In 1986, researchers at the University of Washington built an apartment.
Not a lab that looked like a lab.
An ordinary furnished apartment on campus, with a kitchen and a couch and a view of the water, where newlywed couples came and spent the day doing absolutely nothing special.
They cooked. They read the paper.
They watched TV and ate dinner while cameras recorded all of it.
The researchers were hunting for what predicts divorce, and everyone’s first guess is the fights. The interesting couples, presumably, were the ones who argued.
That’s not what the tapes showed.
What separated the marriages that lasted from the marriages that didn’t turned out to be almost invisible in real time: moments so small that the couples themselves couldn’t have told you they happened.
A comment about something out the window. A sigh from the kitchen.
And, crucially, what the other person did in the next three seconds.
Every tiny “hey, look at this” is a bid, and researchers counted thousands of them
The person who did the counting was Janice Driver, who started in the apartment lab as a volunteer and spent roughly a decade there, eventually as John Gottman’s doctoral student.
Her job: build a coding system for something nobody had bothered to measure.
The throwaway moments.
Their system treated any attempt to interact as a “bid.” Seven types in all, six of them neutral-to-positive and one negative.
A bid can be a question, a touch, a joke, or six words about a bird outside the window.
It’s not a request for a conversation. It’s a tap on the shoulder of the relationship: are you there?
What matters is the response, and the coding system sorted every response into three moves.
Turning toward: engaging, even minimally. A look up, a “huh,” an answer.
Turning away: missing it or ignoring it. The newspaper that never lowers.
Turning against: the irritated swat.
The researchers were granular about it, with five distinct codes just for the different ways of turning toward.
Nobody turns toward every bid. Nobody’s supposed to.
The question the lab could finally answer was what ratio ordinary happy couples actually run at, and what ratio precedes a lawyer.
Couples still married six years later had turned toward 86 percent of the time. Couples who divorced: 33
When Gottman’s team followed their newlyweds over the years after the apartment stays, the numbers split with almost embarrassing cleanness.
Couples still married six years after the wedding had turned toward each other’s bids 86 percent of the time.
Couples who had divorced by the six-year mark had turned toward 33 percent of the time.
Not a difference in how they fought.
A difference in whether “look at that boat” got answered on a random Tuesday, years before anyone was unhappy.
Gottman’s team came to describe the mechanism as an emotional bank account.
Every turn-toward is a micro-deposit, trivial on its own. Conflict, stress, and bad seasons make withdrawals.
The couples who survived their worst arguments weren’t better arguers.
They were arguing against a decade of accumulated deposits, and the lab data connected exactly that: the state of the friendship account predicted how repair went when things got negative.
More Bolde Stories
The dinner table predicted the fight
The cleanest version of the finding is the 2004 paper Driver and Gottman published in the journal Family Process, and its design is almost funny in its smallness.
They took 49 newlywed couples and coded just ten minutes of each couple’s dinnertime in the apartment. Every bid and every response.
Then, separately, a fifteen-minute discussion of an ongoing disagreement.
The question: does the dinner table tell you anything about the fight?
It did.
Couples who were playful and enthusiastic with each other’s small bids at dinner were the same couples who used humor and affection during conflict, the exact behaviors that keep disagreements from turning corrosive.
And when the researchers modeled the direction, the data fit better running from the daily moments to the conflict than the reverse.
The fight didn’t poison the dinner table.
The dinner table was the training ground for the fight.
Which quietly reverses how most people budget their relationship effort. The instinct is to save energy for the big conversations and coast through the boring stretches.
The lab data suggests the boring stretches are where the big conversations are won or lost in advance.
The researchers’ own caveat, and the part you can actually use
One honest note, straight from the source.
Driver and Gottman flagged in the paper itself that the work is “strictly observational and therefore correlational,” meaning it can’t prove that turning toward causes marriages to last rather than happy marriages producing more turning toward.
The 86/33 split is a real and striking pattern, not a mechanism with a warranty.
But here’s why the finding became one of the most-used ideas in couples therapy anyway: unlike most divorce predictors, this one is visible and cheap.
You cannot easily change your attachment style or your in-laws by Thursday.
You can answer the comment about the bird.
The bid is almost never about its content. “Nice day out” is not a weather report.
It’s the smallest possible knock on the door, and the three seconds after it are where, per a soundproofed apartment in Seattle and a decade of tape, marriages are actually built.
Not at the anniversary dinner. At the regular one, in the moments neither person will remember by morning.
