Why Two Wine Apps Rate the Same Bottle Differently

Because they are not measuring the same thing. One app averages what thousands of drinkers enjoyed, another aggregates trained critics, a third models your own past ratings. Same liquid, three different questions. Add the chance that one of them matched the wrong vintage and the gap widens further.

Three kinds of number wearing the same badge

Most ratings you see in an app come from one of three sources, and the badge rarely says which.

Where the score comes from	What it actually measures	What pushes it up
Crowd average	Enjoyment across many casual drinkers	Soft tannins, ripe fruit, low price
Critic aggregate	Quality judged against the wine’s category	Typicity, structure, ageing potential
Personal model	Fit with your own rating history	Whatever you have liked before

A wine can score 3.6 on the first and 91 on the second without any contradiction. The crowd found it austere; the critic found it correct for a Chablis. Both are true statements about different audiences.

The bottle you scanned may not be the bottle they rated

Label recognition is a matching problem, and it fails in ordinary ways.

Producers reuse the same artwork for a decade, so the only thing separating the 2019 from the 2022 is often four small digits, photographed at an angle in restaurant lighting. When the match lands on the wrong year, you get a real rating for a real wine that is not the one in your hand. In a region with a weak vintage next to a strong one, that alone can account for the whole gap between two apps.

Worth checking before you trust either number: does the vintage on screen match the vintage on the glass?

Who rated it matters more than the rating

Crowd scores are built from self-selected samples, and different apps attract different crowds.

An app whose users mostly buy supermarket wine will produce a database where supermarket wine looks excellent, because the people rating it chose it and enjoyed it. An app used by collectors will rate the same bottle lower without anyone being wrong. This is the same reason a lean, food-built wine gets marked down by drinkers tasting it alone on a Tuesday. We went further into what those numbers can and cannot carry in are wine ratings reliable.

What to do with two numbers

Stop trying to pick the correct one and use them as two different signals.

• Big crowd score, modest critic score. Approachable and easy to like. Good for a party, unremarkable with dinner.
• Modest crowd score, strong critic score. Probably structured, possibly needs food or time.
• Both low, from many raters. The one case where agreement is worth heeding.
• Both high, from very few raters. Not yet a signal at all.

The underlying problem is that a single number compresses away the thing you actually need, which is what the wine is like. Knowing a wine is high-acid and low-tannin tells you what to eat with it and whether you’ll enjoy it. A 91 does not, and neither does a 3.6.

That’s the gap AboutWine is built for: it explains what a bottle is and why it tastes that way, showing the reasoning instead of handing down a verdict. If you want the same logic applied to price rather than taste, how to tell if a wine is good value covers that side.

Want the reasoning instead of a rating? Download AboutWine free on the App Store.

Frequently asked questions

Which wine app rating should I trust?

Neither number on its own. A crowd average tells you how much people enjoyed a bottle, which correlates with soft, fruity and cheap. A critic aggregate tells you how a trained taster judged it against its category. They answer different questions, so the useful move is to read what kind of number you are looking at rather than picking a favourite app.

Why does the same wine show two different vintages in two apps?

Because label recognition has to guess. Producers reuse artwork across years, the vintage is often the smallest text on the label, and scanning in dim restaurant light makes it smaller still. If an app matched the wrong year, its rating belongs to a different wine in the same bottle shape.

Do low ratings mean a wine is bad?

Not reliably. A tannic, high-acid wine built for food will underperform with a crowd tasting it on its own, and a soft commercial red will overperform. Low scores often mean mismatch between wine and audience rather than a fault in the bottle.