2 ms·

TFA mentioned several ways in which Google reviews aren't an ideal tool here. Tossing out a couple more, you (1) don't have the same people giving reviews at ea

by hansvm 2y ago

TFA mentioned several ways in which Google reviews aren't an ideal tool here. Tossing out a couple more, you (1) don't have the same people giving reviews at each location, and (2) have a bias in those who choose to give reviews. As a point of anecdata about (1): Saturne in Paris (now closed) served some of the best food I've ever eaten, and it had lower ratings than a tourist-trap fish place on a pier near to where I live, even if you filter the reviews to only those describing the food. I'd be interested in seeing the same analysis with other metrics of quality, like the proportion of negative reviews referencing food vs other things (or a wilson-scored version thereof).

hobofan 2y ago
There is also a bias in the rating range of a city. E.g. in Berlin when looking to go out to eat I'd look for 4.5 and above as it uses quite a wide range for ratings, while e.g. in Rome most time I'd be happy with a 4.0. There are many phenomena (in Rome probably high tourist expectations) that can skew the distribution of reviews in a way that makes it hard to use as a "quality" scale.