What Does a Compatibility Percentage Actually Mean?

The same couple, two calculators on this site, 42 points apart — and neither one is broken.

Is 70% a good compatibility score?

Only if you say which calculator printed it. Here are the same two digits on two tools that sit three clicks apart on this site.

A 70 from the first tool is a shrug. A 70 from the third is close to the best result it hands out all year. Nothing is wrong with either one — they are simply unrelated scales that happen to be printed with the same symbol on the end.

Why do two calculators give us different percentages?

Because they are not adding up the same things, and they are not adding them up the same way. There are three scoring engines behind the compatibility tools here.

That last row is the one that surprises people. Two placements landing in unrelated elements is the most common outcome for any single pair, and it scores nothing at all. Four zeros is a perfectly ordinary chart combination, so the analyzer prints a bare 0 for 6.45% of couples — roughly one in fifteen.

It also means only a short list of totals is even possible. Four values drawn from 25, 15, 10 and 0 can add to 19 different numbers between 0 and 100. 5 and 95 are not among them: the analyzer can print 90, and it can print 100, but there is no way to build a 95.

What can each calculator actually print?

Every scale has a floor and a ceiling that nobody advertises. Enumerating them is the fastest way to learn what a percentage is worth.

Two things fall straight out of that picture. The date-only calculator has no bottom worth worrying about: the worst pairing it can find in a century of birth dates is 41, reached by an Aries with life path 4 and a Cancer with life path 5, and it occurs on 0.04% of pairs. And the Moon-sign tool cannot tell anyone bad news at all, because 50 is its floor.

The analyzer is the opposite. Its verdict bands were written on the assumption that scores spread evenly across 0 to 100, but the scoring rule does not cooperate.

Translating your score into a percentile

The useful question is never “is my number high?” but “how many couples score lower?” That is a percentile, and it is the only figure that carries across scales. Find your score in the column for the tool you used.

Read across any row and the three numbers have nothing in common. At the 90th percentile the three tools print 85, 80 and 50. At the median they print 69, 60 and 25. If you have ever compared your score on one site with a friend's score from another, this table is why the comparison told you nothing.

The Moon-sign column is the coarsest of the three. With only five reachable values, 44% of all couples get exactly 60, which is simultaneously the 10th, the 25th and the 50th percentile. A single number is doing the work of half the scale.

One couple, three calculators, three verdicts

Take two invented people and run them through all three tools without changing anything.

• Person A — born 15 June 1990, 08:30, New York. Sun in Gemini, Moon in Pisces, Venus in Taurus, Mars in Aries. Life path 4.
• Person B — born 3 November 1992, 21:15, London. Sun in Scorpio, Moon in Aquarius, Venus in Sagittarius, Mars in Cancer. Life path 8.

An average couple, a median-ish couple, and a rock-bottom couple — the same two people. The life path row is what rescues the first score: 4 and 8 is one of the strongest pairings in that table at 90, and it carries the heaviest weight.

One small detail worth noticing in the first row. The weighted average is 66.75, and the tool prints 66, not 67 — it truncates rather than rounds. Nothing hangs on it, but it is a reminder that the last digit of a compatibility percentage is decoration.

Across 6,000 randomly generated couples the pattern in that table holds. The date-only score is the higher of the two in 99.1% of cases, by a median of 43 points. In 29.5% of couples one tool says “Very Good” or better while the other says “Challenging Match” or worse. Both call it a good match in just 3.8% of cases.

Which part of your birth data moves the number?

On the date-only calculator, not the part most people assume. The three components are not weighted equally, and the star sign is not the heaviest.

Hold both sun signs fixed and vary only the birth years, and the final score moves by exactly 22 points. Not “up to” 22 for some pairs and less for others: the figure is 22 for all 144 sun-sign combinations, because the life path table always spans 40 to 95 and 40% of that spread is 22. Do the reverse — hold the life paths fixed and vary the signs — and the swing is exactly 30 points for all 81 life-path combinations. So signs do matter more — but only by a ratio of 30 to 22. Keep the two birthdays and change nothing but the years, and you can move the answer by 22 points — more than a fifth of the 0-to-100 dial.

There is one more quirk inside that engine. The sign table and the element table are written independently, and they contradict each other in places:

These are not bugs. They are two separate opinions, written by hand, averaged together. It is worth knowing about because it explains results that look inconsistent: two Fire-and-Air couples can be twenty points apart while the element line of the breakdown reads 90 for both.

Are four placements really four opinions?

Not quite four. The analyzer compares Sun, Moon, Venus and Mars, and presents the four rows as independent readings. Two of them are joined at the hip by orbital mechanics.

Venus never strays more than about 47 degrees from the Sun as seen from Earth, and a zodiac sign is 30 degrees wide. Sampling real charts confirms what that implies:

Because of that, the Venus row repeats the Sun row's verdict for 38.4% of couples — far more often than two genuinely independent comparisons would. The practical effect is that a strong Sun match tends to drag the Venus points along with it, and a Sun mismatch tends to drag them down. The score behaves a little more like three opinions than four.

The Moon and Mars rows are the ones carrying fresh information, which is a decent argument for looking at the breakdown rather than the headline. If you want to see the placements themselves before comparing anything, each has its own tool:

How to read your own result

1. Check what the tool asked for. Birth dates only means sun signs and reduced numbers. Birth times and places means real planetary positions, and it means your answer will change if you get the time wrong.
2. Find the floor and the ceiling. Run two identical birth dates, then two deliberately mismatched ones. Those two answers bracket everything that tool can ever say about anybody.
3. Place your score inside that bracket, not inside 0 to 100. This is the whole trick. A 70 out of a 41–93 range with a median of 69 is a shrug; a 70 out of a range whose median is 25 is remarkable.
4. Never compare across tools. Score everyone you are curious about on the same calculator. Rankings within one scale are meaningful; numbers from two scales are not.
5. Read the breakdown, not the headline. Which component pulled the total up tells you what was actually compared. The single percentage on top throws all of that away.
6. Do not let a low number mean anything. The 0 in the worked example above came from four unrelated elements, which is the single most common outcome the analyzer produces. It is a description of two charts, not a forecast.

If you only have birth dates and no times, one of these will get you the inputs the other tools want:

Terms worth knowing

Percentile

The share of couples scoring below you on the same scale. The only figure on this page that survives a change of calculator.

Reachable range

Every value a scoring rule can actually produce. Rarely 0 to 100, and never stated on the tool itself.

Element

Fire, Earth, Air or Water. Each of the twelve signs belongs to one, three signs per element.

Complementary elements

Fire with Air, and Earth with Water. Both tools that use planetary positions award partial credit for this pairing and nothing for the other combinations.

Modality

Cardinal, fixed or mutable — the other way of grouping the twelve signs, four per group. None of the compatibility scores here use it.

Life path number

A birth date reduced to a single digit by repeated digit addition. Its lookup table is the heaviest ingredient in the date-only score.

Synastry

The traditional practice of comparing two birth charts placement by placement. The analyzer is a simplified, four-placement version of it.

Weighted average

A total built by multiplying each component by a share that adds up to 1. Here: 0.40, 0.35 and 0.25.

All 13 tools in this guide

Authoritative sources

• Percentile — Wikipedia — why rank within a distribution travels between scales when a raw score does not.
• Elongation — Wikipedia — the roughly 47-degree limit that keeps Venus close to the Sun.
• Synastry — Wikipedia — the chart-comparison tradition these calculators simplify.
• Astrology and science — Wikipedia — the evidence on whether birth data predicts anything about people.
• Barnum effect — Wikipedia — why a generic compatibility reading feels specific to you.

Last updated: 2026-09-01 · Published: 2026-09-01 · Written by the miniwebtool.com editorial team.

This guide is for general and educational use. Every count on this page describes how these calculators behave — the ranges, medians and percentiles come from running their exact scoring rules over all 1.36 billion pairs of birth dates from 1925 to 2025, all 144 Moon-sign pairings, and every pairing of 75,138 planetary charts computed from the Swiss Ephemeris — and you can reproduce any of them. What the resulting numbers are said to mean about two people is a symbolic tradition and is not supported by controlled research. Nothing here predicts the course of a relationship, and no compatibility percentage should be used to make a decision about one. Tool behaviour is described as it works today and may change.