Identifying fish by shape when colour is gone

FishingHQ ranks characters for a reason, and the ranking does not relax because the conditions are worse.

A DIAGNOSTIC character settles the identification on its own. A SUPPORTING character is consistent with it and would be shared with some other species. A SUGGESTIVE character is a tendency — usually true, not reliably so.

The failure mode underwater is to collect suggestive characters and treat their number as strength. Right general shape, right sort of place, right approximate size, right time of year: four suggestive characters, and a diver who feels certain. But the species that is genuinely confusable with the target shares all four — that is usually WHY it is confusable — so the four observations have not discriminated between them at all. They have only confirmed that the fish is one of the group nobody has narrowed down.

Stacking weak characters increases confidence without increasing information. That is the specific cognitive failure this family exists to name.

Where no diagnostic character can realistically be established underwater, the correct output is not a lower standard. It is the recognition that this particular identification cannot be completed in these conditions, from this angle, at this range.

Identification guides draw fish in perfect lateral profile. Fish underwater are rarely presenting one.

HEAD-ON OR TAIL-ON, the outline character is almost entirely absent and proportion is unreadable. A fish approaching directly is a fish that has not been identified, however clearly it is seen.

FROM ABOVE, body depth is invisible and width becomes the visible dimension — which is why fish that look completely different in profile can look similar from directly overhead, and why looking down on fish from the surface is a poor identification position even though it is a common one.

FROM BELOW, the fish is backlit against the surface and is a true silhouette: outline and fin placement are unusually clear, colour and pattern are entirely gone. This is genuinely the best angle for shape characters and the worst for everything else.

PARTIALLY HIDDEN — in a hole, under a ledge, behind kelp — a diver typically sees a head, or a tail, or a flank. A partial view can be conclusive when the visible part happens to carry a diagnostic character, and is otherwise exactly the situation the uncertainty rule was written for.

The practical instruction is to move rather than to strain. A second look from a different angle costs a dive; being wrong costs a fish that cannot be put back.

This is a skill built out of the water as much as in it.

WHAT ACTUALLY WORKS. Learning the shape families of the local fish community before diving it — not every species, but the outlines: the deep-bodied breams, the elongate predators, the two-dorsal-finned bass and their relatives, the flatfish, the wrasses. Once the family is instant, the question narrows to a species pair or trio, and species pairs are a manageable problem.

Learning which pairs are confusable BEFORE meeting them, and learning specifically which character separates them and whether that character is visible underwater. Where it is not, the pair is a known no-shoot problem rather than a surprise at four metres.

And learning the local juveniles, which are frequently the fish that look like something else entirely.

FishingHQ's species records carry ranked characters and named lookalike relationships for a large part of the catalogue, which is the raw material for exactly that preparation. What it does not carry, and what no platform can, is the confidence that a particular character will be visible on a particular day. That judgement stays in the water, and the rule it answers to does not move: if the identification is uncertain, do not take the fish.