Methodology

How the numbers are calculated.

Infidelity Meter is a structured self-check, not a validated psychological test. This page shows exactly what it does with your answers, so you can judge how much weight to give the result.

Two separate numbers

Most quizzes about cheating give one score. That hides the most important thing: how deep the wound is and how good the conditions are for it to heal are different questions. A severe betrayal with full truth, an ended affair and real accountability often has a better outlook than a smaller one that keeps being denied. So the meter keeps them apart.

Damage, 0–100

How deep it goes: what happened, how long, how often, how planned, and how the truth was handled.

Recovery outlook, 0–100

How good the conditions for rebuilding are right now: truth, an ended affair, responsibility, empathy, openness, change that lasts, and whether both want it.

Damage

Damage combines four parts:

damage = 0.42 × event + 0.30 × deception + 0.18 × pattern + 12 × (areas hit / 12)

Event

Each thing you tick has a weight from 25 (secret flirting) to 100 (a double life). The heaviest item counts in full; each further item adds 12% of its weight, because several betrayals are worse than one but not simply additive. The result rises with how planned it was (up to +30%) and whether a clear agreement was broken (+8%).

Deception

Built from how much lying there was (50%), how the truth came out (30%), and how slowly the full story arrived (20%). "Trickle truth", where the story grows in rounds, counts heavily, because each new disclosure can reopen the injury.

Pattern

How often (45%), how long (30%), and whether something similar happened before (25%).

Areas hit

An area counts as hit when its estimated impact passes 35%. A double life sets damage to at least 85.

The twelve parts of a relationship

The ring shows twelve areas. Each item you tick has its own footprint across them: a one-off encounter hits sexual safety and exclusivity hardest, hidden debts hit the shared future, contempt hits respect and self-worth. Duration, repetition and planning widen the footprint. Lying adds to trust, predictability and communication; blaming the hurt partner adds to self-worth, emotional safety and handling conflict.

#AreaThe question behind it
1TrustCan I believe what I am told?
2PredictabilityDo I know what to expect?
3Emotional safetyCan I be vulnerable without being hurt?
4RespectAm I treated as an equal?
5ClosenessDo we still feel close?
6Sexual safetyIs our sexual life safe, for body and feelings?
7ExclusivityIs this relationship ours alone, as agreed?
8Self-worthHow do I see myself after this?
9Shared historyCan I still trust our past, or was it rewritten?
10Shared futureCan we still plan together?
11CommunicationCan we talk about it, and about anything else?
12Handling conflictCan we disagree without making it worse?

Each area's health starts from 100 minus its impact, scaled by overall damage. If repair is happening, part of the loss is recovered over time, faster the longer ago it came out. Nothing is recovered while the affair is ongoing.

The LITE result shows the ring without labels. The PRO result names each area and explains it.

Recovery outlook

recovery = 15% truth known + 22% it has ended + 14% responsibility + 12% empathy
         + 10% remorse + 10% openness + 8% lasting change + 9% both want it
         − 12 if the hurt partner is blamed (−6 if partly) − 8 if it is played down

Two caps apply. While the affair or contact continues, recovery cannot go above 25, because rebuilding can't start while the injury is still happening. If any sign of control or danger is ticked, recovery cannot go above 20, and the result shows safety information instead of anything else.

The seven conditions

Separately, the result counts how many of seven conditions for rebuilding are met. These come from the shared ground of the main approaches to affair recovery (for example the Gottman "atone, attune, attach" model and Shirley Glass's work on walls and windows; full list of sources):

  1. The full truth is known: the full story arrived within weeks, not in three or more rounds.
  2. It has ended, including contact with the other person.
  3. Responsibility without blame: responsibility at 60 or above, and no blaming of the hurt partner.
  4. Empathy for the hurt: 60 or above.
  5. Openness: answering and showing without being pushed, 60 or above.
  6. Change that lasts: for months, not days, 60 or above.
  7. Both want to rebuild.

The closest type

The type is chosen from what you ticked, in order of how far it reaches into the relationship: a parallel life, then an exit affair, an affair, a one-off episode or serial cheating, paying for sex, an emotional affair, digital infidelity, micro-cheating, financial infidelity, leaving without leaving, and finally a breach of trust. All types are described here.

The PRO result

The PRO result uses the same numbers as the LITE result, nothing more. It is written by Infidelity Meter and assembled in your browser from fixed texts:

  • each of the twelve areas, named, with its health value and one of three descriptions (below 40, 40–69, 70 and above);
  • the areas grouped into likely to settle on its own (70+), needs deliberate rebuilding (40–69) and likely to leave a scar (below 40);
  • for each missing condition, what would change it, written for your role;
  • questions to ask for your type, and questions that usually reopen the wound;
  • an eight-week plan for your role;
  • warning and protective signs taken directly from your answers (ongoing contact, blaming, playing it down, one-sided wish to rebuild, type, deception and pattern scores).

It is encrypted on your device as soon as it is created. How that works is explained in privacy.

What this can't do

  • It is not a diagnosis and not a validated scale. The weights are a reasoned starting point, drawn from clinical literature on infidelity recovery (sources), not fitted to outcome data.
  • It can't tell whether you should stay or leave. It shows where things stand and what is missing.
  • It only knows what you tell it. If you're unsure about facts, the result is unsure too.
  • It can't detect cheating. On the suspicion page we explain why we won't help anyone spy on a partner.

If you're a researcher or clinician and see something to improve, write to hello@infidelitymeter.com. How we handle corrections is described in how we work.

Published 5 October 2026. Last updated: 8 October 2026 (added links to sources and review status; softened one claim).