RN
There is a whole genre of article devoted to correcting what porn teaches people about sex, and it has an awkward problem: it is mostly built out of statistics that no longer match the studies they came from. A figure gets rounded, then re-rounded, then attributed to a journal rather than a paper, and after a few years of circulation it arrives as common knowledge with nothing underneath it.
That is worth fixing before anything else, because a corrective built on bad numbers is not a correction. It is a second set of expectations to live up to.
Five numbers, checked
"Men last five to seven minutes." The measurement behind this is Waldinger and colleagues, published in The Journal of Sexual Medicine in 2005. They timed intravaginal ejaculation latency in 500 couples across five countries and found a median of 5.4 minutes.
Median, not average — and that distinction carries the whole meaning. The distribution is heavily skewed, the country medians ranged from 3.7 minutes upward, and individual times inside those samples spanned from well under a minute to well over half an hour. "Five to seven minutes" turns a wide, lopsided spread into a target. Half of the couples measured came in under 5.4 minutes, and that is what a median means, not what a failure looks like. Our own guide to lasting longer covers the practical side of this.
"Seventy-five percent of women never orgasm from penetration alone." The best current number comes from Herbenick and colleagues in 2018, working from a US probability sample of 1,055 women aged 18 to 94. 18.4% said intercourse alone was sufficient for orgasm. Another 36.6% said clitoral stimulation was necessary during intercourse, and a further 36% said it was not strictly necessary but made orgasms better.
Note what happened in translation. "Roughly one in five finds intercourse alone sufficient" became "75% never", which is a stronger and different claim — never is a long time, and the survey asked about what is sufficient, not about what is possible. The underlying point survives intact and is genuinely useful: for most women, most of the time, penetration on its own is not the thing that produces orgasm. That is worth knowing. The invented absolute is not. The female orgasm guide goes into the anatomy.
"Average erect length is 5.2 inches." This one holds up better than most. Veale and colleagues pooled data on up to 15,521 men for BJU International in 2015 and reported a mean erect length of 13.12 cm — about 5.17 inches — with a standard deviation of 1.66 cm.
The caveat is in the method rather than the result. That erect figure rests on 692 measurements taken by clinicians using a standard procedure, because self-reported measurements are unreliable in a consistent direction and the review excluded them. So the number is sound and the sample behind it is smaller than the headline "15,521 men" suggests. The reason the review exists at all is that its authors were treating men distressed about size; the nomograms were built as a clinical answer to that distress.
"Ten to fifty percent of women squirt." Here the honest answer is that the research does not support any single figure. Published estimates run from around 5% to around 58%, and the spread is not sampling noise — it is a definitional disagreement. Some studies treat squirting and female ejaculation as one phenomenon; others treat them as two, distinguishing a larger volume of dilute fluid from a small amount of thicker secretion. Depending on which question a survey asks, the prevalence changes several-fold.
An article that quotes one number from that literature has picked a side of an unresolved question without telling you it did. Two of our other guides deal with the anatomy and the claims made about it: is squirting real and the G-spot and squirting guide.
"Thirteen percent of straight people have had a threesome." The 2015 Sexual Exploration in America study found 10% of women and 18% of men reported ever having one. Collapsing two figures that differ by eight points into a single average loses the most interesting thing in the data, which is that the two reported rates do not match.
More useful than the lifetime figure: fewer than 3% reported one in the past year. It is an experience a meaningful minority have had, and almost nobody has often. If you want the practical version, threesome etiquette is more use than the prevalence rate.
Why the numbers keep degrading
Three failure modes account for most of it.
Medians get reported as averages. For anything with a long tail — duration, frequency, count of partners — those are different quantities, and swapping them makes a typical case sound like a standard.
Ranges get collapsed into points. A confidence interval or a spread across studies becomes one clean figure because one figure is easier to write a sentence around.
And journals get cited instead of papers. "A study in the Journal of Sexual Medicine" is not a citation; that journal has published thousands of studies. The vagueness is what allows a number to drift, because there is nothing specific for a reader to check it against.
What porn actually distorts
Strip the statistics out and there is still a real answer, and it is structural rather than numerical.
Recorded sex is edited. Scenes are shot across multiple takes and setups over hours, then cut into something continuous. Everything a viewer might infer about duration, recovery time, or uninterrupted intensity is an artefact of that process, and no amount of correct prevalence data addresses it, because the distortion is in the medium rather than in the facts.
Camera position is chosen for legibility. Positions that dominate on screen are frequently the ones that keep two bodies visible to a fixed lens. Comfort was not the criterion. That is why some of the most-filmed positions are also among the more awkward to actually hold — a point our guide to painful sex positions takes further.
Performers are a selected group, selected partly on physical characteristics. A category of professional entertainment is not a sample of the population, and comparing yourself against it is a category error before it is anything else.
And negotiation is invisible. Consent, boundaries and limits are agreed before recording starts, off camera, in advance. What reaches the screen is the part after all of that was settled — which makes spontaneity the single most thoroughly manufactured quality in the whole medium.
None of that requires a statistic. It follows from how the material is made, it does not drift when it is repeated, and it explains more of the gap between screen and bedroom than any of the five numbers above.