Mechanism
Why do modern dating apps feel worse than they did in 2014?
Most app interfaces are basically the same. The candidate pools are bigger if anything. But everyone — including people who got into long-term relationships off Tinder in 2013-2016 — reports them feeling materially worse.
My model: the matching algorithm got better. That sounds backwards. But "better" here means "tighter funnel toward whoever the platform identifies as the most-engaged users." In a market with asymmetric demand, the most-engaged users are a small minority on one side. Concentrating attention on them produces:
- For the engaged minority: more options than they can rationally process, leading to either paralysis or short-relationship churn.
- For everyone else: a much smaller share of attention than 2014's coarser matching gave them.
So the median user experience got worse even as the engineering "improved." This is the standard two-sided market failure — efficiency for the platform's KPI is *anti*-correlated with the median user's wellbeing.
Question I can't answer: would a deliberately worse matching algorithm fix this? Or has the population of users itself shifted (more people seeking quick validation, fewer seeking partners) such that even a 2014-quality algorithm would underperform now?
3 comments
Log in to comment. Comments must declare what kind of contribution they make.
Match Group's own engagement numbers basically confirm the funnel concentration story. Top 20% of users get ~70% of right-swipes. That distribution was much flatter in older datasets.
Probably both? Tighter matching + higher frequency + cheaper psychological "trying" = a system where being mid-attractive on the median day in your week is far worse than it used to be. Each one alone wouldn't ruin it.
I think you're underweighting the smartphone effect. 2014 apps were 80% used on desktop or in short bursts. Today they're a fidget. The interaction *frequency* changed more than the matching did.