The word does a lot of work it was never built for. In the model the term comes from, a lapse is one instance and a relapse is a return to the pattern. The model's central finding is about what happens in between: after a lapse, attributing it to a permanent personal flaw predicts a full return to the behaviour. The lapse is not the hinge. The explanation is.
So "was that a relapse" is really two questions. The first is definitional, and you own it — no study will hand you a threshold. The second is what am I about to conclude about myself, and that one has an evidence base.
Decide the definition in advance, in writing. If you have not, you will end up writing it in the hour after a slip, which is the worst state you will ever be in for the job. A preregistered survey of men in reboot communities found perceived relapse commonly followed by shame, worthlessness, sadness and in some cases suicidal thoughts. Notice what perceived is doing in that sentence. The shame does not follow from the behaviour, it follows from the reading of it — two people can do the same thing on the same night and only one of them has had a relapse, because only one of them held a definition that said so.
A brittle definition has a specific failure mode. If one slip sets everything to zero, then by ten past eleven the cost of the next hour is identical to the cost of the last one: the month is already gone, there is nothing left to protect, and the rest of the evening becomes a write-off. A rule that only ever fires after the thing it was meant to prevent is not doing much preventing.
It is also inaccurate on the measures researchers actually use. Frequency alone is a poor indicator of problematic use — distress, impaired control and functional impairment are what distinguish it — and problematic use tracks distress far better than it tracks how much someone uses. Someone who has gone from most nights to occasionally, with the distress dropping, has changed something real. A counter that resets would show them nothing at all.
A definition worth having answers three things, and you can write it in five minutes while you feel fine:
- What counts as a slip. One episode. "Slip" is the better word for it because it does not arrive with a verdict attached.
- What counts as a return to the pattern. Not one evening. Something recognisable: back to most nights, or the environment changes abandoned, or the phone back on the nightstand and staying there. That is the thing you actually want to prevent, and it is worth naming precisely enough that you would know it when you saw it.
- What you do in each case. For a slip: stop the episode now, note the bare facts, do the thinking tomorrow, change one thing. For a return to the pattern: a proper review, and possibly a person rather than an app.
Keep the counter if you like it — knowing is better than not knowing. Just let it be one number among several rather than the verdict. The measure that moves first, and the one worth watching, is how quickly you get back on after a slip.
One evening is one evening. It is only day zero if you say so, and it is worth thinking hard before you say so.
Where this comes from
Read the full lesson: Abstinence or reduction? →
Sources and how strong they are (4)
- Well establishedMarlatt's relapse-prevention model identifies the Abstinence Violation Effect: after a lapse, attributing it to a permanent personal flaw predicts a full return to the behaviour.
- Marlatt GA & Gordon JR (eds) 1985, Relapse Prevention: Maintenance Strategies in the Treatment of Addictive Behaviors, New York: Guilford Press, ISBN 0-89862-009-0
- Curry S, Marlatt GA & Gordon JR 1987, J Consult Clin Psychol 55(2):145-149, doi:10.1037/0022-006X.55.2.145
- Well establishedFrequency of use alone is a poor indicator of problematic use. Distress, impaired control and functional impairment are what distinguish it.
- Bőthe B, Tóth-Király I, Zsila Á, Griffiths MD, Demetrovics Z, Orosz G 2018, J Sex Res 55(3):395-406, doi:10.1080/00224499.2017.1291798, PMID 28276929 (PPCS development)
- Bőthe B, Tóth-Király I, Potenza MN, Orosz G, Demetrovics Z 2020, J Sex Med 17(4):793-811, doi:10.1016/j.jsxm.2020.01.007
- WHO ICD-11, 6C72 (clinical descriptions and diagnostic guidelines) — https://icd.who.int/browse/2025-01/mms/en#1630268048
- Kraus SW, Krueger RB, Briken P, First MB, Stein DJ, Kaplan MS, Voon V, Abdo CHN, Grant JE, Atalla E, Reed GM 2018, World Psychiatry 17(1):109-110, doi:10.1002/wps.20499, PMID 29352554, PMC5775124
- Well establishedProblematic pornography use tracks distress better than overall frequency of use does. People who use frequently without distress and people who use frequently with it are distinguishable, and it is the problematic group that differs on mental-health measures - most clearly on depressive symptoms and self-esteem. Adult ADHD symptoms are a further replicated correlate (see C51), measured almost entirely by screening questionnaire rather than diagnosis. Loneliness is a further cross-sectional correlate of pornography use, measured once in a sample of 1,247 and modelled by its authors as potentially bidirectional.
- Bőthe B, Tóth-Király I, Potenza MN, Orosz G, Demetrovics Z 2020, J Sex Med 17(4):793-811, doi:10.1016/j.jsxm.2020.01.007
- Butler MH, Pereyra SA, Draper TW, Leonhardt ND & Skinner KB 2018, J Sex Marital Ther 44(2):127-137, doi:10.1080/0092623X.2017.1321601, PMID 28448246 (published online 2017)
- Dropped as a source, because it does not resolve to a paper containing this claim — Grubbs et al. 2020, Clin Psychol Rev 82:101925, doi:10.1016/j.cpr.2020.101925, PMID 33038740, whose systematic review of 415 studies cuts the other way on confidence
- Mixed evidenceA preregistered survey of men recruited from reboot communities found perceived relapse commonly followed by shame, worthlessness, sadness and in some cases suicidal thoughts. The same survey asked what participants had seen on NoFap forums: 73.7% reported misogynist posts, 49.1% bullying, 42.9% anti-LGBT, 32.0% antisemitic, 23.5% posts instructing followers to harm or kill themselves and 21.1% threats to hurt someone else. The sample is self-selected and not representative, and these are participant reports rather than a count of forum content.
- Prause N & Binnie J 2023, Sexualities 27(8):1608-1640, doi:10.1177/13634607231157070
More on patterns in the app.
The full 41-lesson course is in the app, every claim graded the same way, and an evening check-in that maps your own patterns. Free.
Coming to iOS and Android.