Five categories, one number out of nine, and a top mark nothing is expected to reach.
Every hotel here gets a number out of 9. Not out of ten, and not out of five stars, both of which stopped meaning anything some time ago.
The number is the average of five categories, each scored on the same scale. It is deliberately blunt. It exists so that two hotels can be compared without reading four thousand words about either, and so that the whole ledger can be put on one page in order, including the parts of it that no hotel would choose to have published.
Not the postcode, and not the view from the roof terrace. What is within walking distance of the front door at nine in the morning and again at eleven at night. A hotel in the middle of everything that requires a taxi to reach dinner is badly located.
Whether the bed is good, the shower has pressure, the light is worth waking up to, and the street outside stops at some point. Also whether anything is broken, because something usually is, and how quickly it stops being broken.
The restaurant matters, but breakfast matters more, because you will eat it every single morning and you cannot easily go elsewhere. A hotel with a tasting menu and a bad breakfast scores worse than one with neither.
Not measured by how many people say good morning. Measured by what happens when the booking is wrong, the room is not ready, or you need something at an inconvenient hour. Anyone can be charming when nothing has gone wrong.
Not the same as cheap. Whether the price was fair for what actually arrived, once the resort fee, the breakfast supplement and the fifteen euro flat white have been added up. This is why a £120 room can beat a £600 one.
The five are weighted equally. A hotel cannot buy back a bad room with a good restaurant, which is the correct outcome, because you sleep in the room every night and eat in the restaurant once.
The best of its kind anywhere. Nothing has one, and nothing is expected to. It exists to stop the top of the scale filling up.
Worth building the journey around rather than fitting in. Rare, and the review will say exactly what earns it.
A good hotel with no real complaint against it. Most of the places worth writing about live here.
Good, with something identifiable wrong: the road, the breakfast, the price. The review names it so you can decide whether it matters to you.
The true middle. Nothing wrong, nothing to report. Fine, and the review will be short.
A real fault that a night's sleep does not recover from, or a price that bears no relation to what arrived.
Rare, because most hotels are not this bad and the ones that are do not usually merit the words.
Every ten point scale collapses into a two point scale within about a year. Nothing scores below seven, everything decent scores eight, and the good ones score nine, at which point the number is doing no work at all and you may as well print a tick.
Nine is deliberately awkward. It has a true middle at five, which means a five can be used without it reading as a failure, and it makes a six mean something specific rather than mean disappointing. Reserving the top mark for something that may never arrive keeps the eights honest.
Half marks are allowed. Quarter marks are not, because at that point the precision is invented.
Every review carries one of four verdicts alongside the number. It is usually the more useful of the two, because it answers the question people are actually asking.
Given the same trip and the same money, yes.
Worth seeing, but there is no reason to go back.
The hotel is right and the room was wrong. The review says which one to ask for.
No.
Nothing is removed from the ledger for scoring badly, which is the only thing that makes the high numbers worth anything.
See the ledgerScores are revisited when a hotel changes in a way that matters: a new owner, a new kitchen, a refurbishment. The review says when it was last checked.
Corrections are welcome and acted on quickly. Prices and chefs move faster than a published page does. The questions page covers the rest.