"Launching games is the hardest job I've done in the games industry," says Curve Digital publishing director Simon Byron. "I've been on the press side and the PR side, but this is the job that keeps me awake the most. It's also the most exciting because you never know how a game's going to do."
Byron has no doubt endured many a sleepless night over the past few months, with Curve having launched GTA love letter American Fugitive, physics puzzler When Ski Lifts Go Wrong and the console version of For The King earlier this year.
The publisher is now gearing up for its all-important Q4 releases, such as Zelda-like indie adventure A Knight's Quest and an adaptation of hit Netflix show Narcos, with a video game based on BBC drama Peaky Blinders also in the works.
If you look at the company's line-up, it's hard to identify a common theme -- and that's with good reason. Curve has avoided leaning towards a particular style of game, with Bryon telling us the selection process is "based on what gets us excited."
It's a strategy that contrasts with that of some of the leading indie publishers. Devolver Digital, for example, seems to specialise in irreverent and often comically violent games, while Annapurna Interactive offers narrative-driven and artistic experiences. Byron says Curve has a slightly different priority.
"We like games that are streamable, that are fun, and where players' experience is generally different to what they can see while watching someone else play," he explains. "When I look back at the games we publish over a calendar year, if you rank them by Metacritic and then by sales, they're often very different. It's certainly the ones that are streamable and much more viewer friendly that seem to be doing well.
He continues: "I struggle with whether we should consider our 'brand identity,' if you want to call it something wanky. We tend to stand very much behind our games. Sure, there's a Curve splash screen on everything we do but are those seven million buyers of Human: Fall Flat even aware it's a Curve game? I'm not so sure.
"I wouldn't want us to ever be hampered by a particular style or genre... I equally wouldn't want us to be defined by the amount we're investing in them. Do £1 million plans guarantee anything? It certainly doesn't, but if you're only doing £1 million it means you can't work with smaller titles that might set your heart aflutter.
"So it's true there isn't a typical Curve game, but hopefully that means we can keep surprising."
"I wouldn't want us to ever be hampered by a particular style or genre... but hopefully that means we can keep surprising"
Human: Fall Flat is one of the company's biggest hits to date, with more than seven million sales (plus however many Game Pass subscribers have given it a go). While other publishers -- indie or AAA -- would use this as a benchmark for future releases, hoping to eventually surpass it, Byron says this has not affected expectations for future titles.
"If you forecast assuming you're going to do what Human: Fall Flat did, you aren't going to be in business very long," he says. "We've got realistic expectations [for our titles], but really, really high hopes.
"What we've seen with Human: Fall Flat is that if you continue to update with new content -- particularly free content -- it really does re-energise the player base and enables fans of it to encourage their mates to buy as well."
The publisher clearly has grander plans for the future, however, as it made its first acquisition towards the end of 2018. Curve parent Catalis purchased Brighton-based Runner Duck Games, the team behind Bomber Crew -- a title that made the publisher $1m in two weeks. And Byron tells us this isn't the last purchase Curve has in mind, although we shouldn't expect a Microsoft-style mass studio grab any time soon.
"I wouldn't say we're aggressively pursuing acquisition targets, but if we come across one where the fit is right, we'd certainly be interested," he says. "With Runner Duck, they have been a dream partner from the moment we first encountered them. Bomber Crew was a real privilege to be involved with from the start and we wanted to make sure we had a longer-term relationship with them so we put a ring on them.
"It's certainly the start of where we want to be. We're not talking to anyone else at the moment, but there is definitely aspiration there to find more partners that we love and put more rings on them."
Today, the publisher has announced its newest game signing: Autonauts, developed by Scottish studio Denki and the brainchild of Gary Penn, creative director behind the original Grand Theft Auto.
Due for release this autumn, Autonauts is a sandbox strategy game where you program worker bots to perform tasks for you: catch fish, chop down trees, process wood, etc. Bots are programmed either by physically showing them how to complete their assigned task or with a Scratch-style drag-and-drop coding language, with the ultimate goal of establishing a civilisation on an uninhabited planet.
In line with Byron's earlier comments, the publishing director believe this is a game that will be sold "via GIFs and streaming", and hopes are particularly high for its release.
"It's a game that you can absolutely see sitting at the top of Steam," says Byron. "It's one of those tinkering games where you start off very small, and how you build and colonise your planet is entirely up to you. And it's so cute.
"Any publisher that tells you they know the secret to success is a liar"
"In terms of the tags that are popular [on Steam] at the moment, I think Autonauts ticks an awful lot of those. It just feels right. We fell in love with it the moment we saw it."
The voxel art style makes it eye-catching enough, but the hope is that videos and livestreams of the game will truly convey how sophisticated a routine and far-reaching a robot empire players can build. It's also hoped Autonauts might serve as a gateway to programming -- not the language, of course, but the fundamental process.
"You won't go from playing Autonauts to suddenly being able to write in Python, but in terms of teaching the core concept, it's absolutely up there," says Byron. "Tasking worker bots to go to nearest X, do this, bring this back -- thinking about every single step that comprises what you want it to do. And with coding, if you've not got it quite right, it just throws up an error message, whereas in Autobots you can see where things are going wrong -- it's much more of a visual feedback loop."
He offers the example of an earlier play session by Penn, where one bot was programmed to collect and stack waste. Leaving the bot to its task, Penn concentrated on other areas of the planet -- by the time he remembered it was still working, the waste was stacked so high it nearly reached the moon.
Anecdotes like this are already abundant across the game's fledgling community. As an Itch.io prototype, Autonauts has enjoyed a healthy number of downloads and Byron reports it's becoming popular on YouTube. There's even a thriving Discord discussing the alpha version.
It's not the first time Curve has found Itch.io to be a "fantastic proving ground for concepts" -- Human: Fall Flat also began life on the site. But translating the success of the earliest version into real sales is something of a challenge.
"Any publisher that tells you they know the secret to success is a liar," says Byron. "There are an awful lot of things that need to go right.
"In terms of numbers, it's not huge but the community's really active and there are several individuals that are spending hundreds of hours still playing it. So we know once it's in the hands of people, they tend to like it. So the job on the marketing side is to convey all that stuff, to get across this playful sandbox nature and the fact that it's open-ended, where you can experiment and see unexpected results.
"Can I guarantee it's a Steam best-seller? No, I absolutely can't. But out of the many games we see, it stands a really strong chance."