The most common score in the category, and why

If you have used more than one rating tool, you have probably seen a 7 more than any other number. That is not a coincidence and it is not a reflection of the actual spread of submissions out there. Seven is where three separate effects land on top of each other, and none of the three has much to do with what was in the photo.

Three effects, one number

Most tools centre their results well above the scale's arithmetic midpoint, for reasons covered in full in generosity bias - a tool that scores low loses users, and the mapping between the model's underlying judgement and the printed display absorbs that pressure. That alone would put the centre somewhere in the six-to-eight range, not precisely at seven.

The middle of that range is also where results are most crowded. Distribution shape covers this more broadly, but the practical effect here is simple: a generous centre pulls the bulk of a population into a narrow band, and a narrow band packed with results has more of them landing on whichever whole number sits closest to its actual centre.

Rounding does the rest. A tool displaying one decimal place, or none at all, collapses a continuous range of underlying values onto a small set of visible numbers. If the true centre of a tool's output sits around 6.6 to 7.3, rounding sweeps a large share of that range onto the single digit 7, simply because 7 is the nearest whole number to wherever the actual centre happens to be.

What a seven actually tells you

Not much, on its own. A seven from a tool whose centre sits at 7.1 is the single most default result that tool can produce - it is closer to a null result than a distinguishing one. The informative part of any result is how far it sits from wherever that tool's actual centre is, not the raw digit, and a seven is frequently sitting right on top of the centre rather than meaningfully above or below it.

This does not mean a seven is a bad result or a good one. It means treating it as a verdict - "solidly good," "right in the middle" - reads more certainty into the number than the number is carrying. It is the scale's attractor, and most submissions on most tools are going to land there or near it regardless of what they actually depict.

Checking it against a tool you use

The two ideas above - generosity and crowding - are worth reading in full if a seven from a specific tool matters to you; where average sits on a scale is the piece to go to next for that. A tool that shows a breakdown across several named axes rather than one blended figure at least lets you see whether a seven total is hiding real variation underneath it, which a bare total never can.

The same rounding-toward-a-default happens wherever a bounded scale meets a generous reviewer or a piece of consumer measurement gear built to avoid disappointing a customer - a seven, or its equivalent on whatever scale is in front of you, is very often the path of least resistance for the whole system, not a specific finding about the one submission in front of it.