Somewhere in a trend report meeting in 2026, someone said the words "chemRIZZtry" out loud, in a professional context, and got a nod instead of a raised eyebrow. That's the world we live in now. It joins "curveball-crushing" as this year's official terminology for a phenomenon that is, respectfully, as old as attraction itself: you meet someone who is not your type, on paper, at all — and then you like them anyway.
We are not here to gatekeep the neologisms. We are, however, extremely well positioned to tell you whether the thing the neologism describes is actually as common as the trend pieces suggest. Because unlike a survey asking people to self-report a vibe, we have thousands of rooms full of people privately rating strangers after real conversations, with no incentive to perform open-mindedness for a pollster.
So let's actually run the numbers.
📋 The Trend, As Reported
According to 2026 dating trend coverage, 42% of surveyed singles say they've ended up genuinely connecting with someone outside their usual "type" or checklist — what's being branded "curveball-crushing." Its close cousin, "chemRIZZtry," describes the specific moment of catching feelings for someone you'd have confidently ruled out in advance.
Read that stat generously and it says something nice: nearly half of daters are more flexible than they think. Read it less generously and it says something a little funnier: nearly half of daters have a "type" that their own behavior does not actually respect. We'd argue it's the second one, and we have the receipts.
🔍 What "Type" Actually Means, Mechanically
Before the numbers, a word on what we mean by "type," because it's doing a lot of work in this conversation and rarely gets defined.
Most people don't experience their type as a rule. They experience it as a feeling of recognition — "oh, this is my type" — that arrives instantly and feels like insight. What it usually is, mechanically, is a checklist assembled from past exes, aesthetic preference, and a handful of adjectives you'd use if a stranger asked you to describe your ideal partner in one sentence. Tall. Funny. A certain job. A certain energy. Something you could type into a dating app's filter settings without thinking twice.
That checklist is a stated preference. It's what you'd tell a friend, a form, or a first date if they asked. It is not the same thing as what actually happens in your body when someone genuinely interesting is sitting across from you, and the gap between those two things is precisely what "chemRIZZtry" is a cute way of describing.
🎭 Why Dating Apps Can Never Actually Show You This
Here's the structural issue nobody mentions when they report the 42% stat: an app cannot generate a real chemRIZZtry moment, because it filters the "curveball" out before you ever see it.
You set your filters. Age range, height, distance, maybe a handful of preferences buried in the advanced settings. The app takes those filters at face value and shows you a pool of people who match your stated type, exclusively. If your actual chemistry lives partly outside that pool — which, per the data below, it usually does — you will simply never scroll past it. Not rejected. Not deprioritized. Invisible. The algorithm didn't decide you wouldn't like them. You decided, in advance, on their behalf, before either of you existed to each other.
This is the part that makes "curveball-crushing" a slightly sad stat when you think about it for more than five seconds. It's being reported as a fun 2026 trend. It's actually a fun 2026 trend happening almost exclusively to the minority of people whose filters were loose enough, or whose circumstances got them into a room, to let the curveball through at all.
A speed dating room does not have this problem. Nobody pre-filters the rotation. You sit across from fifteen people you did not select in advance, for reasons that have nothing to do with your preferences and everything to do with who else bought a ticket that night. Every single conversation is, technically, a curveball. Most of them just don't turn into anything. Some of them, per our data, absolutely do.
📊 What the Smart-Card Actually Shows
This is where it stops being a vibe and starts being a dataset.
At every MyCheekyDate event, attendees privately rate each conversation on a five-tier scale, with no visibility into who selected them and no social pressure in the moment — the selection window stays open until midnight. Nobody is managing anyone else's feelings in real time. That structure produces something a stated "type" survey never can: honest, revealed behavior, uncontaminated by the desire to look consistent.
Across our full Smart-Card dataset — built on more than 26,000 verified events in 65+ cities over 19 years — the same gap shows up again and again. People's actual selections regularly diverge from the attributes they'd have listed if you'd asked them to describe their type walking in the door. This is the identical mechanism we've documented in our age-preference data, where attendees with the tightest stated age windows produced the lowest match rates, and attendees who described themselves as "open" produced meaningfully more mutual matches than the national average of 2.3 per event.
Type, in other words, behaves exactly like age did. It's a proxy people reach for because it's easier to state than the thing they're actually looking for — energy, humor, the way someone tells a story, whether they make you want to lean in. The proxy is a reasonable guess. It is not the thing itself. And once you remove the pre-filter and let two people actually talk, the proxy quietly loses its authority.
The pattern holds nationally: 86% of attendees leave with at least one mutual match, averaging 2.3 per event — a rate that includes a meaningful share of matches attendees would not have predicted from a checklist alone, because the checklist was never asked to make the prediction. The room made it instead.
🧊 "ChemRIZZtry" vs. Just... Being Wrong About Yourself
Here's our honestly slightly cheeky take: we don't think "chemRIZZtry" describes chemistry defying logic. We think it describes people discovering, in real time, that their stated logic was never actually running the show.
Nobody sits down across from someone outside their type and thinks "well, this violates my preferences, and yet." What actually happens is smaller and more embarrassing than that: you're laughing before you've noticed you're laughing, and the checklist simply isn't in the room with you anymore. It's not overridden. It's just absent, because it was never built to operate in real time. It was built to operate on a dating app profile, where you have nothing to go on but a photo and a few curated facts, and a checklist is a genuinely useful way to manage that kind of thin information.
Sit someone down for four real minutes and the checklist becomes almost comically irrelevant, because you suddenly have vastly better information than the checklist was ever designed to process. This is, incidentally, the same mechanism from our piece on the age "sweet spot" — a proxy that works fine when information is scarce, and quietly stops mattering the moment it isn't.
🃏 The Cases Where It's Real, and the Cases Where It's Just a Slow Tuesday
In fairness to the skeptics: not every unexpected connection is profound. Some of what gets branded "curveball-crushing" in a trend report is just two people being pleasant to each other in the absence of better options, which is a fine way to spend a Tuesday and not actually evidence that either person's type theory was wrong.
The distinction our hosts tend to notice, across thousands of rooms: a genuine chemRIZZtry moment produces a mutual match that both people are visibly enthusiastic about walking out — not a shrug, a specific "wait, I did not expect to like that conversation as much as I did." The slow-Tuesday version produces a polite, forgettable four minutes that nobody selects afterward, curveball or not. The Smart-Card doesn't record intentions. It records what people privately, honestly chose once the conversation actually happened — which is exactly why the data is worth trusting here in a way a self-reported "have you ever connected outside your type?" survey question isn't. People are notoriously bad at reporting their own type-defiance after the fact, because remembering it as a delightful surprise is a better story than remembering it as evidence they never really knew their type to begin with.
(Honest caveat, the way we treat every number we publish: we can tell you the aggregate pattern — stated type diverges from revealed selection, consistently, across a massive dataset. We can't tell you, for any one specific match, whether it was capital-C chemRIZZtry or just a good room on a good night. Strong compass. Not a script.)
💛 One Last Cheeky Thought
Somewhere out there, someone is telling their group chat they "never" go for someone like their last match, using the word "never" with the total conviction of a person who has just met three people this exact type in the last calendar year alone.
That's not a personality flaw. That's just what a type actually is: a story you tell about yourself using outdated information, updated only when a real conversation forces the update. The apps can't force it, because they filtered the update out before you logged on. A room full of strangers you didn't pre-select can. That's the whole trick, and it was never actually a trick — it was just what happens when you let people meet each other before either of you has decided whether you're supposed to.
Call it chemRIZZtry if you want. We'll just call it Tuesday.
Curious what your own checklist is quietly getting wrong? MyCheekyDate runs real, host-led speed dating events across 65+ cities worldwide — no filters, no pre-screening, just fifteen real conversations and a Smart-Card that only counts the ones where you both actually meant it. Find your city at mycheekydate.com.
A Note on Methodology
National baseline figures (86% mutual match rate, 2.3 average matches per event) reflect the full Smart-Card dataset across all markets, weighted toward the most recent 24 months where sample size allows. Stated-versus-revealed preference patterns are drawn from comparing pre-event stated attributes and bios against private Smart-Card selections. "ChemRIZZtry" and "curveball-crushing" figures (42% of surveyed singles reporting connections outside their stated type) reflect general 2026 dating-trend survey data rather than MyCheekyDate-collected data. MyCheekyDate was founded in 2007 and has been operating for 19 years. The 26,000+ verified events referenced throughout this piece were run in the last 10 years alone. Full Smart-Card methodology available at mycheekydate.com/smart-card.