Oddworld creator Lorne Lanning.
"Back in 1998 there wasn't YouTube, there wasn't Twitch, there weren't even internet news sites for games yet." (There were, but the fact that a 13-year-old me wrote for one should tell you all you need to know about their quality). "When we started Oddworld most people didn't know what 'www' meant," Lanning says.
This necessitated a lot of guesswork during development. The only opportunities for feedback were market research and focus groups. And even when a game was released, developers only had review scores to go on. "We had no idea who was buying our game," he tells me. "Stores weren't taking that kind of data. You'd find out if you were selling, but that's it."
The internet has changed that, but slowly. Stranger's Wrath was released in January 2005, just before YouTube launched, half a year before Steam began selling third-party games and eighteen months or so before Facebook would pivot from a university network to a public site. The game was a hit with critics, but didn't repeat the sales success of previous Oddworld titles. When it came to New 'n' Tasty, which was published by Lanning's company Oddworld Inhabitants, but developed by UK studio Just Add Water with Lanning as co-director, a different approach was needed.
"I've learned the hard way that you need to listen."
"I used to feel like an auteur," Lanning says. "I still do in a way, but I've learned the hard way that you need to listen." During the development of New 'n' Tasty, he put out a trailer. "Immediately [we saw] how many YouTubers cover it, and what their audiences say." He ran polls through Facebook and Twitter to ask the audience questions. "By that night, we'd have 10,000 results, maybe the next day 20,000. No marketing department could ever have gotten that kind of data when we were making the first Abe."
Lanning also made use of Steam, and the vast amounts of data that brings. "If you buy my game I can see you've bought three games from me before. I can see how many hours you've played them, and what games people have in common with mine." That data has completely changed his understanding of the Oddworld audience. "We used to think of all of our players as one person. Now we know who they all are as individuals. Now, I can go to YouTube and I can see how different personalities react." He explains that Pewdiepie's reaction might be totally different to Jacksepticeye's. "It's basically customer feedback. And with Twitch, you can watch the chat and see reactions in real time."
Character art for Soulstorm, the followup to New 'n' Tasty.
Rather than just creating a game with a singular vision, Lanning made use of the vast feedback loop that is the internet to reshape the final article. The main area he points to is difficulty: "With our first games, we found out they were too hard for people," he says. That led to lost sales, as gamers weren't recommending them to friends. New 'n' Tasty is still challenging, but it has a much kinder learning curve than past games in the series.
Listening to fans -– and those new to the series -- worked: Oddworld Inhabitants sold more than 3 million copies of New 'n' Tasty, a very healthy figure for a game created by studio with 16 employees.
When Abe's Oddysee was released, you'd sell a game, and that was that. Games had to be perfect, or your bug would exist forever. Few households had internet access, and it wasn't until 2002, and Unreal Championship on the Xbox, that console games could even be patched. Now, it's almost unheard of for a game not to be patched in its first month of release. That's partly down to the increased complexity of modern games. "Things were simpler 10 years ago. We want more emergent behaviors and possibilities but what comes with that is more things you can't predict," Lanning says. "As a result, there are more ways that people can screw up again."
Lanning says his goal is, of course, to make a highly polished game, but once you have a million people playing it, "you have that 'oh shit' moment when 50,000 people hit a bug that no one saw in testing."
"It's like, what the fuck are you doing? I know you didn't test this software properly and now you're giving it to me."
It's not just complexity, though. Elements of Silicon Valley's "ship it and fix it" business model have taken hold elsewhere. "I hate that. I hate seeing Google do it, or Apple do it. I feel Apple doing shittier and shittier releases ever since Steve Jobs died," he says. "It's like, what the fuck are you doing? I know you didn't test this software properly and now you're giving it to me." He doesn't claim that game developers have this exact mentality, but believes the knowledge that things can be fixed after release "pushes a mentality that while you should get it perfect, it's more important to meet a deadline than make a perfect game, and fix it later." The middle ground, it seems, is to own your mistakes. "What we have to do is put our hands up and say, 'We fucked up, we missed it, we see you complaining about it,' and get a patch out as soon as possible."
Lanning ends our chat with a cautionary tale about what happens when you don't listen to your audience. "A good example of that is Evolve [2K Games' 4-vs-1 first-person shooter]. It was going to be a huge hit. Everyone that I talked to -- publishers, developers, people that have been doing this for 20 years -- they all believed that. But it came out, and it fizzled -- why?" Evolve received decent reviews, and sales started fairly strong, but within months the average player count on Steam was below 500 (the lowest title in Steam's top 100 has around 2,500 at the time of writing, for reference).
Lanning blames this on its strict requirement for four-player teams, saying people didn't find people they wanted to play with. "All of a sudden it wasn't like it was at the conference, where people went back again and again to play it," he says. "Had they done more monitoring and testing, had they listened to their audience sooner, they would've worked out that they had a problem.
"Evolve wasn't a big success. But it should've been. And if they had more audience feedback before it came out, it would've been," Lanning says. "That's what I think, and I know guys that were programmers and lead programmers on the game."