Why review scores are a bad way to choose a restaurant when travelling

You land in a city, open the map app, and the first thing you do is sort by rating. You pick the 4.6 over the 4.4. You sit down, the food arrives, and it's fine — or it's a 4.6 in a city where 4.6 is what a mediocre sandwich shop gets when the owner's aunt writes eleven five-star reviews in the first month.

Review scores are relative to their city, not absolute. A 4.4 in a city where restaurants run a tight operation and the median diner is not a food critic is a very different place from a 4.4 in a city where the baseline is a two-star experience and the 4.4 is genuinely surprising. You are importing your expectations from a different market and the number doesn't tell you that.

This is not an argument against reviews. It's an argument about what the number is actually telling you, and what to do with the rest of the information you can actually get before you sit down.

What the score does and doesn't tell you

A review score, in the best case, tells you where a place sits relative to other places in the same city, at the same price point, with the same kind of service. That's genuinely useful if you know the baseline. You don't, because you've just arrived.

What it doesn't tell you:

• Whether the place will serve you what you want. A 4.6 Italian place may not have a gluten-free option, may be cash-only, may close the kitchen at 9:30 and the room at 10, and the review average won't tell you any of that. If the question is gluten specifically, the gluten-free abroad guide covers the checks that actually settle it. Reviews are written by people who already had the meal; the decision you need to make is before the meal.
• Whether the experience will match yours. Reviews from a business dinner are not reviews from a family dinner. A solo diner's experience of a place is not a group's. A weekday lunch is not a Saturday night. The average flattens all of that into a number.
• Whether the place is still the same place. Restaurants change chefs, change ownership, change the menu. A 4.7 from two years ago is not a 4.7 from last week.

The score is a weak prior. Treat it as one input among several, not as the one that decides.

The signals that survive travel

Some things you can observe before you commit, and they're more discriminating than a number:

The walk-in pace. A place where the room is turning over quickly and the kitchen is producing at a steady clip is telling you something about whether the food is going to be hot and made-to-order when you arrive. A place that's half-empty at dinner peak with a long wait for the food between courses is telling you something else. Neither is a verdict, but both are data.

What the regulars actually order. If you can sit for ten minutes — a coffee on the corner, the newspaper at a table you don't mind at — watch what the people who come back order. That's the menu the place is actually good at. The menu on the wall is the menu the place wants you to order.

Whether the kitchen can do one thing well. A small menu with six items where the kitchen has clearly been doing them for a long time is usually a better bet than a large menu where every dish is an introduction. Special focus is a real signal; it's also the kind of thing a 4.6 average from a tourist-heavy place will miss, because the tourists order the safe thing and the regulars order the good thing.

The language test. If the staff can explain the menu in a language you both understand, that's a place where a dietary requirement or an allergy question has a chance of landing. If the exchange is a series of nods, you have a decision to make before you order, and the review score is not going to help you make it.

The one question worth asking before you order

Most of the time, the useful thing to do is not to read more reviews — it's to ask the one question that the review average can't answer for you.

If you have a dietary requirement, the question is the specific one: Is this cooked with flour? Is it in the same pan as something gluten-containing? Not "do you have gluten-free options" — that gets you a nod and a salad. The specific question gets you a real answer, or a honest "we can't tell you, ask the kitchen."

If you're travelling with a group and nobody can decide, the question is the one that settles it: What is the one thing you'd actually want to eat here, and is it on this menu? If the answer is "I don't know" for three of the four people, the place isn't the problem — the decision is.

If you just need to know whether the kitchen is still open, the question is: What time does the kitchen close, separate from the room? In many cities — especially ones that run late bars — the kitchen closes well before the room does, and the review score will not tell you by how much.

One honest tool for the last of those — checking whether a place is open right now, what diet options it lists, whether there's a step you'll get stuck on — is NearEats, which shows open-now status, diet options and wheelchair access as plain facts rather than a score. It's one way to do it; walking down the street and reading the door works too, and asking a local works, and the review score is a starting point for all of it. But checking first saves the walk.

The short version

• A review score is a relative number, not an absolute one. 4.4 in your home city is not 4.4 in the city you just landed in.
• **The score tells you where the place sits, not whether it will serve you.** Open hours, dietary options, cash policy, kitchen close — none of that is in the average.
• Watch the room, not the number. Walk-in pace, what the regulars order, whether the kitchen has one thing it does well — these are more discriminating than a decimal point.
• Ask the specific question, not the general one. "Is this cooked with flour" beats "do you have gluten-free options." "What time does the kitchen close" beats the review average.
• The score is a starting point. It narrows the field. It doesn't decide.