marcus

This implicitly goes back to the flawed argument that since no rating system is perfect, all are equally deficient. This person is looking for the best restaurant in NY. If he picked up the New York Times and searched for the 4 star restaurants he would have found: ADNY Bouley Daniel Jean Georges Le Bernardin This would have provided a much more useful and better starting point for his investigation. In fact, because of restaurant magazine's ratings, he's asking the wrong question, Daniel vs GT, because he thinks that he already knows the 2 best restaurants, but in fact he doesn't.

Hot off the New York board. Posted without further comment. Hi all Was just checking the latest top 50 in restaurant magazine. The top two rated restaurants in NY are Gramercy Tavern and Daniel. Which would you recommend of these two? I am a bit of a NY virgin when it comes to restaurants. I've been to One If By Land, Two if By Sea, which I loved, and Asia de Cuba. Cheers' DJO

Many of the genuinely best restaurants in the world are quite small and low volume and haven't bothered with websites, Ambroisie is one example. They all take reservations, so you can call them and learn their specific policies. I'm not sure how individuals acquires a sense of entitlement that requires successful businesses to confirm to their desires and tells these businesses what they "should" do.

Carnegie Deli has always made their own corned beef and pastrami at some place in NJ. I don't recal what Ed Levine said, but he came across as very weak on investigative reporting. I believe that he was unable to determine the source of Katz's pastrami.

Well, we've been around on this before, but one of the main reasons that they made these choices is that more of them had eaten in certain restaurants than others. This had a major effect on the results. Is this a valid criterion??

The original concept of Atelier was no reservations, and they followed this for the first few months that they were open, beginning just about a year ago. The then changed there policy and started accepting reservations for the first seating for lunch and dinner only. I personally made a dinner reservation at Atelier for the last Saturday in November 2003 for the first seating at 6:30. As an aside, I found the food highly variable in quality, and every single person sitting on my side, the right hand counter, was English speaking. I am personally confident that Atelier is not the 100th best restaurant in the world, let alone the 4th. With regard to Robyn's post regarding Atelier not taking reservations, this is out of date. I don't know how other people feel, but I believe that eGullet is best served by posters writing about what they know directly. We can all Google the internet and find out for ourselves what's there. Using this kind of information as the sole basis for a posting doesn't seem to me to be generally useful.

Robyn is clearly mistaken, but that's not the complete answer. We don't speak to each other in terms of formal logic which is sterile and lacks utilityy. There is an assumed context of common understanding that allow us to communicate in a useful manner. If I told someone that Atelier takes reservations and did not provide the qualification, but only for first seatings, although it would be a logically correct statement, it wouldn't provide them with the information that they really needed. If they learned that I had known that, but didn't inform them as part of my answer, they would correctly view my response as having been misleading.

Yes, it does. half and half. Atelier takes reservation for its first seatings at 11:30AM and 6:30PM only.

It is not a question of intelligence, but interest level. Many people are reading this information through secondary sources such as a newspaper reporting on the top 10 for example. Most people even reading the article directly are not as focused on this as people are here. Also, what some of us are saying is that when you really look at the survey and its methodology you realize that the results are worthless. Is this what you mean when you say that people have figured out how the survey was done? With regard to rating systems factors, you are correct regarding NYT, but Michelin and Gault Millau as well claim that their rating systems relate only to the food. Michelin provides the additional caveat that for one star restaurants the standards may be adjusted based on the available level of cuisine in a particular geographic area.

To address your second point first, essentially all of these restaurants have been discussed on eGullet at length and comparatively as well, over the last few years. The names of the best reputed restaurants in the world are largely well known, and listing them in an unordered list does not provide any new information. What Restaurant Magazine purports to do that is new and valuable, is to list these restaurants in rank order. If you believe that this ranking is largely invalid, then they have added essentially nothing, and their exercise was meaningless. The fact is that they attempted to accomplish a meaningful ranking. The evolution of this thread has been primarily to discuss the validity of their methodology, which is a legitimate subject of discussion. There are many, many, other threads on eGullet that discuss these restaurants in detail, and if that is your interest, you can seek them out. In my view, you are making the same mistake that others have made on this thread, to whit, that because all methodologies are flawed, they are all equally invalid, and thus equal, and it doesn't matter what methodology you may choose to employ. In fact, although no methodology is perfect, some are much better than others and will yield better results. There are well developed principles of survey design including, sample selection, statistical analysis, that need to be employed to get reasonably valid results. In this case, the methodology was, for many of the reasons that I stated in previous posts and do not intend to restate, egregiously poor.

I believe you mean Buffet de la Gare in Hastings, but what don't you like about these places? Overpriced, unambitious food, less than the very best execution, not that these restaurants are bad. I might exempt La Panetiere from the third criticism, and admittedly I haven't eaten there since they got their newest chef. But, this restaurant has had a number of talented chefs that have gone on to open their own restaurants or work at major NYC restaurants, at which they were allowed to do better. What do you think about these places?

I think that we're saying the same thing. I also don't doubt that there are very good ethnic restaurants serving the various ethnic communities. Getting reliable information about these kinds of restaurants is difficult, and they have to be extraordinary, as some may well be, in order to be destination restaurants to which people are willing to travel substantial distances.

We stopped there for brunch one sunday afternoon and it would be generous to describe the food as fair, although dinner might well be much better. I have a recollection that it was reviewed in the NYT Sunday Westchester section and it received a Good, equivalent of one star, but I can't verify that. Westchester is not a restaurant destination, I worked in White Plains for many years and am quite familiar. Azuma Sushi in Hartsdale is quite good. There are a number of ethnic latin restaurants in Port Chester that can be interesting. I don't recommend La Panetiere in Rye, La Cremaillere in Banksville and Bistro de la Guerre in Hastings which are probably the most well known upscale places. Blue Hill at Stone Barn which is related to the New York restaurant is opening very soon, and may well be the most ambitious restaurant in the county, ever, from a culinary perspective.

What you are saying effectively is that all data points, opinions, are equal.

or more irrelevant?

You've got to be kidding!

Fair enough, I am looking forward to these restaurants as well. But I also think that Schrambling actually does have her finger on the pulse of the principal driving force behind the NY restaurant scene. Her being an out of towner is not really irrelevant, because this whole scene is largely driven by transplanted New Yorkers.

There are many ways around this problem. I suggested a normalization approach in a previous post that would work quite simply. Another approach would be to have the judges score the restaurants, even based on a very simple set of criteria. These scores could be used to create averages which are independent of the number of visitors to each restaurant. I am only addressing here the "fatal flaw" in the methodology, which remains extraordinarily weak, even when this "popularity" problem is fixed.

Fatal to the methodology and any thought that the results have any validity. Let me provide an example. Assume that 2 restaurants in the survey are exactly equal in quality and the judges rate them exactly the same. In this case assume that 20% of the judges that have eaten in each restaurant will put it on their top 5 list. The restaurant that is more visited, will get twice as many votes. Your second argument essentially states that because no subjective methodology is perfect, therefore all are equally bad. Just because we don't have perfect, doesn't mean that aren't important differences between good, fair and poor.

This is a straw man argument, no one is suggesting that a set of rating criteria needs to be restricted to one or the other, they should include both. The issue isn't the quality of the panel, but the lack of guidance that they were given. And, to repeat myself, that no adjustment was made for the fact that the panel had far more experience with certain restaurants than with others. My contention is that the overall methodology for this survey was so weak, that the results have little validity, and the list provides, on balance, a disservice to those who may choose to rely on it.

Suffering isn't the issue, it's truth in advertizing. If you tell someone that a restaurant rates 3 stars, and they go, but it only rates 1 star, they may not have suffered, but they didn't get what they were after, and you didn't do them any favor.

Not so. Every voter for the Oscars is required to see all of the nominated pictures. The fact that some restaurants in this survey were visited by significantly fewer judges than others, is truly a fatal flaw.

I feel largely the same way. All of these restaurants seem to feature attitude and novelty rather than substance. It's very illustrative of what the NY restaurant seen is all about and why some love it and others view it as second rate. I am a lifelong New Yorker, born and raised, but I belong to the latter group.

This is a very important point. This introduces a significant popularity factor into the ratings, all other things being equal, the restaurants that have been visited by more of the judges, will score higher. This causes a major systematic error in the results. This conceivably could have been compensated for by a normalization process where every judge listed the restaurants that they had and hadn't visited, and the scores of the less visited restaurants were appropriately increased. I see no indication that this was actually done.

Is this not a cop out?? Where does it say that this is an impressionistic list? Guides such as Michelin put significant time, effort and resources into doing the best that they can to identify and compare restaurants. They may not be perfect, but at least they're trying and not saying that because it can't be accomplished 100%, then it doesn't matter at all what they do, and anything goes. Atelier Joel Robuchon received no stars from the 2004 Michelin and a middling 15/20 from Gault Millau. Michelin gives one or more stars to close to 100 restaurants in Paris alone.