Zagat discussion split from NY Forum

[rich]

And I think it's an insult to the raters. Just curious: How would you feel if a comment that a restaurant was "not as good as its rating" had been printed?

The same way - it's valid under similar conditions.

Since Zagat is publishing the average of thousands of opinions, this type of comment is not only valid but could be very important.

Let me give an example. A restaurant's average food rating from 1,000 submissions is a 20 and for the last couple of years, the number has been constant (give or take a point). However, in the latest survey, while the restaurant still gets a 20, there are 10-15 ratings (which is not enough to sway the average) that are either six or seven points higher or lower and the comment is made about better or worse than its rating.

Printing this is important because a couple of things could have occurred. The restaurant may have gone through a recent change that a lot of diners didn't experience. Or, the food is more appealing to a certain group (people who enjoy game, offal, etc - the list is endless). Another possibilty is that the kitchen has become inconsistent.

Another factor is timeliness. One fault of Zagat is that there is no way of knowing if the people submitting have actually eaten in the restaurant during the last year. In that scenario, those comments become even more important.

Edited September 1, 2004 by rich (log)

[Pan]

Since you've acted as a rater for Zagat, you know that it's impossible for a rater to rate something 6 or 7 points higher or lower. The rating scheme is a scale of 0 to 3 only. Those ratings are then multiplied by 10 and averaged by the management, so the only ratings possible from individual raters are 0, 10, 20, and 30.

[rich]

That's interesting. I haven't done a Zagat survey in about 10-12 years. When I did it, you gave the actual number.

[menton1]

Since you've acted as a rater for Zagat, you know that it's impossible for a rater to rate something 6 or 7 points higher or lower. The rating scheme is a scale of 0 to 3 only. Those ratings are then multiplied by 10 and averaged by the management, so the only ratings possible from individual raters are 0, 10, 20, and 30.

Interesting that they are so secretive about this... if they really wanted to prove their veracity, they would have the polling results verified by an outside auditor. Plus, Zagat freely admits that he gets a lot of patronage and comps from many of the restaurants around town. One year a restaurant name is printed on the survey, the next year it's off. Others prominent restaurants are conspicuous by their absence. SMARMY!!!

Edited September 1, 2004 by menton1 (log)