The Making Of…
Roll7 reveals the tricks and tumbles behind the creation of slick 2D skating game OlliOlli
John Ribbins had a lot on his mind in 2012. Rolling Sound, a media training company he’d co-managed, had just been shut down after funding cuts by the Conservative-led coalition government. All 100 employees there had lost their jobs and the stress of it had driven Simon Bennett, Ribbins’ friend and Rolling Sound’s founder, out of the country in an attempt to recuperate.
On top of that, the pair’s still-ongoing venture, a game studio based in London’s New Cross Gate, was in trouble. Roll7 had started in 2008 after Dead Ends, a game about knife crime overseen by Ribbins at Rolling Sound, had proved a success. Now, four years later, it was surviving on freelance programming work and a few educational projects for universities. Roll7 was staying afloat, but it wasn’t the developer Ribbins, Bennett or final co-founder Thomas Hegarty had dreamed of. They’d put out one other game, iOS platformer Gets To The Exit, but it was hardly the breakout hit Roll7 needed. With money running out, Ribbins showed up at Develop in Brighton to try to drum up support.
“The idea was to turn up with an iPad and Gets To The Exit, and try to get it in front of as many people from the press as possible,” Ribbins says. “But it wasn’t very successful. We didn’t even have a stand.
“By this point, as well as Gets To The Exit, we’d done a lot of other projects for marketing agencies and places like that. We’d learned a lot about what worked and what didn’t, and we knew we couldn’t just make something and hope that it would pay for itself. We were trying to save up a buffer. But the gun-for-hire stuff was really starting to suck.”
Understandably stressed, Ribbins hit the bar on the closing night of Develop. There he bumped into James Marsden, an independent developer who’d recently launched Coconut
Dodge for PS3. After sharing a few beers, Ribbins took out his iPhone. He had a prototype skating game that he wanted to show off.
“At the time, it was a super-basic three-pixel game I’d made for iOS,” Ribbins says. “James looked like a skater, so I thought he could try it out. After playing for a while, he told me I should show it to Sony. We were both drunk by now, so I just nodded along – like, ‘Yeah, sure, absolutely,’ not expecting anything to come of it. But he emailed me the next day and introduced me to Shahid [Ahmad, business development manager at Sony].
“We set up a meeting and showed Shahid what we were working on. I got a kick under the table from Tom to show this skating game, which back then was called OlliOlliOlli, and Shahid played it for about 20 minutes straight without even talking to us. He put the iPad down and
“IT’S DIFFICULT WHEN SOMEONE LOOKS AT SOMETHING THAT’S TAKEN YOU HOURS TO CODE AND JUST TELLS YOU IT’S SHIT”
asked if we’d want to bring it to Vita.”
By now, Bennett had returned to the UK and was ready to get back to work. With investment from Sony secured, he decided Roll7 would drop its other development projects to focus exclusively on what was now dubbed OlliOlli.
“When I came back, the company was still working on some other projects using the capital we’d saved up,” Bennett says. “I told them that making a game for Sony was the biggest opportunity we were ever going to have, so we shut down all our other efforts and put everything into OlliOlli. At least that way, if it went wrong, we could say we gave it our all.”
Sony, however, wanted to remain hands-off during OlliOlli’s development. Ahmad had previously brought Thomas Was Alone, Hotline
Miami and Lone Survivor to PlayStation, and was spearheading a push for creative, new and, crucially, independently made titles.
“We had this one feature we wanted to get rid of, some music-linked game-editing mode,” Ribbins says. “So we went to Sony and I spent about half an hour talking, nonstop, trying to justify binning this thing. When I’d finished, Shahid just said, ‘I’ll be honest, you had me at the first sentence. I just wanted to see how long you’d talk for.’ Sony always insisted that this was our game, and we could do what we liked.”
Sony left Roll7 to its own devices, and with a lot of decisions to make. “When we realised we’d been given so much trust and control, it was kind of a wow moment,” says Bennett. “However, that quickly turned into ‘Oh, shit,’ because now there was no one guiding us.”
Roll7 had become used to building games based on others’ specifications. Now, it was shouldering the responsibility of making a title of its own imagining, to be sold on one of the biggest labels in the industry. The initial months were especially pressured for Bennett.
“Every milestone meeting we turned up to, we knew there were some requirements for that milestone that hadn’t been met,” he says. “Plus, we’d decided to build our own engine, which meant that, for a long time, unless you know a lot about coding, there wasn’t a lot to see. There’d be big deadlines coming up and we wouldn’t have anything to show. It made me think we weren’t ever going to get it finished.”
Sony might have been hands-off but it did offer support. Roll7 had access to its testing department and picked up a lot of feedback, particularly about OlliOlli’s difficulty. “Sony’s tester couldn’t play our original version,” Bennett says. “There was one meeting with Shahid where we were discussing the game while this guy just sat in the corner saying, ‘Fuck! Fuck it!’ over and over. The game definitely would have been harder – too hard – without Sony’s input.”
OlliOlli also had to pass Sony’s technical requirements checklist, a catalogue of rules and standards that had to be met before the game could launch. “It was so intimidating,” Bennett says. “It was my responsibility to make sure we passed it, but I knew all these stories about even big studios that hadn’t [made it] through. I felt way out of my depth. Fortunately, Spencer [Low, producer at Sony] was very forgiving.”
Still, development on OlliOlli was progressing slowly. Working alongside senior programmer Nikos Asfis, Ribbins was struggling to get the engine to render on Vita. He’d
designed the background tiles for the first level, but transferring them to the screen was difficult. Roll7 had committed to OlliOlli at the beginning of February, 2013. By the end of March, the studio had rendered just one part of the game: the top-left-hand corner of the first level’s background.
“I remember the point that that happened,” Ribbins says. “I was in the office with Niko at about 1am, and we were singing and dancing because we’d finally rendered this grey square. The next morning Simon and Tom came in and understandably weren’t quite as ecstatic as we were. We’d got something to render, but this was about two weeks before we had to go and show Sony a working game.”
Roll7 had a lot to do. Bennett was cutting deals with the musicians who’d provide the soundtrack while Ribbins worked on art and design and Asfis continued to program. That core team, plus Hegarty, was extended by freelancers, who’d come and go whenever OlliOlli needed artwork or improvements to its interface. It was gruelling, which began to take its toll on Ribbins’ and Bennett’s relationship. “It felt like one long argument between John and I,” Bennett says.”I had a breakdown about halfway through, because I never thought we’d finish the game.
“It’s difficult when someone comes, looks at something that’s taken you eight hours to code, and just tells you it’s shit,” Ribbins says. “What caused the most arguments was working out
OlliOlli’s metagame: deciding how people would progress through levels and how we’d ensure they knew enough about a mechanic before moving on. Originally we had this idea that players would begin with no tricks except an ollie and would learn more as they progressed. But to test the game, we had to throw all the tricks in to see how they worked, and we had so much fun playing it that way that it seemed wrong to force people to go for ten hours before they’d unlock [them]. So we had to have a rethink. Things like this were always a bone of contention.” Come summer, Ribbins was hard at work on
OlliOlli’s animation. A skater himself, he could draw the simpler tricks, like kickflips and shuvits, from memory. But the rest required some research.
“I found an amazing YouTube channel where this guy had filmed every skateboard trick in 10,000-frames-per-second slow motion,” Ribbins says.” They’d even filmed it from this side-on angle, so it was basically in the view that we used for OlliOlli. A lot of work was essentially rotoscoping that down to our character. I’d start by mapping every trick animation onto a pixel stickman and then I’d essentially trace over that in the more detailed art style.”
It was a long process, because Ribbins’ enthusiasm for the subject meant he wanted to ensure the details were spot on. “It was important to me was that if you did a 180 spin and landed it, you’d then be riding backwards,” he says. “That was something we hadn’t seen before in other 2D skateboarding games, though we quickly realised why. If you want to have your skater facing the other way, you have to go back and redraw all his animations so his back is to the screen. We ended up having to make something like 7,000 frames of animation. But as a skater, I wanted a 360 flip to look like a 360 flip.”
While Ribbins finalised the animation, Bennett hit the campaign trail. He flew to LA for E3 and demoed the game to whomever he could. Some got it, others didn’t, and Bennett was still feeling shaky. “I wasn’t in the best place,” he says. “We were at the Sony stand, with just this one Vita, basically showing the game to anyone who came past. I think we met with about five journalists and although some picked it up, others just didn’t take to it. So when I got back to London we had this big meeting and tore the game apart, spent about three months dismantling it and testing it over and over. Even so, we’d kind of settled on the idea that it was going to get slated and not do well. We thought it would be the end of the company.”
After a few finishing touches from Ribbins, like a little green bar under the board that would indicate a perfect grind, OlliOlli was ready to go. It launched in January, 2014. Bennett couldn’t believe the reception.
“For about two weeks I’d get up and open Twitter and it was just like a dream,” he says. “I couldn’t understand how this thing that people hadn’t really taken to just a few months before was now getting all this love. It had been the worst nine months of my working career. By the end, my friendship was John was pretty much in tatters. It was only because OlliOlli had become such a huge success that we were able to sort of hug each other and cry.”
Roll7 threw a party and invited everyone along who’d worked on the game. John Ribbins once again found himself at a bar, thinking about OlliOlli. But he’d come a long way since Develop in 2012, and so had Roll7. After a few hours of drinking and playing Nidhogg, which had also released that month, Tom Hegarty checked his phone. “He looked up and shouted, ‘We’ve won Gamespot game of the month!” Bennett explains. “It was the inaugural award, and we’d won. It felt like fate, because it just so happened we were all together right at that moment. We ordered some tequila. There were tears. It was overwhelming.
“People who don’t know Roll7 might think we just came out from nowhere. But we’ve paid our dues. We’ve worked our balls off making games and doing jobs for other people, and it’s been a very long road.”
Originally, the idea was the trick list in OlliOlli would be unlocked gradually, and players started with a basic jump