Let me tell you about the last genuinely weird game I stumbled onto by accident.
It was in a physical GameStop — one of the surviving ones — somewhere around 2019. The cover art looked like a fever dream, the back of the box description made almost no sense, and I bought it entirely because a stranger in the aisle said "that one's strange but you should try it." It was Disco Elysium, and it remains one of the most singular experiences I've had with a controller in my hands. I would not have found it any other way. The algorithm certainly wasn't going to show it to me.
That story matters in 2026 because the conditions that produced it — a physical shelf, a stranger, a random encounter with something genuinely unexpected — barely exist anymore. And the systems that replaced them are doing something that should concern every player who cares about gaming as a creative medium: they're making the world smaller.
How the Machine Thinks
Steam's recommendation engine, PlayStation's "Because You Played" carousel, Xbox's storefront suggestions, and every mobile game platform's "You Might Also Like" feed all operate on fundamentally similar logic. They analyze your purchase history, play time, wishlist behavior, and the aggregated behavior of users who look like you, then surface titles with the highest probability of conversion. In commercial terms, this makes complete sense. In cultural terms, it's quietly catastrophic.
The algorithm is not trying to expand your horizons. It is not a curator with taste, a friend with a recommendation, or a critic with a point of view. It is a conversion-rate optimization tool dressed in the language of personalization. And because it rewards engagement with the familiar — sequels to games you've played, mechanical clones of genres you've logged hours in, safe iterations on proven formulas — it systematically disadvantages anything that doesn't fit an established pattern.
An experimental narrative game with unusual mechanics and no franchise history doesn't just struggle to stand out in this environment. It is actively deprioritized by systems that have no framework for valuing novelty.
The Feedback Loop Nobody's Talking About
Here's where it gets genuinely troubling: the algorithm doesn't just affect what players find. Over time, it shapes what developers make.
Indie developers — who are most dependent on storefront discoverability because they lack the marketing budgets to bypass it — are acutely aware of what the algorithm rewards. When a game with a familiar genre tag, a recognizable art style, and a clear mechanical hook gets surfaced organically while something genuinely experimental disappears into the long tail, the lesson is not subtle. Make something the algorithm can categorize. Make something with a comp title. Make something safe.
This isn't a conspiracy. Nobody at Valve or Sony is sitting in a room deciding to kill experimental games. It's the emergent consequence of optimizing for engagement at scale. But the result is a creative ecosystem where the algorithmic feedback loop is, year by year, narrowing the range of games that feel commercially viable to make.
Speak to any mid-tier indie developer about their Steam strategy in 2026 and the conversation inevitably circles back to tags, genre positioning, and comparable titles. "You need a comp," one developer told me recently. "If you can't tell the algorithm what your game is like, it doesn't know where to put you, and if it doesn't know where to put you, nobody finds you." The implication — that a game without a good comp might not be worth making — is one the industry has not fully reckoned with.
What Discovery Used to Feel Like
It's worth being honest about the nostalgia trap here. The physical retail era was not a golden age of discovery. GameStop's shelf space was dominated by the same major publishers, the same franchises, the same genres. Small games got buried then too, just in a different way — under a pile of Madden copies and licensed movie tie-ins.
But the physical era had something the algorithm cannot replicate: friction. The experience of picking up a box, reading the back, making a decision based on vibes and cover art and the recommendation of someone who happened to be nearby — that friction created accidental discovery. You couldn't be perfectly targeted. The chaos was the feature.
Human curation has tried to fill the gap. The Wholesome Games showcase, Day of the Devs, Double Fine's Amnesia Fortnight, various itch.io collections, and the work done by indie-focused journalists and streamers all represent genuine attempts to surface the unexpected. These circuits matter enormously — they are, in many cases, the only reason certain games find an audience at all. But they reach a fraction of the players that algorithmic surfacing does, and they require the player to already be looking for something different.
The Comfort Zone Is a Cage
The deeper issue is psychological. Recommendation algorithms don't just reflect our preferences — they reinforce them. Every time a player clicks on a suggested sequel rather than scrolling past it to find something new, the system learns that the suggestion worked. Every time a game outside your usual genre goes unclicked because the algorithm never showed it to you, the system learns that you're not interested in games outside your usual genre. The bubble tightens with every session.
This is not hypothetical. Research into recommendation systems across streaming platforms — most visibly in music and video — has consistently shown that algorithmic curation increases consumption of familiar content over time, even when users report wanting more variety. There's no reason to believe gaming storefronts behave differently.
The result is a player base that is, in aggregate, being trained to want less. Less variety, less weirdness, less risk. And an industry that is, in response, learning to deliver exactly that.
Break the Loop
None of this means the algorithm is irredeemable or that discovery is dead. But reclaiming it requires active effort in a system designed to make passivity the path of least resistance.
Seek out the human curators — the indie showcase circuits, the niche gaming newsletters, the streamers who make a point of playing things nobody's heard of. Use Steam's tag browsing deliberately rather than waiting for suggestions. Follow developers whose work you respect and see what they're playing. Let a stranger in a comments section talk you into something weird.
The algorithm knows what you want to play next. That's exactly why you shouldn't let it decide.
