Everything people actually ask us about Sour Grapes — how the Sweet Score is built, what a badge is worth, who the Regulars are, and what happens when something on a listing is wrong. No accordions to click through. It’s all just here.
The part most people came for. Four doors into the same set of listings, depending on how you already think about dinner.
Start with your town if you know roughly where you're going, or search if you know roughly what you want. Both land you in the same place: a list ordered best first, on one scale, with the Sweet Score next to every place that has earned one.
Reviews is everything we track in one long ranked list you can sort and filter. Rankings splits that same board by category — best pizza, best sushi, best steakhouse. Browse cuts it by what you actually need tonight: outdoor seating, open late, takes reservations, good with kids. Same listings, three ways in.
Yes, and it's the nicest way to use the site. Dishes lists every dish reviewers actually named, each one a ranked page of the places serving it. We never invent a menu item — a dish only gets a page because somebody wrote about eating it.
The town index has every village across Nassau, Suffolk and Queens, each with its own ranked page. We know Massapequa is not Massapequa Park, and we know which one you meant.
Roughly what one person spends: $ is under $15, $$ is $15–30, $$$ is $30–60, $$$$ is $60+. It's a band, not a promise — you can always beat it or blow past it depending on what you order.
Because a listing without a grade is still useful. Address, hours, category, photos, whatever the regulars have added — all of that is there. What's missing is the number, and that's deliberate: we publish a score only once a place has clearly earned one. Everything else on the page is a fact, and facts don't need a bar to clear.
One headline number out of 100, built from where a place stands against everywhere we track. The whole method is public on the methodology page; this is the short version.
A rating out of 100, built out of a rank. That distinction is the entire point: a star average tells you how a place felt to whoever rated it, while the Sweet Score tells you where it sits against every other place we track. 90 means only one in ten sit above it, and that is a claim with arithmetic behind it.
Because it is, by construction. 50 is the median of everywhere we track, so a place scoring above it is genuinely above average rather than above some line we liked the look of. It is also the only threshold on this site that can be defended without hand-waving, which is why a number, a badge and a rank all switch on at exactly that point.
One of three places, in a fixed order. It is ours when we have been ourselves — anonymous visits, every check paid out of our own pocket. Failing that it is the Regulars', once enough of them have written the place up. Failing that it is built from the ratings the place already carries, weighted for how much evidence actually stands behind them. The order never changes and the page always tells you which one you're looking at. The full method is here.
Because nothing is perfect and nothing is worthless. The scale is built to land between 4 and 96, so no restaurant is ever called flawless and none is ever handed a zero. A 96 is the top of the board, and it still means somebody could have a better Tuesday.
Two reasons. Either the place hasn't cleared 50 — in which case we publish the facts and skip the grade — or there simply isn't enough evidence yet. Under 25 ratings, with nobody having been and no regulars filed, we don't place it at all. Too thin to be fair in either direction.
No, and not in the careful lawyerly sense of no. There is no internal process by which money reaches a score, because we never built one. Advertising is labeled, lives outside the rankings, and our reviewers don't know who advertises. Scores can't be bought. Only earned. How advertising works →
Yes, in both directions. Ranked places go back into the rotation on a fixed schedule — the same schedule for the ones we loved and the ones paying us for a banner — and nobody can request, rush or delay a return trip. Chefs change, kitchens find their feet, standards move. A score is a snapshot with a date printed on it.
Everything we control is spelled out: the anonymous visit, the receipt, the 10 questions the regulars answer, the thresholds on this page. The one part we keep back is the precise weighting used where a score leans on ratings a place already carries — publishing that would be an instruction manual for inflating one, and the people most motivated to read it are exactly the people it would help least honestly.
Four things we can say about a restaurant. All four are good, because we don't publish the other kind.
Local favorite is 50 and up — better regarded than the typical place we track. Top 25% starts at 78. Top 10% starts at 90. And 💎 Underrated is its own thing, explained two questions down. Badges are free, they can't be bought, and they ride on the Sweet Score alone. The badge key in full →
Exactly what it says. A score of 90 puts a place above nine in ten of everywhere we track — not above nine in ten of the places we felt like comparing it to, and not above nine in ten of some hand-picked shortlist. It is the reason the badge is worth wearing.
The best kind of listing on this site: a place rated 78 or higher that fewer than 40% as many people have found compared with a typical place in its own category. Rates beautifully, barely anyone knows. It is the only comparison we publish, and it only ever points one way.
Because a position is a fact and a grade is a judgement. Every listing gets its ordinal spot in the lists it belongs to — #41 of 73 is simply where it sits, and nobody is insulted by arithmetic. The 0–100 number reads as a grade, so it only appears once a place is above average.
No. There is no bad list here, no wall of shame, no red tier, no adjective for a place that hasn't cleared the bar. Sorting the sweet from the sour is the whole job — we just never show you the sour part. Why we only publish good news →
In this order, every time: the Sweet Score (the headline), the Regulars' score (computed from people who actually ate there), and would you go back (the share of members who said yes). Same order on every page, every card, every email, so nothing can be reordered into looking better than it is. When they disagree, that is usually the most interesting thing on the page, and we leave it visible.
No. It's one question put to every member, one vote per person per place, and you see the result only after you've cast yours. It isn't a rating — it's a recommendation, sitting next to the number rather than inside it.
The half of this site we don't write. Real people, real meals, ten short questions, and a score nobody hand-picks.
Members who write a place up. Everybody knows three or four places worth the drive; the Regulars are the ones who bothered to write them down. Their takes publish under their own display name, and their answers roll up into the regulars' score. Meet the Regulars →
Anyone can. It's free, it stays free, and the only qualification is that you actually ate the food. Sign up, or log in if you're already here.
10 forced-choice statements about one meal you actually ate — agree or disagree, strongly or not. About four minutes. There is no neutral option, because "it was fine" never helped anybody pick a Friday night.
A blank box asks you to be a writer. Ten specific statements just ask you what happened, which is the part a stranger can actually use. Your answers get assembled into a short write-up in the house voice, with your own best lines quoted as you said them.
Nothing publishes until you approve every word. Not a paraphrase you half-recognise — every word. Change anything, or bin the whole thing, right up to the second you hit approve. It goes out under your display name, not "a verified diner".
Once at least 3 takes exist for a place. Below that the page says Not ranked yet and means it. The score is computed from everyone's answers on one shared scale — no single person sets it, including the one who felt most strongly, and regulars never see our number before they answer.
No — and that includes a place you're related to the owner of, or run a rival to, and takes written on somebody else's behalf. Disclose a connection and we'll sort it out; hide one and it ends the account. The house rules →
Both. Editing is worth doing when a restaurant fixes something, and the edit is marked. You can delete your own take at any time. What you can't do is delete somebody else's, and we don't accept deletion requests routed through a third party claiming to speak for the author.
What the food actually looks like, from the listings themselves and from members who were there.
Mostly from the listings themselves. On top of that, members can add their own from a meal they ate — which are usually the more honest ones, because nobody styled the sandwich first.
Yes, once you're signed in. Photos and short clips, a few at a time, with an optional caption. The uploader tells you the size limits before it accepts anything. Join free if you haven't yet.
It goes into a moderation queue and waits for a person. That's a deliberate speed bump: it keeps staff photos, menus of other restaurants and the occasional act of vandalism off a page that somebody's livelihood is attached to.
Anything with staff in it, anything that identifies a customer who didn't ask to be identified, anything you didn't take, and anything that isn't the restaurant. The same rules that govern a take govern a photo. House rules →
Not at all. A great photo of a mediocre plate is still a mediocre plate, and a badly lit one of something wonderful is still wonderful. Photos help you decide; they don't feed the arithmetic.
Nassau, Suffolk, Queens first, then the rest of New York. Density before pins on a map.
Nassau, Suffolk and Queens, down to the village. Every town has its own ranked page, and we're expanding through New York from there.
Eventually, area by area. A new one opens when we can cover it properly — not when the map would look better with more pins on it. If you want to nudge us toward yours, tell us where you are.
Because a discovery site is only useful when the coverage is dense enough that your specific block is on it. Going nationwide on day one is how you end up with three listings a town and no reason to come back.
For now. Restaurants are where we started; they aren't where this stops. The scoring is deliberately written without any reference to food, so the same machinery works the day we point it at something else.
Please do. Add a restaurant takes about twenty seconds and it's on the map for good — searchable, rankable, open to everyone. Then be the first to write it up.
Two things are true at once: this is a business, and the number next to your name is untouchable.
Nothing, now or ever. Write to us through the contact page and we'll get it sorted. Claiming lets you fix your details, add photos and reply to takes in public. It does not move your score, change where you rank, or put you in front of anybody who wasn't already looking. What a claimed listing gets you →
Send us the restaurant, the page URL and the correct detail through the contact form. We fix facts quickly and we say what we changed. No charge, no account required, no queue you can pay to skip.
Yes — in public, underneath it, in your own words, with no time limit and no charge. It's the single most useful thing an owner can do here. The replies that land are short and unglamorous: thank the person, say what changed, invite them back.
No. Promoted placements never appear inside rankings — rankings stay ranked. Every ad is labeled as an ad, there is no native, no partner write-up, no wink. Spend cannot change a score, delay a score, soften a score, or trigger a visit.
Reach for a badge you already earned. A place ranked 50 or higher can pay to show that badge off to more people. A place that hasn't been ranked yet has nothing to buy — no package, no exception, no call to have. Eligibility is computed from the score and never negotiated. Advertising, in writing →
Either add the place yourself or write to us. A listing is free and being listed is not something we sell.
Free to join, free to leave, and nothing clever happening in the background.
Free to join and free forever. There is no paid tier for readers and no plan to build one — advertising is how this gets paid for.
No. Every listing, every ranking and every write-up is readable signed out. An account is for the things that put your name on something: writing a take, voting on whether you'd go back, saving places, adding photos, and adding a restaurant.
A profile with your takes on it, two saved lists — Want to Try and Been There — a vote on every listing, and your standing with the Regulars, which climbs on track record rather than volume. Your account page has all of it.
Your display name, on a public profile at your own handle. Not "a verified diner", not a number. One account per person, under a name you actually use.
Ask through the contact form and a person handles it. Deleting removes your profile, your email address and your saved lists. What we collect, in full →
Facts go stale, kitchens change hands, and sometimes we're wrong. Here is what we'll fix and what we won't.
Tell us and we'll fix it. Bring the restaurant, the page URL and the exact detail, plus anything that shows it — a receipt, a menu, a timestamp. Send a correction. We publish corrections with the same enthusiasm we publish everything else.
Only if it breaks a rule on the house rules page — a meal nobody ate, a conflict of interest, personal information about staff, threats, spam. We don't remove an honest first-hand account because somebody would rather it wasn't there. If you believe something in a take is factually false — not unflattering, false — write to us with the exact claim and whatever shows it, and we'll investigate properly and tell you what we found either way.
Anyone signed in can report anything, including an owner reporting a take about their own restaurant. Pick a reason, and a person reads it — not a scoring model, and not an automatic removal once complaints cross a threshold. If volume decided outcomes, the loudest party would always win.
No. Not advertising spend, not a sponsorship, not a partnership. There is no internal process for paid removal, because we never built the product. The flip side is just as flat: rule-breaking takes come down for free, for everyone, whether or not the restaurant has ever given us a dollar.
Say so, with specifics — dates, dishes, what changed — and we'll go back. One fresh anonymous visit, on our dime, by someone who hasn't read the first write-up. Ask for another look.
The short answers to the questions people actually open the About page for.
Sorting the sweet from the sour is the whole job. We just never show you the sour part. The name is a joke about the sorting, not about anybody's cooking.
Sour Grapes was founded by Michael Szerencsy, who lives here, eats here, and got tired of losing track of the places he actually recommends. It's a small operation, and that's a feature: there's no sales team, so nobody's bonus depends on where a place lands.
Advertising, and we'd rather say so than pretend otherwise. It pays for the meals we buy. Every placement is labeled, none of them appear in rankings, and none of them can reach a score. The wall, in writing →
Never. Clear 50 and you get a number, a badge and a spot in the rankings. Below it you get a full, useful listing and no adjective at all. There is no bottom of the list here because we don't publish one — a discovery site is for pointing at the good ones, not for grading everybody.
How ranking works has the whole method down to the arithmetic. About is why any of it exists, the house rules cover what stays up and what comes down, and contact reaches an actual person.
If the answer isn’t up there, write to us — a person reads it, and if enough people ask the same thing it ends up on this page.
Scores can't be bought. Only earned. Every number on this page comes out of the same code that publishes the scores — see how ranking works.