Hacker News Reader

Comment by Beltalowda | Hacker News Reader

Related anecdote: a friend of mine left a 1-star review for a restaurant a few months back; I was there, and it was definitely worthy of a 1-star review.

It doesn't show on the Google reviews for me. He sees it, I don't. Another friend left a 1-star review as well, and that one does show (and he also doesn't see the "hidden" 1-star review).

I don't know what's going on here, but it was a genuine reasonably written review detailing the problems we had. He never received any notification it was removed: just got kind of "shadow banned" or something.

View on HN

Google does some shadow filtering on their end. I had the same happen to a 1-star review - it showed fine to me when I was logged in but wouldn't show when searching for the business using an incognito tab in another browser. Had I not checked I would never have known that the review was invisible. The logged-in view faked the number of reviews (one higher than the incognito view), the aggregate score (0.1 lower) and showed my review in the info card.

The reason I suspect it's something automatic (at least in my case), and not something targeted by the business I was reviewing, is that I removed half of the review and it did start showing after around 10 minutes of waiting. I edited the removed half back in and it vanished again. I was eventually able to narrow it down to a couple of paragraphs. One of them contained swearing ("fucking" and "assholes", which the management of the business were), which I guess makes sense. The other was very innocuous so I have no idea why Google disliked it. Anyway, by that time I'd gotten bored of having to wait ~10 mins after every step of this bizarro dystopian binary search so I left it at that.

At this point it's received a handful of likes, and I received an email notification that a hundred people have seen it (and that I should register as a "local reviewer", yeah right). So I guess it's going to stay up for good now and don't need to check it daily out of paranoia any more.

And yes, this was the first and only review I've ever written on Google, in case that's part of the equation.

My biggest takeaways from this were that a) holy shit Google employees are a bunch of assholes and can go fuck themselves for writing code that shadow-filters a review instead of giving any indication that it violates their puritanical Christian policies, and b) Google is apparently not smart enough to fool me when I searched for the review from an incognito tab in a different browser but the same client IP, which almost seems like a mistake. Like, if you're going to make the effort to shadow-ban people, why not be thorough about it?

Personally I'm quite happy if the app filters out reviews with profanity. Not because I care about seeing profanity, but because, more often than not, people who leave reviews with profanity are jerks and I'd rather not be exposed to their opinions.

There's a difference between "Sorry, your review violates our guidelines point XYZ because of the sentence ABC and cannot be posted. Please edit it and try again." and "Your review was successfully posted. (haha not really loser, but imma set up a comically elaborate charade to make you think it was.)" Both of them keep the profanity out of your eyes, but only one of them makes you likely to get the honest review in the first place because it doesn't treat the reviewer like an outcast.

I'm not saying I agree with the parent commenter. But their point was not that they didn't want to see profanity, it was that they don't want to see anything at all from the type of people that use profanity. If that's you're position then shadow banning works well, whereas getting them to post reviews with profanity taken out does not.

And I'm saying that the profanity was not the only problem that caused the review to be shadow-filtered. Regardless of whether the system filters profanity or not, the fact that it filters them while pretending otherwise is the problem. If one supports the system that shadow-filters profanity, they're also supporting the system that shadow-filters "some other unidentified thing that seems perfectly fine", and thus leads to them not seeing reviews that contained useful information.

You think I can't write a review without profanity? Of course I can. I can be extremely polite while pointing out the issues I had with the business. That's exactly what I did when I realized this was happening. I just didn't know I was going to be held to that standard when I posted the review, and the system lied to me that what I wrote was acceptable so I had no reason to believe otherwise.

If you were relying on not reading reviews like mine because my use of profanity implied I was a jerk, well I'm sorry but your heuristic isn't going to work any more because I know now. It wasn't even hard for me to figure out, so I'm sure most of the jerks have figured it out too. So sorry if you were relying on that, but you probably can't any more.

Shadow-anything seems wrong. But the argument here is: "I don't want to see the reviews made by people who are vulgar" with the implied "I also don't want the system to ask them to remove profanity and post again, because I still think the review won't be useful to me".

The heuristic reasoning being: a review is of low value if it's written by someone who's default attitude is vulgar, regardless if that person is willing to correct self when requested.

The discriminator being: If a review has profanity initially, don't display it even if the review is improved. Also don't display other reviews by that person.

Not something I agree with, but it seems that's what user esperent meant.

> If you were relying on not reading reviews like mine because my use of profanity implied I was a jerk, well I'm sorry but your heuristic isn't going to work any more because I know now. It wasn't even hard for me to figure out, so I'm sure most of the jerks have figured it out too. So sorry if you were relying on that, but you probably can't any more.

And that's the idea behind shadow banning - the banned person is supposed to never become aware of the ban.

Don't pretend you've never been vulgar when upset. The most valuable reviews are those that are honest.

> Not something I agree with[...]

The purpose of shadow banning is to make it hard for violators to work around the restrictions. If the feedback loop is in the order of seconds, then the bad actor can quickly find a way to just pass the filter while still dumping the crap they want. The more cost each iteration incurs, the more the bad actor is deterred from doing that.

Of course, delaying the complaint and doing full shadow banning are still different things, and the former would suffice, but they chose the easy way out I guess, because their main objective is to keep the bad guys out.

This you get to "they were fscking a$$h0les", or "the management were a bunch of scunthorpes" depending on how your filter works

> There's a difference between "Sorry, your review violates our guidelines point XYZ because of the sentence ABC and cannot be posted. Please edit it and try again." and "Your review was successfully posted. (haha not really loser, but imma set up a comically elaborate charade to make you think it was.)"

There is. The difference is that in the first case, the ranter will tweak their language and continue ranting about the fvcking a55h0les, while in the second case, they will go away quietly. The second outcome is _clearly_ the best for all concerned.

Even better, just shoot all undesirables. Why do we need prisons anyway, and what is rehabilitation? Just kill everyone the second they break the law, so they'll all go away. This outcome is _clearly_ the best for all concerned.

Yes, yes, being murdered by the state is exactly like someone trying to improve the quality of Yelp reviews.

Dude, it’s a product owned by a private company. Do they owners not have rights to you? To compare it to murder is like the slipperiest slope ever.

Nothing stops you from reading only 5 star reviews

> The other was very innocuous so I have no idea why Google disliked it.

Beware of Google employees' agenda on language [1]. Something like "the restaurant service was insanely fast" can trigger a shadow banning.

[1] https://developers.google.com/style/inclusive-documentation

For reference, this was the paragraph (the business in question is an apartment complex I was renting at):

>Their "Courtesy Patrol" service is useless. I tried this service every night for a week but it made no difference; the noise still continued on that night and the following nights. Also, the people at the service told me that it takes them an hour to send someone over after you call them. So if you ever have a reason to call this service, you better be prepared that your problem won't get solved for at least an hour after you call.

How long was the review? It might just be that reducing the length/number of clauses helped. Honestly, if I was naively writing a restaurant review filter, length would be in there; if you look at TripAdvisor or similar, as reviews get longer the probability of the reviewer being a weird crank approaches one.

The length may have been a factor, but it definitely wasn't the only factor. When I eventually narrowed it down to those two paragraphs being the only difference, the attempt with the first paragraph in its original form and the attempt with that paragraph sanitized of profanity were basically the same length.

I'm here for a review, not an essay. I think Google was probably doing the right thing (as in, most useful for their users) with your review. Stick to a small number of quick, highly informative sentences, not lengthy speculation and opinion.

Holy sweet jesus I hate it when Jordan Peterson is right. They're going to police language and make us use words we wouldn't normally pick.

These are two disconnected things.

Google has a style guide for documentation that suggests that people do not include certain kinds of words. This does not imply that users posting content with these words will trigger policy violations.

“puritanical Christian policies” Based on the people I know who work at Google mountain view, and more general descriptions of the google workforce, “christians” are rare (or in hiding), I don’t think you can really blame them for those policies :) But I agree with your overall point

I understand the word choice, it is rampant throughout the Web, and also coincidentally I cannot think of a better description.

Maybe the corporate culture embeds it, or the behavior is used as a performance indicator.

Yeah obvious example is nudity rules which are enforced just fine by completely secular people even though they’re likely derivative of puritanical Christian culture in the US

Reviews can be measured on a number of relevant dimensions - positive/negative, useful/not useful, fair/unfair, etc. Vocabulary, grammar, spelling, typos are all information channels.

Honestly I can’t believe I’m defending Google, but it’s possible that they’ve determined certain language makes the review less useful. Evidence of that might be the other portion of the review which seemed to trigger the shadow ban without any profanity.

What does Christianity have to do with Google?

Keep your prejudices to yourself.

I find the most baffling part of it that Google is so anti-Christian as well.

I don't have anything against Christians. Maybe you should read the word before that.

Wouldn't that warrant some form of anti-trust case?

Wow, just checked a review I left, 1 star with video and phots of the issues, for an apartment complex. Yea, it's not visible unless I'm logged in. As for rep, it tells me i've left 174 reviews and am "level 6" whatever that means. Guess it's not high enough to leave 1 star

There's clearly reviewer tiers in Google Maps, with the lower ones getting shadow banning too negative reviews (I had some of my earlier ones appear after my rank got up)

The higher tiers are more public, like for the 'local guides' for instance: https://maps.google.com/localguides/

Interesting, because I think the people that leave 1 star reviews are typically pretty motivated to do so, especially when it is a single one. Possibly they want to see a balance between positive and negative reviews.

Online restaurant reviews (and doctor reviews) always skew negative because those who have had bad experience are left frustrated and impotent and a scathing online review is one of the few avenues open to them, while those who have had a good experience are unmotivated to write anything.

On a related note, I’ve been getting overwhelmed with “How’d we do?” after the fact online surveys lately - doctor visits, rental car, retail shopping - and I can’t help but wonder if along with actually trying to improve their services, these surveys act to manage the narrative. That if I can complain to them, I won’t complain to Google, yelp or TripAdvisor.

> I can’t help but wonder if along with actually trying to improve their services, these surveys act to manage the narrative.

Absolutely, that's precisely what they are for.

And they direct you to Google/Yelp once the screen you as positive

Friend of mine runs a bar, he loves these 1 stars (does not care for walk-ins, it is a very much regulars only place) and we regularly check them out.

One of the best is a person who wrote: "Shitty service, never been there" and it their only review.

Same thing happened to a few of my reviews on Google Maps - not sure what those few vendors did to get my reviews shadow banned but it was frustrating. I no longer log reviews now since the motivation to contribute to a catch-and-kill operation is low.

Have they left plenty of reviews elsewhere? I wouldn’t be surprised if you have to reach some threshold of reviewing before your score counts.

I have left exactly one review in Google in my entire life, and it was a 1-star review of a car rental place that royally screwed up.

A couple months later I was notified that "x users found this review helpful". So someone must have seen it despite me never reviewing anything else.

Those interactions from other users can be totally fake to boost engagement.

Possible, but it could also be that the reviewing system considers them a high-signal user: very few reviews most/all negative would be a person who doesn’t care for reviewing but had such horrendous experience they were pushed to warn others.

I operate like that, but the other way. I generally only leave very positive reviews when I've had a great experience. I guess I'm fortunate that I've yet to have an experience so bad that I was motivated to leave a 1 star review despite my apathy.

I don’t even run a business, but I constantly get instagram ads for services that promise to delete bad reviews. I wonder if these services are spamming google so much that all flagged comments will get shadow banned until they are manually reviewed (which means never).

I don’t even run a business, but I constantly get instagram ads for services that promise to delete bad reviews

I hear ads for them on the radio all the time.

Broadcast radio, so I know the ads aren't targeted at me.

It seems the lesson here is that if you are unhappy with a place, leave a 2 star review...

This is basically why I pine for the old web with very static code. It came with its own set of issues, but, in a sense, it was a lot easier. Now you have no idea what is happening. Only a big gate keeper does.

In social media, we call this the "bozo mode". If you are bozoed, only you can see your content. A sneaky but very effective moderation technique to keep things civil.

Just guessing, but could it be, that Google is correlating location history with reviews? Would be a good way to weed out fake reviews in many cases.

Happened to me too! I left a company a 5 star review - they were a small company and they deserved it. I even uploaded pictures. Never showed

I had the exact same thing happen to me.

the merchant must be spending enough on Google Adwords to bury bad reviews, just like Yelp

Reply on news.ycombinator.com