "Bond Intro." This intro has been highlighted in reviews for successfully setting the tone of the game, and for being a unique VR experience in its own right. The song was even a nominee for "Best Original Vocal Song - Pop" at this year's G.A.N.G. awards.

It may be surprising to know, then, that originally there was going to be no such intro.

Everyone who works in games has, at least once, been inspired to attempt something "out there" within the context of an existing larger project. Most of these efforts don't go anywhere, but the IEYTD intro exists only because such a personal project successfully snowballed into something the whole team wanted. Hopefully, I can draw a few key lessons from the story that will help anyone else who wants to go rogue.

Find the Gap

The first step, and one that is skipped surprisingly often, is just knowing what people's gut response to a concept is. In the case of I Expect You To Die, it was an easy pitch: "Hey, we have this Bond-themed game, wouldn't it be neat if we had a Bond-style intro?" Even if the pitch seems obvious to you, people have different tastes, and by seeing how people respond initially you can prevent yourself from sinking a ton of time and heart into something no one else wants. And hey, if your conversations go well enough, maybe the idea will just be integrated into the project immediately, and you can skip the rest of these ramblings.

The situation that best lends itself to bootstrapping is this: The conversations go well, but in a wistful sort of way. People fantasize about how cool it would be, if only it could be done. If only we had time. If in response to these statements your thought is I could do it, I have time, and importantly I want to do it, you've got a good candidate for a rogue operation. It's even better if working on it has value to you beyond potentially being in the game - perhaps you are dusting off old skills, so even if it doesn't go anywhere, you are still getting something out of it.

For IEYTD, the person who took up that initial mantle was, well, me. I'd dabbled in 2D animation since I was a wee lad - it wasn't a skill I used often, but it was one I had in reserve. I'd become familiar with Unity's 3D animation tools through previous projects, and I knew enough to throw some unlit shaders onto assets to get a stylized look. And I sang in the shower sometimes, which made me totally qualified to sing the first version of the intro, right?

Eh. You gotta start somewhere.

Convert Trash to Treasure

Even if you've done a good job finding a gap, getting something like this going is a massive undertaking. You are willingly taking more work onto yourself, but you still need to respect your needs; get to where you want to be at a quick but steady pace, giving yourself plenty of time to relax and sleep. To do this, it's good to take every possible shortcut, and cover for the areas where you are weaker.

Every game project, especially one as big as I Expect You To Die, goes through many iterations. Ideas and environments are tested, and while many become part of the final product, many more are cast off. These abandoned assets and code are the perfect place to give yourself a boost - especially in areas where you are weak, where even the abandoned might be better than anything you could make on your own.

For the intro, my weakest point was, by far, modeling, so I borrowed assets liberally from existing levels. Even in its final version, the Intro sequence is twisted memorial of previous level concepts. The tesla coils at the end were an earlier version of the virus vat from the lab level, and the missiles were cast off concepts from the lodge. Oddly-posed male models used to make sure environments were scaled correctly became our "seductive" spy puppets. And where I couldn't find assets that worked, I cobbled together the nearest approximation. The most painful example: originally, the eye-logo spiral was lopsidedly approximated by unity cubes

Here is the key thing to keep in mind while harvesting: whatever you are cobbling together, you don't need to be as good of an animator as the animators, or as good a designer as the designers; it doesn't need to be perfect. If you want it to become part of the main project, going solo is a means to an end. It just needs to be good enough to inspire others to get excited about how they could make it better.

And thus we get to the next phase:

Build Outwardly

The core idea here is simple: don't try to recruit everyone before your experiment has momentum. Let people know what's going on, and let them come to you.

For our intro efforts, the most early converts were those interested in the creation of the song itself. Even if the whole animation thing didn't work out, we could have slipped the theme song in somewhere. Bonnie Bogovich and Tim Rosko could actually sing and compose, and Jared Mason - one of the studio's main writers - did a lyrical pass that made the rhyming scheme more interesting and song actually flow. By tapping into their early excitement, we quickly developed a well-rounded version of the song that we could build on.

It's important to note that if you are bootstrapping, you never want to depend on or demand the contributions of co-workers - that's how you wind up creating unwanted overtime for other people and turn your passion project into a burden. But if someone is excited, and thinks they can make your Unity cube version of the eye logo actually look like the logo it's supposed to be (thanks, David!), then by all means let them. Chris Cleve, who would become my main co-conspirator once the sequence got officially "turned on," was already contributing animation improvements long before that point. The more people get involved, the more other people will believe in it and want to contribute, creating a slowly expanding circle of engagement.

Ultimately, these people are your best tool in deciding when to inflict "judgment day" upon your efforts - because at some point, you have to put it in front of the people who have the power to decide if it should become part of the main game, or live forever in experiment exile. If you can't sell the people on your own team and your immediate leads on what you currently have, you shouldn't try to go over their heads - that'll end poorly for everyone. Only when the intro sequence had achieved the enthusiastic endorsement of the team did we even approach the studio heads. It was a make or break moment, and I will always remember Jesse Schell's reaction the first time he saw it:

"You bastards," he swore. "Oh, you magnificent bastards."

Honestly, this approach was risky in its own way, but I can't recommend any other way to get a side-project approved. You only get one chance to get that snap judgment, and that moment of joy was just too rewarding.

Use Meterstones, not Milestones (And not just because the metric system is more sensible.)

For the final stage of the project, we had bi-weekly reviews with the heads of the studio. We brought in Jeff Hoffman to do treatments of a few sections that hadn't quite been nailed down yet, and created videos so we could review individual sequences quickly and easy. Every meeting we were given a list of tasks, and we triaged them right then and got to work. Chris Cleve rightly calls this out as "one of the best experiences I've had." These microtasks were smaller than the usual milestone, could be accomplished within hours, and gave the intro sequence its final shine.

The funny thing is, in retrospect I realize such short work cycles were a key part of the whole process. In order to keep momentum going, I would often focus on one specific goal each night I worked instead of jumping between tasks. These goals would start broad, and vague, and then become more specific over time -  "Today, I'm going to figure out a visual style" became "Today, I'm going to figure out how to pull off a smooth transition," which would in turn become "Today, I'm going to make a shark" or "What's the perfect timing for the bomb to go off." This short-term focus kept scope from running away on a project that could be mistaken as unconstrained, keeping me on a path that was actually accomplishable.

As a bonus, when you give yourself tiny, concrete action items, feedback gets incorporated quickly instead of being put off. People can see that you care about and are responding to their thoughts. In addition to keeping you focused, it will also help get those other people to the point where they feel involved and want the project to succeed on that inevitable judgment day.

Conclusion

So there we go - all the things that allowed the intro sequence to take off. Admittedly, a lot of these tricks were made extra effective by the abstract nature of the intro sequence - but that's kind of what made it such a good fit for bootstrapping. If you got through step one, and have honestly assessed whether or not you can accomplish much of what you want with the resources on hand, there is a good chance you will also be fine.

For posterity's sake, here were the key takeaways as I see 'em:

One final note: When you embark on a endeavor like this, you are donating your free time to your project. I don't think I need to spend too much time talking about how that could get complicated, for you and your studio. Schell Games was a very supportive environment, but you will need to assess your company culture for yourself to determine whether an undertaking like this would even be appreciated. But if you think your company might be amenable to this type of rogue project, it can be an incredibly rewarding experience.

Good luck! And please - try not to die.

Schell Games is a team of over one hundred people who strive to make the world’s greatest educational games, including Yale Medical's PlayForward: Elm City Stories, Water Bears VR, SuperChem VR, the Daniel Tiger’s Neighborhood games, and Happy Atoms. Schell Games also creates pure entertainment content, such as the award-winning VR game, I Expect You To Die, and the comedy space game Orion Trail.

Connor Fallon is a game designer at ArenaNet who spent three good years at Schell Games. In addition to figuring out the future of Tyria, he's working on the indie narrative game Elsinore: A Time Looping Shakespearean Tragedy.

Both I Expect You To Die and Elsinore will part of Indie Megabooth at PAX East in Boston on March 10 through 12, so swing by and say hi if you are attending!