How it works

1. The personality assessment

Thirty-two bipolar questions, eight for each of the four classic dichotomies. Half the questions on every scale are worded in the opposite direction, which cancels out the common tendency to agree with whatever you are shown.

You get a four-letter type, but more usefully you get the strength of each preference. Someone at 52% Thinking and someone at 95% are both "T" and have almost nothing in common. Our compatibility maths uses the real numbers, not the letters.

2. The birth chart

We calculate your chart from scratch: planetary positions, the Ascendant and Midheaven, house cusps and every aspect. Your birth time is converted using the full historical timezone record, so a birth during a year your state skipped daylight saving is handled correctly.

If you do not know your birth time we say so plainly and leave the Ascendant and houses out, rather than quietly assuming noon and presenting a guess as fact.

3. Feelings and values

Thirty-two questions about how you behave in close relationships — reassurance, conflict, affection, independence, ambition, routine. In practice this is the layer that predicts whether two people can share a life, and it is the layer no zodiac or type site asks about.

4. Likes and dislikes

Forty-eight things — hiking, horror films, dogs, late nights, whether you want children — rated from “hate it” to “love it”.

Not scored by counting matching ticks. Where you both feel strongly and agree, it counts for a lot; where either of you shrugs, it barely registers. Your ratings are also measured against your own average, so being an enthusiastic person does not make you compatible with everybody — which is the failure mode of every shared-interests quiz on the internet.

A handful of items are treated as hard differences rather than tastes, and those are subtracted after the averaging so a long list of shared hobbies cannot bury them.

5. What shaped you

The deepest one. Your attachment position on two axes, eight core emotional needs, and — separately — how much of each of those you naturally give other people. It also asks about the home you grew up in.

Measuring what you need and what you give as two different things is what lets this work in one direction at a time. Whether a partner meets your needs and whether you meet theirs are separate questions, and the answers are often different. A relationship can be genuinely good for one person and quietly starving for the other, and we would rather say that than average it into a single number.

We ask about your childhood rather than guessing at it. Working backwards from someone’s taste in films to a theory about their upbringing is guesswork in a confident voice, and it is convincing precisely because you cannot check it. Asking directly, and then telling you where your answers line up, is the honest version. Everything in that report is written as something to weigh, not something we found.

The algorithm

Each layer produces its own score from 0 to 100. They are then blended by weight. Any layer with missing data is dropped from the blend rather than counted as zero — not having entered your birth details should not make you look incompatible with everyone.

Within each layer, we are specific about where similarity helps and where difference does. Two people who take in information very differently find conversation hard work, so similarity is rewarded there. A little difference in how you make decisions tends to help, each of you covering what the other under-weighs, so a moderate gap scores higher than none at all.

For the chart comparison, every cross-aspect is weighted by how personal the two bodies are, how exact the aspect is, and whether it is easy or challenging. Challenging aspects are discounted rather than treated as pure negatives — a comparison with no friction at all usually reads as pleasant and inert.

Astrology carries the smallest weight of the five. It is a real reason people come here and we calculate it properly, but it is also the layer with the least behind it, and it would be dishonest to let it outvote what you have told us you need.

Confidence

Every score is published with a confidence figure. It goes up when both people have completed both assessments and entered exact birth times, and down when they have not. A 91% built on two full profiles means something. A 91% built on one questionnaire does not, and we would rather tell you than let you believe it.

What this is not

This is entertainment and self-reflection, not science and not advice. Type assessments are self-report questionnaires with real limitations, and astrology has no accepted physical mechanism. We think both are genuinely interesting lenses for thinking about yourself and the people you know. Please do not make a decision that matters on the strength of a number from this or any other site.