Scientific dating: how evidence-based matching works

What if your dating decisions weren't based on an attractive profile or a smooth opening line, but on 40 years of relationship research? What if every match you received was filtered through the same scientific insights that relationship therapists use to help couples? That is the promise of scientific dating. And it is precisely what Onedayte is built on.

Scientific dating is not a marketing term. It is a methodological approach that sets three requirements: the matching must be based on factors that peer-reviewed research has shown to predict relationship success, the measurement methods must be validated, and the results must be transparent to the user.

The three pillars of scientific dating

The first pillar is attachment theory. Developed by John Bowlby and Mary Ainsworth, refined by for adult relationships, and empirically supported by decades of research at the Fraley Lab. Attachment style predicts how you respond to intimacy, conflict and distance. It determines which partner combinations are stable and which are destructive.

The second pillar is the relationship psychology of John . Forty years of research among thousands of couples yielded concrete, measurable behaviours that make the difference: the 5:1 ratio, the four horsemen, repair attempts, Love Maps, turning toward. These behaviours are not subjective. They have been filmed, measured and statistically validated.

The third pillar is the scientific critique of existing online dating, particularly the work of Eli Finkel (2012) and Samantha Joel (2017). Their conclusion is clear: the matching algorithms of existing platforms measure the wrong things. Personality similarity and shared preferences barely predict whether two people will be happy. What does predict it: the dynamic between two people.

"The variance in desire was almost entirely a property of the dyad, not a property of the individual."

— Joel et al., Psychological Science, 2017

This insight is confirmed by additional research.

"The matching algorithms that these sites use are not likely to be any more effective than the alternatives that they have replaced."

— Finkel et al., Psychological Science in the Public Interest, 2012

How Onedayte matches scientifically

The Onedayte process consists of multiple phases that together build a relational profile that goes deeper than any other dating platform.

Phase 1 is the profile: basic information, photos and dealbreakers. Phase 2 is the Attachment Scan: 12 scenario questions that measure your position on the dimensions of anxiety and avoidance. No abstract statements ('I feel comfortable with intimacy'), but concrete situations: 'Your partner hasn't responded to your message. It has now been 3 hours. What do you do?' The answers reveal your actual behaviour, not your self-image.

Phase 3 is the Doctor Conversation: an AI-powered conversation of 12 to 15 messages that measures emotional responsiveness, conflict style, repair ability and shared meaning orientation. The conversation progresses from safe to vulnerable, based on EFT principles. The AI analyses not only the content but also the pattern: defensiveness in word choice, avoidance in answer structure, openness in emotional expression.

Phase 4 contains the Match Boosters: optional deep-dives such as the Values Ranking (ranking 15 values), a friend reference (5 questions answered by a good friend), conflict scenarios and an external data integration. Each booster adds data points that make the algorithm more precise.

The result is a three-layer matching system. Layer 1 filters on dealbreakers (binary). Layer 2 filters on attachment compatibility (the most destructive combinations are excluded). Layer 3 calculates a weighted compatibility score based on emotional responsiveness (22 per cent), conflict style (18 per cent), shared values (15 per cent), fondness and admiration (12 per cent), repair ability (10 per cent) and additional dimensions.

Source: Onedayte evidence-based dating framework