What do the numbers say about your names together?
Two names either sound right side by side or they don't, and almost nobody can say why. Psychologists have measured a version of that pull, and what they found is real, faint and nothing like the answer the playground flame test used to hand out. Put both names in and see what the frequency records make of the pair. This calculator on Find The Norm uses name frequency and pairing analysis to score the linguistic and demographic compatibility of two first names.
What is name compatibility and is it real?
Name compatibility is entertainment with a small psychological footnote attached, because nothing in the relationship literature shows that the letters in two first names predict how long a couple stays together or how satisfied they are while they do, however neatly a score lands. The playground version, counting letters through F-L-A-M-E, has been running since at least the mid-20th century, and the Japanese o-mikuji fortune paper carries a name-based reading of its own. What the research keeps turning up instead is duller. Communication quality, shared values, attachment style and emotional intelligence carry the weight, and none of them has a documented link to spelling, which is why those predictors and not this score fill the relationship statistics page.
One real effect does sit underneath all of it. Jozef Nuttin, working in Belgium in 1985, found that people rate the letters in their own name as more pleasant than the letters they don't own, and they can't tell you why when asked. The finding held across 12 European countries, the US, and Japan. Pelham, Carvallo, and Jones reported in 2005 that the same faint pull turns up in where people settle and who they pair off with. Their estimate puts it at roughly 0.4% of variance in partner selection, small enough that you would never feel it working on you, and small enough that a shared initial is a coincidence rather than a reason.
Does the name-letter effect change who you end up with?
Barely, and only in the sense that a thumb resting on a scale counts as weight, because the name-letter effect belongs to a family of quirks researchers call implicit egotism, the unconscious pull towards anything carrying a trace of yourself, and every one of those quirks comes out small once somebody puts a number on it. People like their own birthday numbers better than other numbers, and they marry someone who shares their surname a little more often than chance allows. Both findings replicate, and both are small. What the effect describes is a faint positive charge sitting on your own letters, which then colours how you read those letters somewhere else. Frameworks like Myers-Briggs compatibility are at least trying to measure a psychological pattern.
The other half of the score is doing something more honest. Names rise and fade, so two names that topped the lists in the same decade usually belong to two people born within a few years of each other. Susan and Barbara came up in the 1950s and 1960s. Emma and Liam belong to the 2010s and 2020s. Whatever compatibility follows from that comes from age proximity and the shared references that travel with it, and the spelling has nothing to do with it. The SSA and ONS popularity series this calculator reads is simply a count of how often each name was given in each year, which makes the era half of the score the only half with a mechanism behind it. Our I love you timeline calculator shows how you compare against the full data set.
About this data
The figures behind this page come from published academic work and government name registers, and the score is shown as a position in a distribution rather than a verdict on anybody. Two names can land near the bottom and belong to a couple who are perfectly happy. Individual variation is the normal state of affairs here, not an exception that needs explaining.
Frequently asked questions
Take it as a conversation piece. The score is built from real records, but no part of it has been tested against relationship outcomes, so it cannot tell you whether a pairing will work. If you want an answer that holds up, watch how the two of you handle a disagreement. That predicts far more than any arrangement of letters ever will.
Two government registers. The Social Security Administration publishes every US first name given to at least five babies in a year, and the Office for National Statistics does the equivalent for England and Wales. Both are counts of registrations rather than samples, so the popularity curve behind each name is about as solid as population data gets.
No. Nothing in the published literature supports the idea that a name combination predicts romantic success, and no name test has ever been checked against couples followed over time. These tools run on name frequency and pairing analysis, which describes the names rather than the people carrying them. Shared values, conflict style and attachment security do the actual predicting. (Source: Gottman Institute; Sternberg RJ, 1986)
The Social Security Administration's 2023 figures put Olivia, Emma, Charlotte, Amelia, and Sophia at the top for girls in the US, and Liam, Noah, Oliver, James, and Elijah for boys. ONS 2023 has Olivia and Amelia leading in the UK, with Noah, Oliver, and George for boys. Since the 1960s the spread has widened a lot, back when John, James, Mary and Linda took a huge share between them. The top 10 names now cover a smaller slice of births than at any point on record. (Source: SSA 2023; ONS Baby Names 2023)
Small enough to be invisible in any single couple. Jones et al. (2004) found mild evidence that people hold slightly warmer associations with partners whose names share letters with their own, though the effect has not replicated consistently and it shrinks further once confounds are controlled. Proximity and mutual attraction swamp it completely in how people actually meet. (Source: Jones et al. 2004; Nuttin JM, 1985)
Mostly because the pool is lumpy. There are 26 letters in the alphabet, but every generation crowds its babies into a handful of popular names, so matching initials turn up far more often than a flat guess across the alphabet would suggest. Pelham et al. (2005) did find a slight overrepresentation on top of that base rate, and follow-up work has been mixed. Coincidence explains most of what you notice. (Source: Pelham BW et al. 2005; Jones JT et al. 2004)
Five things carry most of the weight in the research literature. (1) Positive sentiment override, reading your partner's ambiguous behaviour generously by default. (2) Conflict styles that fit, so a disagreement can be raised and settled without contempt, criticism, defensiveness, or stonewalling. (3) Shared meaning, by which researchers mean rituals and goals both people actually invest in. (4) Secure attachment, confidence in the relationship without constant anxiety or avoidance. (5) Similar appetites for physical and emotional closeness. Names, star signs and numerology sit outside all of it. (Source: Gottman JM, 2015; Johnson SM, 2019; Bowlby J)
You can read the context, never the person. Bertrand and Mullainathan (2004) sent out matched CVs and found that the ones with traditionally white-sounding names drew 50% more callbacks than identical CVs with traditionally Black-sounding names, which shows how much social weight a name carries whether it deserves any or not. A name also dates someone roughly, because popularity peaks in particular birth years. What it cannot tell you is a single thing about their personality. (Source: Bertrand M & Mullainathan S, 2004, AER)
Screenshot it, send it to them, and forget it. A low score means your two names share few letters, sit awkwardly together out loud, or peaked in different decades. None of those has ever been shown to predict relationship quality or how long a couple lasts. The score is a description of two words. It has no access to how either of you behaves at eleven at night after a bad day.
Three inputs. How the names sound together, which mostly rewards contrasting rhythms, since a short name against a longer one sits better than two names with the same stress pattern. Then the count of letters they share, which is the name-letter effect written into arithmetic. And whether the two names peaked in the same decade, which is really a proxy for age. The output is consistent and built on real records, and it is still a judgement about two words.
- Peer-reviewed academic literature and government population surveys. Each claim above carries its own reference.