Over my adulthood, I’ve been surprised numerous times upon hearing of a couple getting divorced because it seemed that they had a solid relationship. On the flip side, I’ve also been surprised that some married couples stay together because they seem to have a stormy relationship. If you think about the factors that would suggest marital happiness, you might conclude that couples with shared interests, being close to the same age, having a similar upbringing, being the same religion, having the same political views, being about the same intelligence and having the same level of education, and so on would be important.

But when it comes to predictions, there are two primary problems. First, more information doesn’t necessarily lead to better predictions. As I’ve written in a previous IFOD, our brains can only store about seven items of information in our working memory, which means that we reach information overload quickly. So, trying to consider all the myriad aspects of a relationship to make a prediction isn’t something our brains can do.

Second, when trying to predict outcomes for complex systems like the economy, stock market, or marital relationships, it is hard to determine which factors matter and which don’t. We might think that being about the same level of attractiveness is a relevant predictor because we know a couple that got divorced where one spouse was hotter than the other, but maybe that had nothing to do with the divorce.

Robin Dawes was a psychologist who was a pioneer in the field of human judgment. He was best known for championing the idea that applying linear models with simple algorithms better inform judgment than trying to juggle all sorts of factors based on experience.

In 1976 he published a study that found that a simple formula predicts whether a marriage would be successful. Here it is:

A positive number is associated with marital happiness. A negative number suggests unhappiness.

The study was of 27 married couples who kept a journal for 35 days and noted:

• Acts of lovemaking;
• Quarrels ; an
• A self-rating of happiness on a 7-point scale.

Note that a “quarrel” is a fight — an interaction where a couple is mad at each other. Not just a passing disagreement.

The conclusion of the study was that “the rate of sexual intercourse minus rate of arguments was highly predictive of self-ratings of marital happiness.”

A follow-up study by different researchers replicated the results of the original study.

The takeaways:

First, a simple formula often bests expert judgment based on experience. The marital happiness formula was more predictive than the clinical judgment of marriage counselors with decades of experience.

Second, it’s okay to have quarrels, but it needs to be balanced by affection.