Our method and its limits

Three systems, one analysis

SoulResonance combines three perspectives into one profile: astrology (35 percent weight), Human Design (40 percent) and numerology (25 percent). Each system looks at a different layer. The weighting is a deliberate editorial choice, not a law of nature, and is shown with every analysis.

How we calculate

We compute all planetary positions ourselves, based on the astronomical models VSOP87 and JPL DE405. Your birth chart is derived from date, time and place, not from a sun-sign table.

Every change to the calculation engine runs against more than 1500 automated verification checks and 150 reference profiles. The same input always produces the same result: reproducible, consistent, verifiable.

To be equally honest: we measured the trickiest edge value, the exact moment a planet turns retrograde, against NASA's ephemerides (JPL Horizons). Our engine hits it within a few minutes: across all six Mercury stations of 2026 the largest deviation was just over five minutes, and an automated test permanently guards a ten-minute bound. For daily impulses even that does not matter. If you need second-precise professional astrology, specialised tools remain the better choice.

Where the texts come from

All interpretation texts come from a curated library of more than 10000 entries across more than 300 categories. They are editorially maintained and continuously extended. No interpretation is invented by an AI at the moment of your request.

The algorithm selects which text matches your constellation. The library is created with AI assistance, curated, checked and owned by humans. At runtime it is fixed: no interpretation is invented anew at the moment of your request.

What you answer yourself

Alongside the birth moment there are five self-reports. We wrote every question ourselves, in the spirit of well-known models. The reason is simple: other people's questionnaires are protected by copyright, and a copied test would be neither permitted nor better.

• Attachment style, 20 questions. Following the model of Bowlby and Ainsworth, with the four-category classification of Bartholomew and Horowitz (1991). Our own questions, nothing taken from the ECR-R.
• Love language, 30 questions. Following the model of Gary Chapman (1992). Our own wording; Chapman's original is protected by copyright.
• Conflict style, 15 questions. Following the model of Thomas and Kilmann (1974). Our own questions; we do not use the TKI instrument itself.
• Enneagram, 36 questions. In the spirit of the Gurdjieff, Ichazo and Naranjo school. Our own questions, nothing copied from Riso-Hudson, Palmer or Chestnut.
• Values compass, not a questionnaire. You rank three to seven out of twenty values. That selection is our own curation, not a model from the literature.

Every questionnaire requires at least 75 percent of the questions to be answered. Below that we store no result, because there would be none.

How well founded this is

Not everything we use stands on the same ground. Three classes, named openly:

• Empirically well supported: attachment theory (Bowlby, Ainsworth, four categories after Bartholomew and Horowitz 1991) and the conflict model of Thomas and Kilmann (1974, from organisational psychology).
• Widespread and useful, but not validated: the love languages (Chapman 1992) and the enneagram. Both are established in couples counselling; their factor structure does not hold up to academic scrutiny.
• Interpretive systems with no scientific claim: astrology, Human Design and numerology. What that means is stated further down.

What goes into the score

The compatibility score rests on seven pillars. How much each one weighs depends on whether you said you were interested in the birth systems during onboarding.

Since August 2026, the Enneagram and conflict style feed into the score as their own, deliberately small pillars. Both are only shown if you allow it , by default they stay private.

The strongest caveat first: the score is not a prediction. In 2020 Joel and colleagues showed across 43 longitudinal studies that the success of a relationship cannot be predicted from the traits of the two people involved. In 2024 Impett, Park and Muise tested the three core assumptions of the love languages and none of them held, which is why this pillar carries the least weight here. Both findings sit in the source code that sets the weights. The score answers a more modest question: how closely does this person match what matters to you.

What SoulResonance is, and what it is not

Important: SoulResonance is a tool for self-reflection. It does not replace therapy, medicine or professional counselling. In a serious crisis, please turn to professionals.

SoulResonance is:

• A tool for self-reflection: prompts that help you look at yourself and your relationships more consciously.
• A reference library: all core concepts openly explained in Knowledge and the glossary.
• Reproducible: the same birth data always produces the same analysis.

SoulResonance is not:

• Not a diagnostic tool: the analysis describes tendencies, not findings.
• Not a decision machine. Never base major life decisions on an analysis alone, ours or anyone else's.
• Not science: astrology, Human Design and numerology are interpretive systems with long traditions, but without scientific proof of effect. We do not claim otherwise.

Your data

SoulResonance runs on Swiss infrastructure and follows the Swiss data protection act (nFADP). The free calculators work without an account. Statistics are collected cookieless and anonymously, and only with your consent. Details in the privacy policy.