How similar should couples be? The two things worth matching on

August 11, 2026 · 12 min read

How similar should couples be? For most of what holds a relationship together, similarity is the wrong axis. Trust, communication, empathy and self-regulation are qualities you want more of, in both people. Two partners who match at a low level are not compatible; they are simply both low on it. Of the nine dimensions LoveScore measures, five are scored on level (higher is better for both) and two on a hybrid rule — level plus a correction for how differently the partners see the same thing. Exactly two are scored on match: there is no better pole, and the couple score depends only on how close the two answers are, not on how high they sit. Those two are Togetherness & Space and Future Vision.

The distinction is not cosmetic. Trust can be built up, and more of it is always good news. A need for personal space has nowhere to be built up to: “a lot of autonomy” is not an achievement and “not much” is not a defect. A couple who live almost as one person and a couple who each keep their own trips and their own circle earn the same high score, because what gets measured is not the size of the need but the distance between two sizes. What follows is how that is calculated, what these two scales do inside the overall Love Score, and what a couple can actually do when their answers come apart.

Do opposites attract? The evidence, and the better question

“Opposites attract” is one of the most durable pieces of folk wisdom about relationships and one of the least supported. When researchers look at who actually ends up with whom, they keep finding resemblance rather than contrast: partners are more alike than chance would predict on age, education, religiosity, political attitudes and health habits, and — more weakly, but still in the same direction — on personality. Complementarity turns up in the story people tell about a first meeting far more often than in the data on who stays.

But “more similar is better” is a sloppy conclusion too, and a scoring model has to be more careful than a headline. Sameness is only good news where the trait itself is neutral. Two people who are equally quick to escalate are not compatible in conflict — they are a couple with two short fuses and no brake. Two people equally inclined to check a partner's phone are not compatible in jealousy either; mutual control is a configuration of tension, not of agreement, which is why jealousy stays on a level scale rather than a matching one.

So the useful question is not “how similar should we be” but “similar in what”. Some parts of a relationship are skills, and skills have a direction: there is a better and a worse end, and both people benefit from moving toward the better one. Other parts are settings — they have no better end at all, only a distance between two of them. Two of the nine dimensions are settings, and everything unusual about them follows from that.

How a matching score is calculated

Each partner answers five items on a scale, separately and from their own device, and gets their own score from 0 to 100. Level scales are then combined into a couple score that leans toward the weaker of the two results. A matching scale is combined by a different rule: take the distance between the partners' scores, and the wider it is, the lower the couple score. That is the whole operation. The average level does not enter into it at all — on a level scale a disagreement between partners is noise that blurs the measurement; here the disagreement is the measurement.

Reading a score like that takes three corrections. First, a high value says only that you agree — not that you have “a lot” of autonomy or the “right” plans for the future. Second, a small difference in answers is not a rift; it is ordinary human variation, the kind that closes with one conversation and an agreement about format rather than years of work on yourself — and the score stays high in that case. Third, when two people answer at opposite poles there is no partial credit on offer: that is not two shades of one life but two different lives, and the score drops to zero with no consoling middle.

Then the symmetric consequence, the one that raises most eyebrows: two people with equally low values get top marks on a scale like this. It is not a glitch. A couple where neither wants separate vacations is exactly as aligned as a couple who plan them every year — they simply live differently. Individual levels are not lost in the process: the core keeps them separately, so the report can show what you agreed on and not merely that you agreed. How all nine scales are built is laid out in our guide to the nine dimensions.

The first setting: how much personal space each of you needs

Here is the scene most of this starts from. It is Friday, and one partner says: “I'll go see friends this weekend — on my own.” To the one saying it, that is how you come back on Monday still a person. To the one hearing it, the sentence sounds like “I'm better off without you.” Both are right inside their own logic and neither has done anything wrong; they simply have different natural distances, and each reads the other's behavior on their own scale.

Togetherness & Space is about exactly that: how much time alone a person needs, how many pursuits of their own, how many decisions that do not require checking in. Inside the methodology the scale is called “Need for Freedom”; the public name is neutral because the construct is neutral. A high need for personal space in a relationship is not detachment and not an attachment problem; a low one is not codependence. The LoveScore report does not rate this need on a good-to-bad axis at all — it describes it as a setting, the way you would describe a preferred room temperature.

That is also why the report talks about a difference here the way it does. On level scales it can say that one partner is ahead of the other. On matching scales it cannot, by construction: if a gap is wide enough to be worth mentioning at all, only the direction gets named — one of you leans toward space of their own, the other toward shared time — with no leader and no laggard. A person with the “wrong” need for personal space does not exist in this methodology. The far end of the scale even has an archetype of its own, Free Spirit, and it is written as a way of living with its own costs, not as a diagnosis.

The second setting: where the relationship is headed

The other matching scale covers a shared household, official status and children. Same logic, higher stakes. The future is the territory where “working on yourself” does not help: “I want kids” and “I don't” do not average out, however good a couple's conversational skills are. That is precisely why the scale is scored on the gap — a mean here would lie comfortingly.

A couple who both value a relationship without a shared household and without paperwork are fully aligned on Future Vision. A couple where one is waiting for a proposal and the other is perfectly happy with the status quo get a low score — which is not a verdict on anyone's choice but a pointer to a conversation that keeps being postponed. We wrote about having it separately: how to talk about the future with your partner.

One softening is built into the algorithm. Very early in a relationship a gap on this scale is not treated as firm: for a couple who have only just come together, answers about marriage and children measure today's orientation more than a mismatch of plans — and the methodology says so out loud instead of pretending it has measured the future. The report handles the topic more carefully in that case and does not build alarming conclusions on it. For people whose orientation toward the long horizon is strongest, there is an archetype too: Long Haul.

How much the two settings actually weigh

It is tempting to assume that a dimension with no “right” value cannot matter much. The arithmetic says otherwise: the weight of these two scales changes from block to block, and almost everywhere it is larger than it looks.

• In the Love Score they are the lightest of the eight — lighter than trust and communication. But “lightest” is not “invisible”: a couple who come apart on both at once lose a noticeable amount of the overall score, and two mismatched settings pull the total down harder than the word “light” suggests.
• In the friction points block the ranking flips. A mismatch of future scenarios and a mismatch on personal space are among the heaviest components of that index and together account for a visible share of it. Disagreement on these topics strains a couple more than it moves the headline number — and the report shows both effects.
• The overall score has a ceiling. Future Vision belongs to the group of critical scales alongside trust, communication and peace of mind: a very low score on any of them caps the Love Score from above, however good the remaining dimensions are. Togetherness & Space is not in that group — a gap there lowers the score but does not put a lid on it.
• Across scenarios the weight changes radically. In the “moving in together” scenario, agreement about personal space is the heaviest component of all: once there is a single address, different needs for quiet stop being theoretical. In the “getting married” scenario the same component barely registers, while the intention to formalize the relationship becomes decisive. The same couple can score high on one scenario and low on another — a normal result, not a contradiction.

Hard flags deserve a note of their own. A wide gap on either of the two scales is marked out separately — as a sharp divergence in your need for personal space, or a sharp divergence in your pictures of the future. Those marks do not touch the score: they do not punish twice, they name a topic the report is obliged to state plainly instead of dissolving it into a general phrase.

When agreement does not count as a strength

The most interesting consequence of the matching type is that a high score here is not always news. Take a couple who answered “it depends” to everything. Both get the same value, the distance between them is nil, and the couple score on both scales comes out at the maximum. Formally a perfect match; in practice two people who have not thought about any of it.

This property of the model is known and accounted for. In the strengths block, agreement on personal space and agreement on the future are not counted automatically: both components carry a relevance condition. If the topic is not pronounced for either partner, the strength is simply not awarded — agreement on a question that matters to nobody is not a resource. And these two strengths are counted more strictly than the rest: it is not enough for a match to exist, it has to be distinct to qualify as load-bearing.

A similar guard sits at the entrance. If a person answers in an agreeable style — saying yes to a statement and yes to its mirror image — their level on a matching scale is treated as unusable, and the scale drops out of the calculation for that couple entirely. The reason is simple: a false level in one person would manufacture a false gap between two, and the gap is the only thing being measured here. How these checks work, and what the answer-quality status means, is described in the methodology.

A side effect: why your personal read stays quiet about these two

The rule “the score is the closeness of two answers” has a non-obvious consequence for privacy. A matching score depends on exactly one quantity — the distance between partners — which makes it give the other person away more cheaply than any other scale. You do not even need the number: the scale's position in a list is enough. A line saying “your picture of the future is among your strongest resources”, sitting next to your own high score, reads as “my partner must be up there too”. On a level scale that inference needs at least one number; on a matching scale a rank will do.

So the personal read — the one you read about yourself — shows neither the names of the couple scales nor their order. It is assembled from the reader's own answers plus the two shared numbers both partners already see in the free couple report. Answering separately is worth something exactly as long as the report does not hand your partner back to you by arithmetic.

What to do when the two of you have diverged

There is no exercise here that “raises the score”, and that is the honest part of the news. But a gap on a matching scale is the one kind of divergence that is closed by an agreement rather than by self-improvement. Three things that genuinely work:

• Name the amount, not the verdict. “I need two evenings a week to myself” is a negotiating position. “You're suffocating me” is a sentence, and no decision follows from a sentence.
• Negotiate the format, not who is right. A difference in the need for personal space has no correct value, but it does have workable arrangements: an evening each, a hobby that belongs to one of you, one trip apart per season with nothing hidden. The argument about which of you is normal produces no decisions.
• Say the future out loud, and early. A gap in scenarios rarely appears out of nowhere — it usually goes unspoken for years. A difference discovered over dinner is a conversation; the same difference discovered while signing a mortgage is a crisis.

And here is what this frame does not do. It does not predict what will happen to your couple, it hands out no diagnoses, and it does not compare you with other couples: a matching score measures the distance between two of your own answers, not your place in somebody's ranking. It does not say which of you is right. It says where exactly the gap between you sits — and what to do about that is decided by two people, in words.

Frequently asked questions

Keep reading

LoveScore is an entertainment and educational service for adults 18+. Articles and reports are not a psychological consultation or a diagnosis, and they do not replace working with a specialist.