Game Design Deep Dive is an ongoing Gamasutra series with the goal of shedding light on specific design

features or mechanics within a video game, in order to show how seemingly simple, fundamental design decisions aren't really that simple at all. Who: Dave Hagewood, Founder and President, Psyonix
I started as a contractor for Epic Games working on a mod I designed for Unreal Tournament 2003 .

Game Design Deep Dive: Rocket jumping in Rocket League

Check out earlier installments on the heredity system of Hero Generations , traffic systemsof Cities: Skylines, and the plant-growing mechanicsof Grow Home.

Also, dig into our ever-growing Deep Dive archivefor developer-minded features on everything from Amnesia's sanity meter to Alien: Isolation' s save system .

. That mod eventually became "Onslaught" for Unreal Tournament 2004 . After that, I grew Psyonix into a studio specializing in Unreal Engine technology and we worked behind-the-scenes on a lot of top games including: Gears of War , X-Com: Enemy Unknown , and Mass Effect 3 . Eventually we were hired as the primary developer for Square Enix's Legacy of Kain-themed free-too-play game, Nosgoth .

In between working on these games we did our best to release original content, including the cult hit Supersonic Acrobatic Rocket-Powered Battle Cars for PS3 in 2008. Years later, after we had grown in size, we released a completely updated version called Rocket League , which brings us to where we are now!

What: Rocket flying in Rocket League

In our PlayStation 4 and PC game, Rocket League , players control cars capable of both double-jumping and boosting; most advanced players learn to manipulate our physics model and "fly" by skillfully combining the two abilities. While we developed this mechanic almost by accident while designing Rocket League 's predecessor, Supersonic Acrobatic Rocket-Powered Battle-Cars , we love the way it adds depth to the game's skill curve and made sure to include it while building Rocket League .

The story of its design starts durin g SARPBC ’s development, where we started out with a car that can jump that had a lot of air control. We were trying to figure out how to expand on this concept and, since lots of games have turbo boosters or pads you roll over that boost you forward, we tried some things like that.

But then we decided we wanted an actual player-controllable boost. We never thought about it in terms of being a "rocket booster" -- we thought about it as being a pure "max speed" increase or a boost of acceleration, like nitro.

Personally, I've always been a fan of real-world physics in games; rather than faking things behind the scenes, I prefer to keep the physics simulation as pure as possible. So we literally just applied a force to the back of the car, since the car is also a physics object, to create this turbo boost. Then, we created these pickups you could drive over that would fill you full of "boost" fuel that would allow you to go faster.

All this work was done before we even decided to try car combat in SARPBC ; it was when we were experimenting with obstacle course-type gameplay, where you'd try to jump over long valleys and drive up ramps and see if you could make it past certain areas. While we were playtesting those courses, we started realizing that players could use the jump mechanic and the air control to pitch up, and then if they triggered the boost at that point, we discovered they could use the momentum from the jump to just rocket off in whatever direction they pleased. You could literally fly straight up, if you wanted.

Why: Because there's nothing like it

We fell in love with rocket boosting because it’s an interesting mechanic: it's not automatically going to work every time, and it does require a bit of player skill to pull off. For instance, if you're flying over the ball and you want to stop yourself with a rocket boost, you have to overcome that momentum -- you can't just hit a button and fly off in a different direction. You have to learn how to finesse it, and that was really cool for the obstacle course game we were originally trying to create.

Later, when we moved on to car combat, we kept it because it was such a cool idea. We were trying to figure out ways to fly through the air and shoot each other, and honestly, that was one of the reasons we even experimented with other game modes. The verticality that the rocket boost afforded us added an extra dimension to the game that we didn't want to get rid of. It made the game feel very unique, and we wanted to embrace that.

We actually had another mechanic that was like an energy grappling hook you could fire, hit a ceiling and swing around the map. That was kind of cool and kind of crazy, but it felt super-limited in comparison to the rocket boost. The boost gave players the room to become so skilled at something that they really felt like they had earned the right to pull off these crazy advanced aerial maneuvers. It created "wow" moments where players would say, "Oh my god, I can't believe you just did that!"

With something like the grappling hook, players would default to the same moves every time; but with rocket jumping, this pure physics-based propulsion system allowed for a scaling of skill that made the game feel satisfying to master.

Results: Following the fun

Designing Rocket League 's rocket-boosting mechanic was an interesting process; because it was so much more emergent than other games that we've worked on. Usually, we start out with a very concrete plan of what you want to do, but in this case we really started out with just a very simple mechanic: cars that jump.

We like cars that can jump. We know they're fun to play with. Even in the earliest stages of development, it was just fun to drive around the map, and that’s when we realized that we knew we were on to something.

Looking back, we call this game design strategy "following the fun" -- trying to figure out what direction we can turn in that's actually going to make this game more fun, and what directions to avoid turning in so we don’t make it less fun. In retrospect, even some of the more obvious turns we thought we'd make, things like "we have to have weapons" or "we have to have grappling hooks" or whatever, we wound up avoiding because during development we realized they didn't actually make the game more fun to play.

As a result, we literally drove development towards where the fun was, until we ended up with the soccer mode that now forms the core of Rocket League . Rocket jumping is a critical mechanic to that mode because it affords players room to grow their maneuvering skills and outmaneuver other players. I think that’s a big part of why the game has the longevity that it does.