In 1956, a Bell Labs researcher named George Miller published a paper arguing that the human mind can hold roughly seven items in working memory at a time — give or take two. Sixty-plus years later, the average dating app user can swipe through hundreds of potential matches in an evening, and social platforms surface thousands of strangers as potential connections. The gap between our cognitive architecture and the options available to us has never been wider. And something strange happens in that gap.
Barry Schwartz's Jar of Jam
The foundational experiment in choice overload research involved jam. Psychologists Sheena Iyengar and Mark Lepper set up two tasting tables in a grocery store: one with 6 varieties of jam, one with 24. The table with 24 attracted more browsers. But the table with 6 sold more jam — ten times more, in fact. More options, paradoxically, produced less action and less satisfaction.
Barry Schwartz synthesized findings like this into what he called the Paradox of Choice: past a certain threshold, more options don't liberate us, they paralyze us. And even when we do choose, we're more likely to feel regret, because the unchosen alternatives remain vividly imaginable. You didn't just pick Option A — you also declined Options B through W.
Most of the original research was about consumer goods. But the same dynamics apply with uncomfortable precision to social discovery online.
What Happens When People Become Options
When you meet someone at a party, on a train, or through a mutual friend, the encounter has a kind of specificity. This person appeared in your life through a particular chain of circumstances. There are maybe five people you could realistically talk to tonight. You invest attention accordingly.
Now imagine opening an app that could, in theory, connect you with any of several hundred million people. The abundance immediately changes the psychology. Each individual encounter becomes mentally cheaper — if this conversation isn't working out, the cost of moving on feels near-zero. Not because you're callous, but because your brain has registered the size of the alternative pool and done the arithmetic.
Research on online dating has documented this extensively. Users who are presented with larger match sets report lower satisfaction with any given match. They're also more likely to discontinue conversations early, because "next" is always right there. Behavioral economists call this the "option value" effect: when you can easily access alternatives, the value you assign to your current choice drops, even when the current choice is objectively fine.
Filters, Algorithms, and the Illusion of Control
The standard platform response to choice overload is filtering. Age, location, interests, mutual connections — add enough parameters and the haystack shrinks to something manageable. This feels like it should help, and in some respects it does. But filtering introduces its own distortions.
When you filter for what you already know you like, you exclude the unexpected. And much of what makes a conversation genuinely memorable — a perspective you'd never considered, a shared reference you didn't expect, the discovery that you find a certain type of person fascinating when you'd have never predicted it — comes precisely from the unexpected. Tight filters optimize for comfort and efficiency. They're lousy at producing surprise.
Algorithmic recommendation has a similar problem. When a platform learns your preferences and serves you more of what you've engaged with before, it's doing something useful. But it's also quietly narrowing your social world, nudging you toward a personalized bubble of people who are likely to feel familiar. Serendipity — the accidental encounter that opens something new — requires a degree of randomness that optimization tends to eliminate.
The Case for Constraints
There's a counterintuitive design principle buried in all this research: artificial constraints can improve experience. The finite — a set list, a prix fixe menu, a conversation that happens because two people were simply placed in the same room — can feel more meaningful precisely because alternatives aren't dangling in the periphery.
This is partly why some people report that random pairing platforms produce conversations that feel different in quality from filtered matching. When there's no optimization layer selecting your partner, the encounter takes on a different character. You didn't choose each other. You ended up here. That small shift in framing changes how both parties engage. There's less evaluation, more exploration.
The psychologist Mihaly Csikszentmihalyi spent decades studying the conditions under which people enter states of deep engagement — what he called flow. One recurring feature was clear constraints: defined rules, a bounded problem, a specific context. Unconstrained freedom, it turns out, is often the enemy of full engagement.
Satisficers, Maximisers, and the Social Search
In the 1950s, economist Herbert Simon introduced a concept that cuts to the heart of how different people experience platforms with large option sets. He distinguished between two decision-making strategies: maximising and satisficing. Maximisers search exhaustively for the objectively best option — they cannot settle until they are confident they have found it. Satisficers, by contrast, adopt a threshold: the first option that meets their minimum criteria is good enough, and they stop searching.
In the context of consumer choice, maximisers tend to perform better on objective measures — they do sometimes find better deals — but they report lower satisfaction and higher regret than satisficers, because the mental cost of the search and the lingering awareness of alternatives erodes the pleasure of the outcome. Satisficers are, counterintuitively, happier with objectively inferior choices because they aren't constantly comparing.
These tendencies map directly onto social platform behavior. Maximiser-type users scroll endlessly, skip rapidly, and apply increasingly fine-grained filters in pursuit of the theoretically ideal conversation partner. They are, in a measurable sense, harder to satisfy — not because the platform is failing them, but because their search strategy guarantees dissatisfaction. Every real person they encounter is implicitly competing with an imagined perfect match still somewhere in the queue.
Satisficers behave very differently. Given a random connection that clears a basic threshold of interest, they engage. They invest in the conversation in front of them rather than mentally auditing it against alternatives. And this difference in engagement posture tends to become self-fulfilling: the satisficer's willingness to lean in produces a better conversation, which produces a better outcome, which reinforces the strategy.
This is one reason constrained and random-matching platforms can produce disproportionately positive outcomes for a significant segment of users. By removing the scroll — by presenting a single person rather than a ranked list — these platforms structurally nudge maximisers toward satisficing behavior. They take away the mechanism the maximiser would otherwise exploit. The result isn't a worse experience; for many users, it's a substantially better one. The constraint does the work that willpower couldn't.
Toward Better Social Discovery
None of this is an argument against choice or platform design. It's an argument for thinking more carefully about what we actually want from social discovery online. If the goal is to find the theoretically optimal person according to specified criteria, maximum filtering makes sense. If the goal is genuine connection — the kind that surprises you, that reveals something about yourself, that results in a conversation you actually remember — then the endless scroll may be exactly the wrong tool.
The people most satisfied with their social lives, research consistently finds, aren't those with the most options. They're those with the deepest investments in a smaller number of connections — people with whom there was, at some point, an encounter that felt unplanned, a little risky, and unexpectedly real.