The Critics Are The First To Go

Anyone playing bingo cards reading “Smart Mobs v. Experts” or “How Distributed Intelligence Changes Journalism” will want to look at this pair of food articles from today’s New York Times. The basic story goes like this: the 2004 Zagat’s Guide is out today, and right in the middle of the Great and the Good of the Manhattan restaurant world, clocking in at #7 in overall rankings, is a small Brooklyn restaurant called the Grocery, run by a husband and wife team who happen to cook really really well. The Times coverage is unintentionally hilarious. Zagat uses cumulative anonymous ratings from diners who send in their opinions of various restaurants. The Times journalist, Florence Fabricant, goes on and on about how these ratings draw on as few as 100 people, obviously casting about for some way to explain how a 30 seat restaurant in Brooklyn could be rated above Alain Ducasse, where a bowl of soup will set you back almost forty dollars, while never noting that the alternative method of judgment — the impressions of a single restaurant reviewer — are a more limited sample. This is not to say that Fabricant’s criticism of Zagat statistics is flawed — it is non-existent. She makes not the slightest attempt to critique or even explain statistical sampling. She simply takes it as self-evident that a rating system that values anything other than the gilded charm of Manhattan’s most expensive establishments must be wrong. Her story is accompanied by a William Grimes re-review of the restaurant, which he previously gave one star, and the poor guy nearly gives himself a hernia back-pedalling on his earlier review. While noting that everything Grocery does they do perfectly, he also busies himself defending the idea that it is better to dine at Le Bernadin, no matter what those anonymous dopes who contribute to Zagat’s might say. This is all standard fare: an expert like Grimes, once brought to book, must find a way to change his mind without being seen to do so, because even more important than any actual opinion he might hold is the need to defend the idea of expertise itself, a pattern we’ve seen before. “Oh sure,” went the story in 1998, “Google is interesting for quick and dirty searches, but if you really want to understand the Web, you have to use Yahoo. They have a real live ontologist on staff.” Oops. And of course the third (and unpublished) part of the story is the one that really matters: the most important restaurant reviewer in the country was called to revisit an opinion because his earlier work was so at odds with the judgment of an anonymous and distributed group; he had to admit that yes, on sober reflection “The Grocery deserves a nearly perfect score”; and having made that admission, it is obvious to anyone who cares about food that the NY Times is now an also-ran compared to Zagat’s in terms of tracking quality over time. I forget if it was Enver Hoxha or Billy the Kid who said “Come the revolution, the restaurant critics will be the first up against the wall”, but truer words were never spoken, and with coordination costs of smart mobs so dramatically lowered, aggregate judgment continues to challenge expertise in an increasing number of arenas.