I put on a podcast on Spotify called Strangers on a Bench most often on long walks, the kind I take specifically to bring my nervous system back down after a day that ran too hot.
It’s a simple format: the host sits on a bench somewhere and starts talking to whoever happens to be sitting there, and for twenty or thirty minutes you get a stranger’s actual life, unfiltered, in your ears, while you walk somewhere that has nothing to do with either of you. It’s become one of my more reliable ways of calming down. Someone else’s story, briefly, standing in for my own.
A few days ago, one of those strangers said something I haven’t been able to put down since:
Over the course of an entire lifetime, he said, most people probably only have three or four connections that actually count. Not friends in general. Not the wide, pleasant collection of people who know your name. Three or four relationships, total, that genuinely change you.
I don’t know if that number holds up to scrutiny, and I want to say that plainly before I do anything else with it. I couldn’t tell you what study it’s from, if it’s from one at all, or whether the man on the bench was reciting research or just describing his own life and rounding it into a rule. What I can tell you is that I haven’t stopped thinking about it since, which is its own kind of evidence for something, even if I’m not sure what.
The question I expected to ask myself, and the one I actually asked
The obvious anxiety, the one I assumed would show up first, was some version of what if I never find them. What if the three or four are still out there somewhere and I miss them, or run out of time, or I’m simply not paying close enough attention when one of them walks past.
That’s not the question that actually kept me up. The one that did was quieter and considerably harder to sit with: what if some of them have already happened, and I didn’t understand what I was inside of while I was inside it.
Someone I don’t speak to anymore. Someone who was in my life for a matter of months and then simply wasn’t, through nobody’s fault, just the ordinary drift of two people’s lives pointing in different directions. A friendship that reshaped how I saw myself and then ended the way most good things eventually do, without a villain, without a single decisive fight to blame it on. Someone I assumed, at the time, would simply always be around, the way you assume about weather, until the day I noticed they weren’t and couldn’t say exactly when that had become true.
The metric we default to is the wrong one
We tend to grade relationships by how long they lasted. Still together after ten years reads, almost automatically, as a success. Ended after eight months reads as a failure, or at best a phase, something you were still working out how to do properly when it stopped.
I don’t think that metric holds up to a single honest look at an actual life. Someone can know you for six months, closely and specifically, and change the direction of your life more than someone who has known you for fifteen years at a comfortable, familiar distance.
Permanence and depth get treated as though they’re the same measurement.
They aren’t.
One tells you how long something lasted. The other tells you what it actually did to you while it was here, and those two numbers frequently have nothing to do with each other.
Grief researchers have spent a long time studying a related version of this question, mostly in the context of bereavement, and their findings translate further than I expected them to. The continuing bonds model argues that maintaining a continuing relationship with someone who is gone, changed in form but ongoing, is not a sign of pathological grief but a healthy and adaptive feature of mourning for many people.
That research is specifically about death, and I don’t want to blur that distinction carelessly. But the underlying claim, that a relationship doesn’t have to keep actively existing in order to keep shaping you, seems to hold for relationships that ended in far gentler ways too.
Endings of any kind, not only the final one, might not erase what a connection already did to you on the way through.
What research can and can’t tell you about the number three
I looked, a little skeptically, for anything that might back up the specific number from the podcast. The closest I found wasn’t really about three or four at all.
One recent paper studying Dunbar’s layered model of social relationships describes an innermost tier called the “support clique,” a group of people who provide significant support in times of need, with a mean size of five. Five isn’t three or four, and a support clique isn’t quite the same claim as a once-in-a-lifetime transformative bond. But it’s close enough, and specific enough, that I no longer think the man on the bench simply invented the shape of the idea, even if the exact figure was more feeling than finding.
I’m a researcher myself, though not in this particular corner of psychology, and I want to be honest about where the evidence actually stops. I don’t think there’s a fixed, universal quota of profound relationships waiting to be discovered in anyone’s life. What I do think is true, because I’ve now lived enough of it to trust the pattern, is that transformative relationships are rare regardless of the exact count, and that rarity is precisely what makes them worth noticing while they’re still happening rather than only afterward.
The frightening thing about hearing a number like that isn’t the number itself. It’s that it makes you start counting.
The show that ended and the walks that didn’t
Strangers on a Bench reached its 100th episode this year and then stopped. Tom Rosenthal, the creator of the podcast, just announced the ending on his Instagram rather than letting the show simply trail off.
I have to be honest here — I felt that loss more specifically than I expected to, for a podcast, which is a slightly embarrassing sentence to write down. Something that had reliably talked me down after difficult days was, quite suddenly, not going to keep doing that.
Except it still does, in a way.
I know I’ll still walk with the old episodes playing. Sooner or later, the same hundred conversations are going to circle back around, and they will still do the thing they always did, which is to quiet me down and hand me someone else’s life to think about instead of my own for half an hour.
The show ending didn’t undo any of that. It just meant the effect had to keep happening on what already existed, instead of on what might come next.
Maybe that’s the more useful response to the number three, in the end, more useful than trying to identify who does or doesn’t qualify. Not counting anxiously backward through everyone you’ve known, and not treating every new person as an audition for a permanent role.
Just paying closer attention while someone is actually here, on the theory that the ones who matter most rarely announce themselves in advance. If a specific loss is sitting heavier on you than a podcast ever could, that’s worth bringing to an actual person, a therapist or someone you trust, rather than carrying alone.
The podcast ended at a hundred episodes. I still put the old ones on and walk. Apparently, a thing doesn’t have to keep happening in order to keep accompanying you.
