What I Would Change About Chowhound
Having visited this site for the last 6 months or so I have found that I learn very little about the quality of food at a restaurant. Usually half the people will say they love the food and the other half will say they hate it. Case in point is the recent thread on Hiro Sushi. After reading all the posts I still have no idea if the place is any good. Why not have a rating system on this site? Here is what I envision:
1. I want good Sushi
2. I go to Chowhound and type in Sushi
3. I see a list of top 10 Sushi places rated by the Chowhounds
4. Ratings could be for food, ddecor, service, etc.
5. Now I can pick one and read the posts on that particular place
This site could also allow users to rate other memebers' posts. That way I can see the top rated members and I can see what they recommend.
All of this doesn't seem very difficult to implement and would be a lot better than the state of affairs on this site today where after reading fifty posts on a particular place I still have no idea if it's any good.
On amazon it's possible to rate reader's reviews. Some form of this might be helpful here. Also would be super if there was data base by address. Some way to search good recommendations in say, the east 40s, Manhattan.
The original comment has been removed
Sounds very Zagat-ish or Citysearch-like to me, which I hope to God this site does NOT turn into! A top rated member may not be a person I personally follow as they're in another part of the country, so that has the potential for being completely useless for so many others.
Everyone's got different tastes; find members who seem to like what you like, and add them to your Tracking.
And if half of the people like Hiro Sushi and half don't, see if those people you're tracking like it, and go there yourself. Or, perhaps just GO there and see if you like it and become one of the half that like it. Sometimes you gotta eat some worms to get to the good stuff.
I agree - I often use Zagat to identify restaurants in a certain neighborhood, to get a list of ones offering a cuisine I'm interested in, to find an address/tel etc. - CH I use to see what other CHs think of those places - for good or for bad. Then I decide for myself whether to try a place. And, CHs often identify places that would have otherwise flown below my radarscreen.
I don't agree that you can't tell by reading a post like the one about Hiro Sushi whether it is good or not. The people that like it there explained why they liked it and if those are things that would appeal to you, then you should give it a try and assess it yourself. The people that didn't like it, explained why, and if those are things that would bother you based on your experience at other restaurants, then you should take that as a good indication that you wouldn't like it either.
I think that rating other posts is too subjective. I agree with LindaWhit that if you keep track of people's posts that you like you can get a fairly good idea yourself what you would like to try.
exactly, if people aren't telling you the information you want then you either have to further seek it out or find people for whom your opinions jive with. rating users and restaurants is highly subjective to so many things that in the end it might not be as helpful. people who do the fly by postings might heavily tilt things one way and especially with restaurants... consistency is a really big issue over short and long periods of time.
and personally... a big tip on if the place is good or not is to not trust the opinion of a poster that only orders tempura and california rolls in a joint commended for fresh fish. especially if they've demonstrated their lack of knowledge.
And Chowhound is nothing more than what we, its users, make of it - it's completely interactive. So if you read a post from someone who likes Hiro Sushi who briefly mentions something they liked that catches your eye, you should post a follow-up asking for more details. For that matter, post and say you'd like to hear from everyone in more explicit detail what they liked or didn't like about it so you can decide. A lot of posters give really fast "drive-by" posts without a lot of detail, but in my experience, if you ask for specific details about dishes or meals, you'll get much fuller responses.
One thing that's great about regular posters on local boards who take the time to post in some detail about their eating experiences is that, as LindWhit and pescatarian have pointed out, you get to know their tastes and how they jibe with your own. With the current software, you can track posters you like if you want via your MyChow page and see what they're up to. A rating system doesn't make much sense in light of this, because each chowhound has different tastes and will therefore find the posts and preferences of others valuable to a differing degree.
A rating system is pointless without solid controls and an unbiased audience. CH may have the first, but doesn't have the second. The audience here is decidedly biased (and for good reason).
Forum rankings are meaningless and either lead to flame wars or ego stroking, neither of which I would welcome here at CH. There would be no way to prevent people from giving a person one star because they disagreed with them about Restaurant X or, conversely, giving a long time poster five stars just because they post "I agree!"
A rating system serves no purpose because it is exceptionally superficial. If you doubt that, go to CitySearch and note that Denny's always has an average rating of 9.0+ out of 10.
re: Seth Chadwick
Long post - forgive me in advance.
I agree with Seth above, and actually, that's my problem with Zagat's. Ratings systems, unless the community is deeply expert (and this is enforced in some way), inherently end up reflecting conventional wisdom. Which results in a skew towards older, famous establishments and chains, with the smaller chowhound-type "finds" getting driven out. Lots of reviewers without deep expertise IMO = the antithesis of Chowhound.
Does anyone here use www.movielens.umn.edu ? It uses collaborative filtering to predict how much you'll like various movies, and continually "learns" what you'll like by having you rate movies you've already seen.
The beauty of MovieLens is that the system doesn't purport to, or need to, explicitly know *why* you feel the way you do. It's just math - the aggregate of all my ratings intersects in complicated algorhythmic ways (sorry, I'm a math dope) with other people's ratings, and results in good, reliable personalized recommendations.
I'd love to see the same principle applied to restaurant reviews. A system that would in effect be able to tell me "People who like x [in Toronto] also like y [in Oakland]" or even "People who already like x [in Toronto] really dislike the new y [also in Toronto]."
Having said that, Chowhound does something different, and just as valuable if not more so. I love it here :-)
I have been fascinated with collaborative filtering since its early days, with sites like firefly. I think it would work beautifully with restaurants.
Collaborative filtering is making a bit of a comeback, especially in so-called "discovery engines" that attempt to filter items for you to browse. Music sites like Pandora don't work quite the same way, but with similar results.
More than anything, collaborative filtering is fun, and the more input you have, the better. What a great database Chowhound could provide.
re: Dave Feldman
That's an interesting idea, but I think one of the great things about Chowhound is that we aren't assigning overall good/bad ratings to restaurants, necessarily, but instead saying, "This place has good fried fish, but everything else is terrible," or "The grilled meats at that Pakistani place are great, but you should go to Dimple for anything fried," etc. That would be hard to do in with collaborative filtering. The best way to experience Chowhound is to read it often, find people whose taste agrees with yours, and go with that. And do a lot of your own legwork.