Method
Measuring a curve
Published lengths are straight-line distances from bone to tip; following the curve with a tape gives a longer number that compares to nothing.
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
Measure a curved shaft in a straight line: length is the chord from the bone landmark to the tip, not the arc along the curve. A rigid ruler held across the curve gives you the chord automatically. The published method answers the question the same way every time, and the reason is simple geometry.
The convention
Length is the chord: the straight-line distance from the bone landmark to the tip, regardless of whatever path the shaft's surface actually takes between them. It is not the arc: the distance you would get by laying a flexible tape along the curve and reading where it ends. A rigid ruler naturally measures the chord, because a rigid ruler cannot bend to follow a curve even if you tried to make it. That is one reason rulers, not tapes, are the correct instrument for length. A flexible ruler that bends to follow the surface defeats this purpose entirely - why a flexible curve ruler is the wrong tool here explains what it measures instead.
Why the arc is longer
A chord connecting two points is always the shortest path between them; any curved path between the same two points, arc included, is necessarily longer. This is true regardless of how gentle or pronounced the curve is - a slight curve adds a small amount, a pronounced one adds more, but the arc never comes out shorter than the chord, and the two are only equal when there is no curve at all.
The picture
The solid line is the arc, following the curve's own surface. The dashed line is the chord, running straight from the same two endpoints. Length is the dashed line.
Why this is the convention, not just a preference
Two people with the same degree of curvature will always get the same chord for the same underlying dimensions, because a straight line between two fixed points has exactly one length. An arc, by contrast, depends on how closely the tape was made to follow the surface - press it flush against every part of the curve and it reads one number, let it bridge slightly across the tightest part of the bend and it reads a shorter, different one. The chord is the reproducible measurement; the arc is not, even before you get to the question of whether it is a longer or shorter number. This is also, separately, why a rigid ruler is the correct tool here and a flexible tape is not: a rigid ruler cannot trace the arc even by accident.
Holding the ruler on a curve
Rest one end of the rigid ruler at the base landmark, found the way bone-pressed technique describes, and let the other end reach toward the tip in a straight line, bridging across the curve rather than trying to follow its surface. On a pronounced curve, part of the shaft may lift away from the ruler in the middle of the span - that is expected, and it is exactly what a chord measurement looks like against a curved object. Read the tip where the dorsal line meets it, as defined elsewhere, not at whatever point on the ruler happens to be closest to the shaft's surface at the peak of the curve.
What this is not covering
The angle a curved shaft is held at relative to the ruler, and keeping the ruler parallel to the shaft's own overall axis rather than to the floor, is a related but separate discipline covered on its own. Curvature itself is also sometimes recorded as its own figure, an angle between the base and distal axes, taken with a goniometer rather than a ruler - that is a distinct measurement from length, with its own method, and this piece is only about how curvature affects the length reading, not about recording curvature as a number in its own right. Even clinicians find that angle depends on how it is measured. In Ohebshalom et al. (2007), The Journal of Sexual Medicine, 68 men with Peyronie's disease averaged 42 degrees by goniometer after an injection-induced erection, but 34 degrees from at-home photographs, which the authors concluded underestimate curvature.
Still a length, nothing more
Measured this way, the number is a straight-line distance in centimetres, comparable to what the pooled clinical literature reports, with the same technique-driven error bar as any other bone-pressed reading. It has nothing to do with how a photograph of the same shaft would be scored - a rating from an AI-judged site like Rate Cock or a person's opinion on Rate Penis is a judgement about an image, not a chord across an object, and a photo compared against other photos on Penis Rater works the same way. An app estimating length from a picture runs into the same wall as any of these - a photograph flattens the curve into two dimensions and loses the depth a chord measurement depends on. A rigid ruler bridging the curve in person is the only one of these actually measuring the distance the literature reports.