感谢你，Hades——Evil Empire 谈独立游戏发现、Steam 抢先体验变化与 AI 工具兴起

Ken coolly serves the hottest news stories and has played, we think, every game ever made.

Published: October 17, 2026

Evil Empire began life carved out from Dead Cells developer Motion Twin as a standalone studio that would carry on supporting it while the rest of the team moved onto its next project, Windblown. Fast forward to today, and it's blossomed into a multi-game studio, most recently bringing us Castlevania: Belmont's Curse, while also supporting the indie game scene through its Triple-i Initiative. I caught up with CEO Thomas Pfeiffer and Head of Business Ben Laulan to learn about making that journey, and what advice they'd pass on to other budding indie teams.

Laulan says the team's success with Dead Cells grew from its four co-founders into 15 working full-time on the roguelike, before it began to rapidly scale up to its modern size of 50 permanent contractors to take on projects with bigger publishers. We first got The Rogue Prince of Persia in conjunction with Ubisoft, and just this week Evil Empire has released Castlevania: Belmont's Curse from its partnership with Konami. There's already more in the works, although nothing to be said on that front yet.

"Right now we have multiple lines," Pfeiffer remarks. "You can imagine us working on two or three games in parallel at different stages." The team also has its business unit for live ops, which has seen it supporting other studios' with projects such as arena-shooter roguelike Brotato from Blobfish Games, and its marketing team. As part of this work, it created the Triple-i Initiative, a recurring indie showcase that first took place in 2024 and just held its fourth incarnation in early October, the second installment of this year.

"The last event in April was extremely successful, because we made something like three million wishlists in just a week of events across all participating games," Laulan tells me. That estimated out to an average of around 65,000 wishlists per game, but with quite a few reaching six figures.

The team was still waiting for the dust to settle on its October showcase when I spoke to them. "There are chances it's going to be a bit less," Laulan notes, "But we do believe we managed to still create an impactful event for the participating studios." In part, he says this was affected by Steam visibility - while this latest Triple-i showcase had specific sales on all three major consoles for the first time, Valve's platform has a "one featuring per year per event" that had already been used with the April showing.

Among the heavy hitters this time were the likes of Croteam's The Talos Principle 3, a crossover between Subnautica 2 and Dave the Diver, and the reveal of Aftergods from Tribes of Midgard developer Norsfell. One that Laulan was particularly pleased by the response to was Machinarium 2, from adventure game veterans Amanita Design, which continues the tale of robot partners Josef and Berta almost 20 years on.

"It's a game I grew up with, I loved it," he recalls of the 2009 original, "and I was so happy to get the reveal of a sequel. So I was very eager to see players' feedback to that - is it too old to come back? Will it really make a splash? And it was a huge performer in the showcase, so we are really happy about that."

The count of new games being released on Steam every week is astronomical nowadays, making discovery tougher than ever. Indeed, Laulan remarks this was "the main reason" for the creation of the Triple-i Initiative. "We had to announce The Rogue Prince of Persia, and we couldn't really see a good fit in the landscape of showcases; a fit that really brought us the audience we wanted to target and reach. So we said, 'Hey, why not create our own thing?'"

He believes being spotted today "lies not necessarily in showcases, but in talking to a specific audience - to make sure what your game is, and who your game is for, and trying to find a way to connect to that very specific audience." He says he recommends to any developers interested in being part of Triple-i or other showcases not to look at the total view count, but whether similar games are performing well there. "You're not trying to reach a million views, for instance, you want to make sure to communicate to players that will be there to buy your game."

I ask what the team's thoughts are on Valve's plans to trade out Steam's curated discounts for fully algorithmic suggestions. "I'm not too worried because Valve is really good with algorithms," Laulan comments. "I just hope it won't prevent discoverability and can avoid having your own echo chamber, because I play certain types of games, but once in a while I don't mind discovering a genre I'm not used to." He hopes the new setup ensures "ways to discover new and exciting experiences I wouldn't have otherwise discovered by myself."

Having worked with both Ubisoft and Konami, the pair say they'd absolutely encourage other indie developers to pursue similar collaborations. "It comes with a cost," Pfeiffer cautions, noting that you'll need to be prepared to deal with the structure of a bigger company. "Security stuff, IT stuff… very basic stuff but important to make the contract possible." Then, once you're underway, there are factors such as time zones and language differences to factor in. "But yes, for many studios, I think it would be a good idea."

Both Dead Cells and The Rogue Prince of Persia went through early access, but Castlevania avoided doing so. Does that mean the team's perspective has changed? "It's not an easy answer," Laulan responds, "it's not a done deal like, oh let's go early access - for us it was very important, because it's a way to reach our audience early and we love to build games hand in hand with the players, so getting in touch with them and collecting their feedback very early was super important for us."

"It's been great in the past, but I honestly don't have any clear answer today," he continues. Discussions are ongoing for the other projects Evil Empire is currently working on. "We haven't figured out yet what makes the most sense for us." Laulan says a lot has changed since ten years ago: "The need for content and quality is off the charts now - thank you, Hades." He laughs. "There are many ways to collect feedback nowadays that don't necessarily need early access, and we'll probably be looking into that."

"We need to find another approach to get in contact with players as soon as possible, because that's what we like - to be connected with real players," Pfeiffer remarks. "When I was much younger, when we were creating games on web platforms, we were connected to players through forums. It was easy to do." Echoing Laulan's sentiment, he says nowadays "early access is not an easy approach because expectations have changed. But we still need to get in contact with the community of players as soon as possible."

I can't wrap up without touching on the hottest topic of today - the use of AI-supported coding and generative-AI assets in game development. Pfeiffer says Evil Empire doesn't use it in their production, and on a personal level, he would miss the shared struggle of creation as a team. "If AI makes the effort disappear - if we don't have the effort of creating alignment, creating pleasure, having difficulties and troubles to create that team effort - I'm not sure I'm going to be happy with creating games any more."

"Even if it's faster, even if it's maybe better because knowledge will probably be better at coding with AI than with individual people, the reason why people make games is because it's a fantastic moment," Pfeiffer emphasizes. "We are creating something together from scratch, and we need to create that same understanding of what we are going to achieve; to welcome new ideas. If we are losing that, I don't know what games are going to be. I'm going to be very frustrated for sure. But I know it's coming."

"I think we'll keep crafting games the old way for a bit more time," Laulan chimes in. "A game is something we do," Pfeiffer concludes. "It's a collaboration on a creative subject. If you are collaborating with a machine, whatever you train it for, it's not going to be the same thing." He taps his hand to his heart; I can hear the passion in his words. "Maybe the product is going to be good, but the meaning of the product is very different."

Keep scrolling for next article