Disclosure: The organizers of the Slush game conference paid my way Iceland, where I was the master of ceremonies. Our coverage remains objective.

When I was in the second grade, I told my teacher that I wanted to be a unicorn. And while the other kids may have laughed at me, I may still get the last laugh.

Valve, the company that owns Steam and is developing the Steam VR platform, has built an internal test game for virtual reality that lets you play as a unicorn. In this game, which Valve writer Chet Faliszek simply called “balloons,” it’s your job to use the horn on your head to pop the balloons that are falling from the ceiling.¬†Faliszek didn’t mention any plans that Valve has to release this as a consumer product, but he did highlight how building experiences like this is helping Valve better understand VR.

“Before we had controllers, and we just had visual tracking, we experimented with all kinds of ideas,” Faliszek said in a presentation at the SlushPlay conference in Iceland earlier this week. “We just tried all kinds of things. Balloon was this game where you’re a unicorn with a horn atop your head. We then added a bit of a game where you only pop the pink ones and popping the other ones would take away from your score.¬†It was stupid. It was fun. People played it, and people loved it. We learned really quickly that balloons — the physics of it — just work.”

That discovery provided Valve with valuable data that it is still using today.

At the Game Developers Conference that happened in San Francisco in March, I got to try out the HTC Vive, which is the first consumer product built on the Steam VR platform. The first thing that the person running the demo instructed me to do was to press a button on the left controller. This caused a balloon to immediately start inflating out of the tip of the motion-tracked joystick.

I then, unprompted by the Valve representative, used my other controller to bump the balloon in the air. It was very satisfying, and it felt very real.

But Valve only came to this place because it was messing around with the limited tech it had early on, and that was one of the messages that Faliszek was trying to impart on the developers in attendance at SlushPlay.

“Don’t just go straight at the problem,” he said. “You’re experimenting. Just start doing stupid things because who knows where it will lead.”