chefjohnpaul sent along an article (Nov 2000 Commentary, at
http://www.commentarymagazine.com/, can't link to it but can be found under back issues) that pretty well chiffonades Zagats, at least as a reliable instrument.
The author writes: "Perhaps most surprising of all, respondents to the Zagat survey are not required to document—through, for example, copies of receipts—that they have actually eaten the meals they claim to be evaluating. The survey’s voting controls consist only of computer programs that scan for major irregularities and discrepancies, which would hardly prevent a shut-in living in Bozeman, Montana from voting in the New York survey. Nor is there anything to prevent determined people from voting multiple times for their friends’ restaurants, evaluating restaurants visited long ago or ranked according to the opinions of others, or engaging in conduct calculated to teach a restaurant a lesson—or make it number one. Online voting, recently introduced, will no doubt make it even easier to cast fictitious and/or ill-considered ballots."
Then it gets worse...
The author also notes in a postscript that part of the reasons Zagats is popular is for the dearth of good food criticism...
[This message has been edited by Live_to_cook (edited 12-18-2000).]