Conventional wisdom may dictate that indie game devs should put all their PC games on Steam, but Jason Rohrer's never been one to simply follow the crowd.
The veteran indie is known for doing things differently, whether he's making a game about burglary and home defense in which the perpetrators are often simultaneously the victims (The Castle Doctrine), exploring life and morality in a five-minute journey through a side-scrolling 2D maze (Passage), or burying a game in the desert in the hopes that a distant future generation might discover and play it (A Game for Someone).
And now he's selling and distributing his latest work, the civilization-building (and parenting) MMO game One Hour, One Life, entirely independently — without Steam, GOG, Humble, Itch, or any other digital distribution platforms. I spoke to Rohrer recently to find out why, and how, he's done it.
"For me and what I'm doing," says Rohrer, "it's not just about the way it feels. Like, 'Oh, I want to be indie.' And it's not just about the amount of exposure I would get, or the 30 percent cut [that Valve takes on Steam sales] or any of this other kind of stuff."
It's about doing what's right for the game. "The type of game I'm making is a game where you're not paying for a digital download," he explains, "you're paying for an account on the server that I'm running. Because it's a multiplayer-only game, and it's actually in the public domain — the source code's all available."
"The Steam store doesn't support weird little dynamic pieces of content...when you come to this website, it feels like this cohesive portal for this very particular experience."
That "paying for a server account" business model takes care of one of the biggest points in favor of Steam right off the bat — piracy isn't a concern because the value of his game, from a player's standpoint, rests on experiencing it on the official server with the thousands of other players trying to build and rebuild civilization.
On Steam, everything he'd need is there, but he'd have to spend weeks integrating his code with the latest Steam APIs and then accede to the horrors of technical support when player accounts are generated from Steam IDs — 64-bit strings that are hard to look up manually.
Steering clear of Steam is at its core a matter of expediency for Rohrer. "By doing it off Steam and off any of these other services," he explains, "I have full control over all of that kind of stuff and can make a very sort of customized experience in terms of how the accounts work, how the website and everything else is sort of integrated with your account. How the forums work in connection with that. and, you know, it's kinda like these other things are sort of a one size fits all."
One Hour, One Life's website presents perhaps the clearest case for why selling on Steam might not only be unnecessary but also counterproductive — especially for a regularly-updated multiplayer game such as this one, which has new content patched in every week.
Besides the usual fare — trailer, description, price, links to buy — the website also includes all sorts of extra things. There are tallies of lives lived within his game, lives lived past age 55 (i.e., close to the full hour-long natural lifespan of a character in the game, if they're not killed or left to starve), and monuments currently under construction, plus the length of the longest family tree.
There's also statistics on the size of the codebase and the amount of content available in the game, plus a wiki and a tech tree and a browser for family trees that drills down through the generations — complete with fun trivia like each character's cause of death, age, and final words.
"The Steam store doesn't support weird little dynamic pieces of content like that," notes Rohrer. "So when you come to this website, it feels like this cohesive portal for this very particular experience."
And it's constantly evolving. Whereas a Steam store page is largely static, other than with user reviews and Steam Workshop activity at the bottom, One Hour, One Life's website grows and changes organically in conjunction with the game. Its family tree browser came out of the realization that people were interested in the family relationships and lineages of their characters. The fan art section only appeared a month or so ago, again in response to community activity.
More impressively, given the game's one-man development team and website-only distribution, there's also a reviews section with more than 1,700 player-authored reviews.
It used to be that Valve handpicked every Steam release and helped developers through the process, with guaranteed launch week exposure that translated into significant extra revenue for indies. But that exposure is no longer a sure thing, and Valve has moved to a hands-off approach. Dozens of games come out on Steam every day — by Rohrer's count it was 83 the day One Hour, One Life went up for sale on his website. Even great games can get lost in the shuffle.
And there's plenty of precedents for finding success outside of the Steam ecosystem — Minecraft foremost among them, but also several of the top commercial games (Fortnite, Overwatch, League of Legends, etc), while Prison Architect's paid alpha was a hit before it shifted to Steam.
If it's not marketing and sales, then, he wondered, "What do I lose if I'm not on Steam?"
"Steam has a pretty comprehensive system there for filtering reviews," notes Rohrer. It has options for filtering reviews by date, playtime, helpfulness, negativity, positivity, and more, and these reviews can make a big difference to the sales rankings of a game. "So that feels like the last thing that makes Steam valuable," says Rohrer. "That's where the last bit of virality would come from."
His game could never benefit from that ecosystem without going onto Steam, but he reasoned that at least he could provide people with "an unfiltered outlet to express themselves about the game" and to issue recommendations and/or warnings to prospective buyers.
So he built his own reviews server and solicited reviews from registered players, with their total playtime and the date of the review counted along with their positive or negative rating and review text. Then he put them on the game's homepage, initially in a list of the 10 most recent reviews, with a link to see more, but now split into proportionally-sized positive and negative sections (when we spoke, the overall rating was 93 percent positive, so there were the most nine recent and nine longest-playtime positive reviews followed by two of each type of negative review).
This proportionality system was a response to the fickleness of the crowd. Whenever Rohrer changed or added something that the fans didn't like, he was finding that they would write negative reviews en masse to punish him and try to force him to roll it back. "It's like you have this angry, fickle God that you're trying to appease," says Rohrer. "And I don't think that leads to good game designs." Hence the current system, which better reflects the consensus.
Scrolling through the One Hour, One Life website feels almost like falling down a rabbit hole. There's just so much stuff there to discover about the game and its community, and Rohrer is well aware of the potential this has to overwhelm and scare off potential new players ("it's kind of mind-bogging how many different little weird features are on this homepage," he comments at one point while looking at his own site).
"But at the same time," he says, "a Steam store page has quite a bit of scrolling, right? It's got the description, then it often has some quotes and then it has maybe an extra video someplace and then it also has all these system requirements, so I was kind of thinking that if someone's really about to plunk down $20, more scrolling is not bad."
It's about helping people to make an informed decision about how they spend their money. "They want to get a really good feeling for like the ecosystem of the game, how active it is, how much stuff is happening. And so seeing fan art [and 3,530 people online on Discord], you know, is another sign that the game's community is lively right now. That it's a good community to be joining."
Rohrer says that rolling all of these systems on his own took relatively little effort, at least when compared to the scale and scope of the game itself (which runs thanks to some 84,000-plus lines of code). The reviews system was "maybe" half a week of work, for instance, while the Web interfaces for exploring family trees and interesting statistics took a similar amount of work.
He takes advantage of the rapidity of web development, with mySQL databases, PHP scripts, and basic HTML allowing him to work much faster and with fewer errors than if he'd tried to code these into his game engine.
It's that expedience coming through again. He's just one person, so he picks his battles carefully and takes any reasonable shortcuts he can to improve the experience both in and around his game.
That said, Rohrer is quick to state that staying off Steam isn't going to solve any marketing and sales problems. Indie game development is hard. Most indie games — on or off Steam — fail to earn enough for their developers to even hit the poverty line on sales revenue alone. The challenge is not in distribution; it's making a great game, then getting people to play it.
Rohrer's riding on the cumulative learnings of 18 previous games across 15 years as an indie dev, and a reputation as a leading indie game designer, along with a mailing list of 22,000 email addresses, and still his long-term sales plan is reliant on word-of-mouth virality from streamers and (especially) from players recommending One Hour, One Life to friends.
"People are coming to Steam to buy the game because they happen to have Steam accounts, but they heard about the game some other way," he argues. If he's right, that means that the real questions are "how do I draw attention to my game, wherever it's sold? How do I generate buzz? And how do I maintain sales momentum?" For Rohrer, the key is weekly updates that keep fans engaged, but, as with everything marketing and sales related, your mileage may vary.