By the numbers
Porn recovery in numbers.
Every figure on this page links to its primary source. Hedge language is preserved where the underlying research is genuinely contested. If a claim can't be sourced to a peer-reviewed paper, a clinical handbook, or an Apple/WHO/government doc, it isn't here.
Last verified: May 2026.
How widespread porn use actually is (data through 2023)
| Finding | What it means | Source |
|---|
| Pornhub is consistently among the 10-15 most-visited websites globally, with recent (2023) estimates around 5+ billion monthly visits. Comparable in traffic scale to Wikipedia, Netflix, eBay. |
The category isn't a fringe corner of the internet. It's mainstream-traffic infrastructure. |
Pornhub article · Wikipedia (2023+ data) |
| ~75-80% of adult-site traffic is mobile, overwhelmingly smartphone — a pattern that consolidated over 2018-2023. A decade ago this was a desktop activity. |
The smartphone-era shift is why the 11pm-in-bed pattern emerged. The phone in your pocket is the access vector. |
Pornhub article (year-in-review summary) · Wikipedia |
| 60-80% of US men aged 18-30 view porn at least monthly across surveys through 2023, with variance from question-wording. Wright's (2013) baseline of ~70% has held up across subsequent surveys. |
Among young adult men, porn use is closer to a majority behavior than a minority one. "I'm uniquely broken" is mathematically wrong. |
Internet pornography (Wikipedia, aggregating 2018-2023 surveys) + historical baseline Wright (2013) |
| Average age of first exposure clusters around age 11-13 for boys across studies, with the trend going younger as smartphone access has become universal. |
For most current adults, first exposure happened pre-puberty or early-puberty — before any frame existed for what was being watched. |
Internet pornography article · Wikipedia |
| OnlyFans: 4M+ creators, ~$5B paid out to creators in 2023. Top creators report monthly earnings in the $100K to $1M+ range. |
Adult content is no longer centralized in studios. It's a creator economy, structurally closer to YouTube/Twitch. |
OnlyFans article · Wikipedia |
| Compulsive Sexual Behavior Disorder (CSBD) recognized by WHO in ICD-11 (2018) as an impulse-control disorder. APA's DSM-5 does NOT recognize it as standalone. |
Major clinical bodies disagree, but both acknowledge a real subset of users experience compulsive symptoms. |
CSBD article · Wikipedia / WHO ICD-11 |
The brain & reward system
| Finding | What it means | Source |
|---|
| Wanting ≠ liking. Dopamine drives the motivation to pursue rewards (wanting), not the actual pleasure (liking). |
Why "I crave it but it doesn't satisfy" is a real neurochemical pattern, not a personal failing. |
Berridge (2007) · Psychopharmacology |
| In compulsive sexual behaviour subjects, sexual cues activated the dorsal anterior cingulate, ventral striatum, and amygdala — same regions activated in drug-cue-reactivity studies. |
Neuroimaging evidence that compulsive porn use shares mechanisms with substance addictions. |
Voon et al. (2014) · PLoS ONE |
| The strict "willpower as finite resource" model (ego depletion) failed to replicate. 23 labs, N = 2,141, effect d = 0.04 (95% CI [-0.07, 0.15]). |
The "I'm out of willpower at 11pm" experience is real, but the depletion-of-fuel theory behind it is no longer well-supported. |
Hagger et al. (2016) · Perspectives on Psychological Science |
| People systematically over-predict the intensity AND duration of future emotional reactions ("impact bias" + "durability bias"). |
The brain's "this will feel amazing" prediction is reliably wrong about how good it'll actually feel and how long it'll last. |
Wilson & Gilbert — canonical affective-forecasting research |
Recovery experience & behavior change
| Time to reach 95% habit-automaticity ranged from 18 to 254 days, with average ~66 days. |
The popular "21 days to a habit" claim is wrong; "66 days" is the median, not a guarantee. Range is wide. |
Lally et al. (2010) · European Journal of Social Psychology |
| In the same study: missing one opportunity to perform the behavior did NOT materially affect the habit-formation curve. |
One missed day doesn't break a streak in any meaningful neurological sense. Forgiving streak-systems are evidence-aligned; punitive ones aren't. |
Lally et al. (2010) · European Journal of Social Psychology |
| Meta-analysis: physical exercise increased odds of abstinence in substance-use-disorder subjects by 69% (OR = 1.69, 95% CI 1.44–1.99) across 22 studies. |
Exercise is one of the strongest non-pharmaceutical interventions for craving across addictions. Generalizes to porn recovery cautiously. |
Wang et al. (2014) · PLOS ONE |
| Self-compassion is more effective than self-criticism at sustaining behavior change. Self-critics procrastinate more and learn less from failure. |
"Be hard on yourself" doesn't work as recovery strategy. The "be honest with yourself, kindly" framing is research-aligned. |
Neff (2023) · Annual Review of Psychology |
Post-arousal & sexual function
| 46% of women in a sample of 230 reported experiencing post-coital dysphoria (PCD) at least once in their lifetime; 5.1% in the past 4 weeks. |
The "crash after" is a real, documented experience — not just a subjective bad mood. |
Schweitzer et al. (2015) · Sexual Medicine |
| Internet pornography may condition sexual arousal to features that don't transition to real-life partners (clinical review). |
The mechanistic basis for porn-related sexual dysfunction in some users (the "PIED" pattern). Hedge: clinical review, not large-N RCT. |
Park et al. (2016) · Behavioral Sciences |
Compulsive sexual behavior & relationships
| Compulsive Sexual Behaviour Disorder (CSBD) is recognized by the WHO in ICD-11 (code 6C72) as an impulse-control disorder. Requires 6+ months of pattern + significant impairment. |
There is a clinical category for what some people are experiencing. (Note: APA's DSM-5 does NOT recognize it as a standalone diagnosis.) |
WHO ICD-11 |
| Betrayal Trauma Theory: severity of trauma response correlates with degree of trust placed in the betrayer. Intimate partnership = highest-risk context. |
Why discovery of partner's compulsive porn use produces PTSD-like responses in some partners — not over-reaction, but a documented trauma profile. |
Freyd, Betrayal Trauma Theory · University of Oregon |
| In a clinical sample of wives of sex addicts, ~70% met criteria for PTSD diagnosis after disclosure. |
Generalizes carefully — clinical sample, specific population. But the trauma response is not rare. |
Steffens & Rennie (2006) · Sexual Addiction & Compulsivity, 13 |
Physiology of distress tolerance
| Cold water on the face triggers the trigeminal nerve → vagal response → bradycardia and parasympathetic activation (the mammalian dive reflex). |
One of the fastest non-pharmaceutical ways to drop arousal in an acute moment. Used in DBT distress tolerance. |
Diving Reflex (StatPearls / NCBI Bookshelf) |
| Cortisol (stress/alertness hormone) peaks within 30–60 min of waking and decreases through the day to its nadir near bedtime — the diurnal cortisol slope. |
Why decisions feel different at night: the alertness hormone is at its low point. |
Hirotsu, Tufik & Andersen (2022) · PMC / Sleep Medicine Reviews |
The Escape app: hard numbers
| 11,868 | Adult domains in the bundled Safari content blocker. | Internal to the iOS app. Updated with each release. |
| 27 | Recovery courses in the website + app library. | Source: Course.allCourses.count (iOS app, v4.6). |
| 205 | Lessons across the 27 courses (each its own URL on this site). | Source: aggregated from Course.swift. |
| 211 | Messages in the Recovery Timeline (hour-by-hour through year one). | See /timeline.html. |
| 0 | Personal data fields the iOS app collects, transmits, or stores off-device. | App Store privacy nutrition label: "Data Not Collected" for personal data. See /why-escape. |
