Method
Two kinds of spread
Spread across three readings in one session is technique; spread across sessions on different days is state. Telling them apart tells you what to fix.
Guides on Method: Bone-pressed length, start to finish, Stretched flaccid length, and why the clinics take it, Girth, done the way the studies do it, The measurement glossary
Two numbers can both be called "spread" and mean opposite things. The spread between three readings taken minutes apart, in one sitting, is telling you about your technique. The spread between sessions taken on different days is telling you about your state. Confusing the two means fixing the wrong thing.
Within a session: technique
Three readings taken back to back, in the same sitting, under the same erection state, should be close together if the method is being applied consistently. When they are not - when the three lengths spread by more than a couple of millimetres, or the three girths visibly disagree - the cause is almost always something in the execution: pressure that varied between readings, a tape that dragged skin differently each wrap, a grip that compressed the site unevenly. Where measurement error comes from covers this territory broadly; within-session spread is the piece of it you can see directly, because the state is supposed to be held constant and anything that still moves is technique showing itself.
The fix for within-session spread is procedural: press harder and more consistently, lay the tape rather than pulling it, hold the steadying hand further from the site. It responds to doing the same steps more carefully, because the same steps are what varied.
Between sessions: state
Take a full set of readings today and another full set next week, both done carefully, and they can still disagree by more than either session's internal spread would predict. That gap is not technique failing twice. It is state: erection firmness that was not identically full both times, a room that was warmer or colder, a different time of day, or ordinary day-to-day physiological variation that has nothing to do with how well the ruler was held.
The fix for between-session spread is not "measure more carefully" - both sessions may already have been careful. It is controlling the conditions that vary between sessions: measuring at the same time of day, confirming full rigidity rather than "mostly" both times, and keeping the room at a similar temperature. Where that is not fully possible, the honest response is to report the between-session spread as part of your uncertainty rather than to treat one session's figure as definitive.
Seeing the two apart: a 3 by 3 log
The way to tell which one you are looking at is to take three readings in one session, log them, then repeat the same three-reading session on two more days and log those too. Nine numbers, in three groups of three.
| Session | Reading 1 | Reading 2 | Reading 3 | Within-session spread |
|---|---|---|---|---|
| Day 1 | 14.2 | 14.4 | 14.3 | 0.2 |
| Day 2 | 14.6 | 14.7 | 14.5 | 0.2 |
| Day 3 | 14.1 | 14.3 | 14.2 | 0.2 |
The figures above are illustrative, not a claim about any real distribution. What the layout shows is the pattern to look for: a tight, consistent within-session spread across all three days (here, 0.2 cm every time) paired with a wider gap between the session medians (14.3, 14.6, 14.2) than any single session shows internally. That pattern says technique is fine and stable, and the remaining movement is state, day to day. A log where the within-session spread itself varies a lot from day to day - 0.2 one day, 0.8 the next - says the opposite: technique is not yet consistent, and that is the thing to work on before trusting any single session's median.
What this does not replace
This is about separating two sources of spread, not about how many readings to take in the first place - that groundwork is covered elsewhere - and not about the general vocabulary of systematic versus random error, which is a related but broader distinction that applies beyond sessions and days specifically.
Why the distinction is worth the extra sessions
A single session tells you a number and an internal spread. It cannot tell you whether that number would hold up next week, because it has no information about between-session variation at all. Two or three sessions, logged this way, are the cheapest way to find out - and knowing which kind of spread you are dealing with is what decides whether the next fix is a technique adjustment or a controlled condition, rather than a shrug.
None of this has an equivalent on the sites next door, because none of them are taking repeated physical readings of the same thing. Rate Cock scores a single photograph once; there is no within-session spread to log because there is no session in that sense. The same is true further along that chain - an image model's score is a single pass, a numeric rating does not repeat itself across sittings, and a human judge's opinion does not have a technique-versus-state distinction to draw, because an opinion is not a reading in the first place.