Marks from the first reading

A ruler or tape held firmly against skin leaves a visible mark for a short while after it comes off. That mark is displaced tissue, and a reading taken on top of it is not the same reading you took the first time.

What the mark actually is

Skin and the tissue just beneath it are compressible. A tape wrapped snugly, or a ruler pressed at the base for a length reading, temporarily flattens whatever it touches, and that flattening does not spring back instantly once the instrument is removed. For a minute or so afterward, the site sits slightly lower than its resting state. Girth in particular depends on tissue that has not just been compressed, so a second wrap taken immediately over the same mark reads a genuine circumference, just a temporarily smaller one than the site's normal state.

Why this quietly biases repeat readings

The practical effect shows up in exactly the situation you would expect: taking three readings in a row, back to back, at the same site, to get a median for the session. If each reading follows the last with no gap, the second and third are not independent measurements of the same stable thing - they are measurements taken on tissue still recovering from the previous one, and they will tend to land a touch lower than the first. That is not the technique noise the repeat protocol is meant to capture. It is a one-directional bias stacked on top of it, and it can make a set of three readings look tighter and lower than the site's true value, which is the opposite of what repeating is for.

The fix is simple

Either wait between readings - long enough for the indentation to visibly fade, usually under a minute - or shift the site slightly each time and note that you did, rather than landing the tape or ruler in exactly the same groove. For length, moving off the exact same skin contact point matters less, since the reading is against bone rather than surface tissue, but the same courtesy of a brief pause between attempts costs nothing. Recording the method next to the number is where a habit like this belongs - noting that readings were spaced, not stacked, is part of what makes a set of three trustworthy rather than quietly self-reinforcing.

This is a mechanical detail about the physical reading, not a claim about what any number looks like once it exists elsewhere. A photo-based score at Rate Cock, the AI model behind it at AI Penis, the published scores at Penis Rater, or a human judge's opinion at Rate Penis are unaffected by any of this - they work from an image, not a tape left pressed against skin.