Numbers in human reviews

Some human judges give a number out of ten, and many do not. When one appears it is part of the performance - a closing gesture inside the register rather than a measurement - and a brief that asks only for a number usually gets the least interesting version of the clip.

The default assumption, and where it comes from

A lot of buyers arrive expecting a score because scores are the shape a rating usually takes elsewhere: star ratings, app reviews, a number on a card. Human rating borrowed the convention without borrowing the reason it exists in those other contexts, which is comparability across many raters. Fields that depend on human raters check that comparability directly: McHugh's 2012 review of interrater reliability defines it as the extent to which raters "assign the same score to the same variable". A single judge's number has no second rater to be checked against, so it cannot do that job, and most judges know it, which is part of why the convention is looser here than the expectation suggests.

Which registers tend to include one

A number fits naturally into worship - a big score is part of the enthusiasm - and into playful or honest registers where it can land as a light punctuation mark on the verdict. It fits less naturally into reassurance, where a number risks undercutting the specific reassurance just given by reducing it to a digit, and into clinical, where a score reads oddly against an otherwise flat, descriptive delivery. Judges working those registers are more likely to skip a number entirely, and that is a register choice, not an omission.

Why some judges refuse outright

A number invites exactly the reading this site keeps warning against: treating a subjective opinion as data. Some judges decline to give one specifically to head that reading off, because once a number exists in a delivered clip, buyers tend to fixate on it over everything else that was actually said. Other judges will give one but frame it heavily in register - "a nine, and here's why that means something different from what you think" - to keep it inside the performance rather than letting it float free as a fact.

Neither position is wrong. A judge sets their own boundaries around what they deliver, and a refusal to score is one of the more common ones. Why a judge's seven is not a score covers the mechanics of what that number is actually doing when one is given.

What a buyer loses by insisting on one

A brief built around "just give me a number" tends to produce the least interesting version of a clip, because it directs the judge's attention toward a single output rather than toward responding to what was actually sent. What a review actually responds to is the specifics in the brief, and a request that narrows down to a number gives the judge fewer of those to work with. The number itself, once delivered, also will not do what a buyer often hopes: it cannot be lined up against another judge's number to mean anything, because two reviews are two performances rather than two data points, and a seven from one judge and an eight from another are not on the same scale in any sense worth trusting.

Where a real number belongs

If comparability is genuinely what is wanted - a figure that means the same thing today and next month, from the same instrument every time - Penis Rater's scores exist for exactly that job, because consistency is the entire design goal rather than a side effect of one judge's mood. That consistency runs on a model rather than a person; AI Penis explains what the scoring is actually doing underneath, and it is worth reading before assuming a human number and an automated one answer the same question. If the number wanted is really a length or girth figure, that has an actual method and an actual unit, and a judge estimating from a photo cannot replace it.

A human number, when a judge gives one, is closer kin to a critic's star rating than to a measurement: informative in context, useless stripped of it. Knowing which kind of number you actually want before you brief a judge is most of the decision - Rate Cock's comparison of a human read against a scored one walks through the rest of it.