Four of the eight cat-breed ratings cannot tell breeds apart

The cat breeds here carry eight trait ratings each, on a one-to-five scale, from the source that records them.

It's tempting to rank breeds on all eight. Four of them can't be ranked at all, and knowing which is more useful than any list built on them.

The four that say nothing

Affection rates 54 of the 59 rated breeds at the maximum.

Child-friendliness puts 50 of them on the same value — and gives no breed the top mark at all. Dog-friendliness has 45 at the maximum, intelligence 39. A column where four in five entries are identical can't separate them, so an article ranking the most affectionate cat breeds would be ranking a tie and inventing the order.

That isn't a flaw in the source so much as a fact about it. It's describing cats generously, one breed at a time, rather than scoring them against each other. Reading it as a league table is the reader's error, and most sites that quote these numbers make it.

It's worth being precise about what a compressed column does and doesn't tell you. It isn't useless — 54 breeds rated affectionate is a real statement, roughly that the source considers affection a property of cats rather than of particular cats. What it can't do is answer the question people bring to it, which is always comparative: is this breed more affectionate than that one.

Why these four collapse and the others don't

The pattern isn't random, and the reason is worth knowing because it generalises well beyond this catalogue.

The four that fail are all judgements about temperament — how loving, how good with children, how clever, how sociable with dogs. Those are exactly the traits that depend on the individual animal, its upbringing and the household it's in, far more than on its breed. A source writing honestly about them has very little to differentiate with, so it describes most breeds positively and moves on.

The four that work are closer to physical facts. How much a breed vocalises, how much coat it has to maintain, how much it drops, how much it wants to move — those sit nearer the body than the personality, and bodies differ between breeds in ways temperaments largely don't.

So the rule of thumb for any breed data anywhere: the closer a column is to anatomy, the more it can tell you, and the closer it is to character, the more it's describing the writer's goodwill. That holds for dogs as firmly as for cats.

The four that work

Vocalisation splits the catalogue almost in half — 25 breeds at the quietest rating against 17 at the loudest.

Grooming is lopsided but real: 31 breeds at the lowest level, four at the highest. Shedding separates ten breeds at the bottom from a large middle. Energy puts 23 breeds at the top against four at the bottom. Those four are worth filtering on, and the breeds page does.

Even these have limits worth reading. Vocalisation has almost nothing in its upper middle, which means it sorts breeds into two camps rather than ranking them within either. Shedding puts 37 of 59 in one middle bucket, so it's informative at the ends and blank in between. Useful is a comparative judgement here, not a clean bill of health.

How to read any breed rating

Three habits make the difference between using a column and being used by one.

Look at the spread before the value. A breed scoring five out of five means nothing until you know how many other breeds also scored five. If most of them did, the score is a category label wearing a number's clothes.

Prefer the ends to the middle. Almost every rated column in this catalogue is informative at its extremes and vague in the centre, because that's how people rate things — the obvious cases get the obvious marks and everything else gets filed as typical.

Ask what would have to be true. If a column claims to rate intelligence across 59 cat breeds, ask how anybody could have measured that, on what cats, doing what. Where no plausible answer exists, the column is an opinion, which is fine as long as it's read as one.

Those three would improve most of what's published about breeds, including the parts of this site that quote numbers. They're also the reason this page exists rather than a ranking of the most affectionate cats, which would have been easier to write and more popular.

What a five-point scale can and can't hold

Part of the problem is arithmetic rather than judgement, and it's worth separating the two.

Five levels across 59 breeds gives you, at best, five groups of about a dozen. Even a column using the full range can't rank breeds within a level, and nothing in this data pretends to — so the finest answer any of these columns can give is which group a breed belongs to.

That's genuinely useful for a coarse question and useless for a fine one. Which of these two breeds sheds more is unanswerable; whether this breed is in the heavy-shedding group is answerable. Most disappointment with breed data comes from asking the first question of a tool built for the second.

It also means a column can look better than it is. A trait spread evenly over five levels is doing about as well as five levels allow, and that ceiling is low. The four working columns here aren't precise instruments — they're coarse ones that at least point in a direction.

Why this site says so out loud

Because the same discipline runs through everything here.

The name catalogue counts real registrations and leaves a field blank when nothing was recorded, rather than filling it with a plausible default. A breed rating that can't discriminate is the same problem in a different column, and pretending otherwise would buy a few pages at the cost of the thing that makes the rest trustworthy.

Dog breeds now carry the same eight ratings, filled from the American Kennel Club's published figures — editorial judgements, the same kind these cat numbers are, not measurements, and cited on each breed's page. They come with the same warning: the affection column rates most dogs at the top too, so it separates them no better than it does the cats. A breed the AKC doesn't rate is left blank rather than guessed.

What to use instead, for the things ratings can't tell you

If the four broken columns are the ones you actually care about — and affection and child-friendliness usually are — the answer isn't a better dataset. It's a different method.

Meet the animal. An adult cat's temperament is already formed and already visible, which is why a shelter can tell you what a specific cat is like with children, dogs and strangers in a way no breed table can. That's the single strongest argument for adopting an adult when a particular trait genuinely matters.

Ask the people who've been watching it. Shelter staff, a foster home, a breeder who's honest about a line's temperament — all of them are describing individuals, which is the level at which these questions have answers.

And be sceptical of anybody who tells you a breed is affectionate, good with children or clever without saying how they know. That's the sentence this page exists to make harder to write.

The trait ratings on this page are TheCatAPI's, on a one-to-five scale, and they are somebody's assessment rather than a measurement — which is why the dog columns are left empty. Where the breed data comes from.