Goodreads asks users to help combat 'review bombing'
npr.org
npr.org
They show a graph of positive/negative reviews over time, making review bombing stand out.
If Valve identifies the anomaly as being off-topic they can even mark it as such an exclude the reviews from affecting the overall review scores. Users who don't like this can even disable the exclusion and get the "raw" review score.
More info from when this was introduced: https://steamcommunity.com/games/593110/announcements/detail...
Things like this is so nice. If I don't trust Gaben's bullshit review filter, I can see what the result would be without, just to check.
Reviews are a dime a dozen, positive or negative; they normally aren't useful for judging whether a product is worth buying.
Review bombings on the other hand only happen when something happened to trigger the internet's sweaty wrath, and that is very useful as a unit of measure. Any particular review in the bombing is worthless like any other review, but the collective act of review bombing is very valuable information.
Reviews by themselves don't tell me anything important, especially in times of peace. Review bombings on the other hand tell me there is or was a problem I might want to look into before I open my wallet, and that's very valuable information for making informed purchases.
would you mind qualifying that? I'm not familiar with what "woke" actually means
This could cover almost any social, economic or political issue. Generally though it was used around issues related to race, gender or sexuality. People would sometimes use the phrase "Stay Woke" as a shorthand for something like "Stay aware of racial or gender disparities and work to counter them".
The term was subsequently used in a derisive, pejorative or mocking manner by people who perceived the first group (the "woke" people) to be factually incorrect or excessive in their demands.
People using a phrase like "woke nonsense" are expressing disdain for inserting race/gender/sexuality issues where they don't belong or being frustrated by too-extreme measures being advocated to address such issues.
There's my attempt at writing a very neutral etymology of the usage of the term.
But I wouldn't have that chance without the review bombing telling me there's something I might want to check first. Developers and publishers are very good (and incentivized) at hiding the problems review bombings can otherwise reveal.
Goodreads really has only one thing going for it: a pristine data set of books, authors, cover images, etc. There are so many other reading apps with a better UX but they are frustrating to use because the data is spotty or low quality.
And spoilers in book titles (the goodread ones, not the physical books)
Just a quick search for "a shocking twist" on goodreads finds many many books:
https://www.goodreads.com/book/show/198614844-the-perfect-fa...
https://www.goodreads.com/book/show/137436113-1800198272-978... (i have no idea who/what inputed the title here...) and how "unputdownable" is a word...
https://www.goodreads.com/book/show/30514112-the-sister?from... ... i'll see the twist coming, you just told me about it!
...
Great, now I know that the killer (or whatever) they catch right before the end isn't the real one, and that there will be a twist in the last few pages.
For The Sister I googled to see its cover images.
I personally cannot stand the former where publishing teams will buy or fake mass amounts of reviews to play into the NYT bestseller formulae.
It is fairly easy to tell an ad hominem in a "review bomb" in my experience. Usually it relates around politics or something controversial rather than the actual thing itself.
The harder we police against fake positive reviews, the more we incentivize this sort of anticompetitive sabotage.
Accounts need to verify themselves somehow.
Accounts should be old enough before contributing reviews.
Accounts should only be able to review X amount of titles an hour/day/etc.
Moderators should have tools to identify influx of reviews and decide mitigation strategies.
Moderators should be able to put up a banner on a title that may be review bombed so the community can report suspicious reviews.
Etc.
These aren't honestly policing much. Just general common sense for these platforms. Also heavily needed on many other platforms that get spam in similar ways.
Maybe you decide that sounds too difficult, so they should stop banning listings for that and focus on policing the reviews themselves. But that actually turns out to be a very difficult problem in itself! Sellers of crappy products will often include notes that induce/bribe the customers into leaving fake reviews. A seller of cheap products can rack up thousands of verified purchaser fake reviews this way.
They can also do things like hire a botnet to purchase the product in bulk using thousands of hacked PCs all over the world and get those PCs to review the product, generating more fake verified purchaser reviews.
In the end, if you don’t ban sellers for fake positive reviews then you run into these tactics and it’s very hard to police individual reviews. If you do decide to ban sellers then you have the opposite problem of dealing with sabotage.
As for detecting “influxes”, that problem is far from trivial. If a review bomber is using a botnet of thousands of PCs then it’s quite trivial to stagger the timing of the fake reviews over a period of weeks. It’s not like they’re forced to push a button and have them all hit the review page at the same time.
You could still record unverified reviews, but hide them and their ratings by default until the user explicitly wants to see them.
I think it is if the product itself is cheap enough. A $10 product where you get a free $10 gift certificate is essentially free. I think lots of people would be willing to leave a quick 5 star review in exchange for getting the thing they just bought for free, even if the product is kinda crappy. As long as it’s not totally broken and useless!
I doubt it would cost a lot of money from Amazons perspective.
But yeah I agree Amazons incentives are not the same as their customers. They obviously have a lot of insight into what drives and lowers sales.
Maybe, although I seriously doubt it, at some point they will see enough reduction in sales to take action against poor quality products and their mispresentative reviews.
I’m sure many reviewers will enjoy having a special badge against their name. Not all, sure. We don’t need everyone’s reviews in this situation.
Proof of purchase would require some modest effort on the part of publishers to create a workable and shared system, but it should otherwise be trivial to get going. Publishers have a strong motivation to opt-in since fake reviews are probably more detrimental than not to their bottom line.
That person could no longer write verified reviews. Maybe it’s a permanent ban, maybe it’s a 1-5 year ban. They wouldn’t be able to verify another account with their identity until the ban is listed.
Yeah Steam is good, but mostly because it is a very closed loop system as they own the entire transaction-review-lifecycle. Not soo easy with other things like books.
As for the individual reviewers, there are tens of thousands or millions of these, and review bombers can trivially cycle these in and out, so no individual reviewer is a “serial fake reviewer.”
This is really just a special case of the wider issue of fraud. And I'm fairly convinced of the argument that the optimal amount is nonzero [1]. A further wrinkle to this problem is that fraud tends to only get more sophisticated over time, so the costs of fighting it always go up. At some point we may reach a scenario where fraud is so rampant that entire business models become non-viable.
Although really the current concept of "review" is terrible. Few reviews are detailed and specific enough and multi-dimensional enough. They cannot be searched well enough (Although we might be nearly there with LLMs - asking them to critique a book from our point of view and preferences.) It's hard to follow reviewers that you have learned to trust (hard enough to notice them in the first place). Star ratings are not useful in the intended way and only a few sites ask for multi-dimensional star ratings (generally asking to rate irrelevant things :-) Bleh!
I have also heard that it is a backdoor way of spending political campaign money. Candidate "writes" a book, the Campaign to Elect X buys a warehouse of the books, candidate then gets the profits.
But seriously, I have a similar feeling about book deals. That is, it's a great - and legal - way to pay (off) someone for "prior service", meanwhile the book's sales are shite. But selling books was never really the point.
Separating account verification and restricting the amount of reviews allowed - separated from privacy-respecting reviews might be a direction. But currently it's hard to trust anyone's "privacy-respecting" features.
(My impression is that in consequence, currently, "verified" means "mostly positive".)
If a thread started for ‘best X of 2023’, I’d be pretty sure of the results.
Why is it important that you, a person apparently unwilling (hyperbole aside) to prove that you have actually read a particular book, should be able to transmit (ostensibly) your opinions about said book into the heads of other potential readers via a software platform? A small number of verified reviews is almost certainly better than a large number of unverified reviews, for exactly the reason highlighted by OP's article: when there are few or no controls on who or what can post however many reviews, the system is inevitably exploited by bad actors, and said actors often operate at scales that dwarf the impact of a single legitimate reviewer.
Probably you could even generate new questions every time with GPT, so that people couldn't cheat. This system would have a million problems, sure, but would it have more problems than the current system?
For example, Facebook/Instagram/OpenAI/etc. requires non-VOIP phone number verification on signup. This alone would cut spam to near-zero.
I’m sure if Goodreads required this on signup, the problem would solve itself.
Some of the problems with this proposal
1- Books borrowed from friends or libraries 2- Book bought from different platforms or stores 3- Books downloaded from from the internet legally (or illegally)
Most of my readings are not books that amazon can verify I actually owned the books and I usually write reviews if I want to share it with my friends on Goodreads. That would be hard to implement and if so, I think personally I will stop using Goodreads, probably with most of the people I know.
https://janefriedman.com/i-would-rather-see-my-books-pirated...
Instead they prioritize having as many reviews as possible - when possibly a majority of them are fake.
And the cost is - the reviews are completely meaningless.
All they care about is engagement - and if users have to dig through reviews manually trying to sort out signal from noise - that's engagement to them, not an awful product.
https://www.amazon.com/Bill-Dave-Hewlett-Packard-Greatest/dp...
might look fake, since I borrowed it from the library.
It's a way to force the fraudsters to buy from Amazon.
Don't fix the problem - make more money from it. That's what companies do.
If costs were no object they could simply hire well-paid, trustworthy, professional reviewers to review every product and listing. These high quality reviews could establish their site as the gold standard, strengthening the brand (think Apple levels of brand value) and drawing even more users. But all of this could be so expensive they’d actually lose money for doing it. So they don’t.
But you, the author, have to pay them. When I checked, it was $450.
If you're an author, you also get a deluge of offers to write a review. I didn't follow up on any of these, but you'd assume that they couldn't get any repeat business if they panned your book.
Amazon has a Verified Purchase label, but still too much noise.
For goodreads they'd need a different system (reputation based? helpfulness?)
I'd rather have a much smaller set of verified reviews than a bunch of fake ones. Decreasing the quantity in this case often means improving the quality and signal to noise ratio.
It would be thought of as an advertising expense to buy books from the platform in order to qualify to write some of the first reviews.
Even though it's an imperfect barrier to review quality, at least it is a barrier versus none, and time and time again retailers who've experimented with it have found it a useful signal.
Yes, you'll lose out on the borrowers who fell in love with a title, but they can always spend the $20 to buy that book to review it if they really want to (I've done that for things I loved, just to leave a review). But even if they don't, it's still usually a net win: you lose not just the honest reviewers but the 100x spam that outnumber them.
When shopping for low-cost products these days, it's not the individual review that matters anymore, just the aggregate ratings. So it doesn't really matter if a few borrowers don't get to leave their rating, it's still more useful for buyers to cut out the spam.
----------
Maybe a simple compromise is to just allow filtering (and adjust the rating) by "verified purchasers only", so you can see both the unfiltered and filtered scores. I'd just leave out the unverified ones altogether, myself, but others could see the raw scores and reviews if they wanted to.
There is a difference between reviews on Amazon, where limiting to purchaser makes sense, and Goodreads that is open to everyone.
And yes, agreed that Goodreads needs a different system. I was responding to a post about Amazon, and even there mentioned that goodreads would need something different, maybe reputation based or whatever.
Not sure it would really improve the signal to noise ratio.
https://www.ftc.gov/about-ftc/bureaus-offices/bureau-consume...
I don’t believe the main problem is that they don’t know how to identify fraud, I think the main problem is they don’t have any real incentive to.
Look: the FTC is already sending notice that enforcement is coming: https://www.ftc.gov/news-events/news/press-releases/2021/10/...
Because at the end of the day this is fraud?
That combined with moderation is a pretty good system.
I think publishers and authors want to free ride such platforms to reduce the marketing spend but then they face trolling and there is nothing to do but cry (but the monetary effect is unclear).
Instead they could invest more into multichannel ads and build alternative platforms for the readers to check what the book has inside. But yeah lets depend on amazon and just wait until it will start asking for money to mitigate review bombing ;)
Besides who reads the new titles when there are thousands of amazing time tested books?
> Corrain's downfall came after internet sleuths published a Google document detailing a number of Goodreads accounts praising Crown of Starlight and giving low reviews to works by other writers, many of them people of color.
That does strike me as the author being aware of her own choices and 'getting ahead' of the story.
I can't find any way on the site to ask questions, either of the developers or of the community. What I'm trying to figure out is the semantics of their reading preferences exclusions.
The actual question asks about "types of content that you never want to read about", and that "about" confuses me. Suppose that I choose "racism" there. Does this indicate that I never want to read anything that somebody has tagged as racist (which I think is probably the case, but is less than useful because it seems like darn near everything gets tagged with that one way or another), or is the "about" saying that I don't want to read any meta-discussion of racism, that is, I don't want any books that are "about" racism - which is absolutely the case.
I can see a "contact us" button both on mobile (in the hamburger menu, top right) and on desktop (bottom right corner).
Right now it seems to be a bit too sensitive and hides books a bit too eagerly, but the team is aware of that. That's about the extend of info I have though
I asked all my friends, all my contacts, no one would review it!
So, it's frustrating.
I actually get great book recommendations there via "Personally recommended" when I go to fetch something new to read. Far better than anything commercial.
While sites like Netflix use all the AI they can muster to tune and individualize their recommendation systems, the review systems we know best rely on straight democratic votes that can be gamed. Or worse, they rig the votes for profit.
The AI objective function should be to synthesize advice relevant to each user. As a component of this, treat each reviewer as a tennis tournament, to develop a tennis rating system summarizing reviews? One could apply weights tuned to each user for the relevance of each other user's reviews. If there was a shadow economy behind these uses, as if a hedge fund had to pay pennies to include each voter, the "review bomb" voter's inputs would quickly become useless, and be disregarded.
What we have instead is singularly dumb. Is there room for a startup here? Google page rank was obvious in hindsight. Won't this too be obvious in hindsight?
I don't quite see how that applies? Those are straight up/down/tie. And not subjective at all, like book quality.
While on first glance it might seem useful to have the nuance of being able to award between 1 and 5 stars, those intermediate star rankings tend to mean different things to different people, so a 2-star rating from one person might mean something fairly different coming from someone else.
Also consider how Uber/Lyft ratings are decidedly nonlinear. Anything below 5 stars -- even 4 stars -- essentially means there was a fairly large problem, and drivers will get "fired" by the platform well before their average even drops to 4.0. If any platform would benefit from a simple thumbs up/down rating system, I think it's that one.
> And not subjective at all, like book quality.
Certainly in chess and tennis they're not subjective, but a simple thumbs up/down just means "I liked this" or "I didn't like this". Sure, if you only have a handful of reviews, that maybe doesn't tell you much, but as you accumulate more, the average starts being more useful. And of course this doesn't preclude the option of allowing people to write something in addition to the binary choice.
I was opening up the question of how different models for aggregating reviews would be naturally immune to review bombing, using Chess and Tennis ranking systems as inspiration. There's nothing about that style of aggregation that locks one into a particular kind of base signal.
If you know someone IRL, you don't magically have something in common with them that you didn't before. A stranger might be born in a foreign country so we don't have that in common. After I get to know them and become friends IRL, we're still not going to have that in common. So the idea that you don't have anything in common with people you don't know seems weird to me because of course you have things in common, you just haven't discovered them yet.
Is a book review by John Updike at all useful to me? I didn't go to Harvard and don't have that in common with him, so why would I value his book review?
Do the negative reviews focus on SJW topic I don’t care about? Do the negative reviews focus on opinions or do they focus on evaluating expertise/facts? do the positive reviews focus on world-building, writing, quality, or characters; what do they like about it? What do they compare it to?
Sometimes I do this when I’m 10 or 20% into a book and I’m trying to evaluate if I care enough to continue. I’ll see if other reviewers say the book picks up or slows down etc.
How this is any different from their poor handling of fake reviews on all the Chinese garbage that floods their normal sales I don't know.
I'm surprised Amazon is actually doing something, anything, related to Goodreads, including 'asking users'. The website feels abandoned.
What does that mean?
They’ve admitted this publicly. https://amp.theguardian.com/books/2023/dec/13/cait-corrain-p...
> “I boosted the rating of my book, bombed the ratings of several fellow debut authors, and left reviews that ranged from kind of mean to downright abusive,” she tweeted.
> Many of the books given one-star ratings were written by people of colour
That is not 'specifically targeting.' Why do you have to make this about race?
> They posted a 31-page Google document containing screenshots of the activity of a number of accounts, with usernames including “Chantal B” and “Oh Se-Young”, suspected to have been created by Corrain.
> “There’s something extra despicable about using clearly POC [people of colour] names in the fake accounts to upvote every negative review on POC books so the top ones are all 1 star and 2 star. Like what in the yellow face,” Zhao tweeted. Many of the accounts linked on the document Zhao shared have now been deleted.
AIUI, authors in the subgenre are disproportionately non-white. If that's the case, it would be expected for those targeted to be similarly non-white.
None of this is a defense of the offender: she's definitely a big jerk. But it doesn't necessarily mean that we have to make every freakin' issue about race.
You have a lot more confidence on a bunch of people doing OSS "research" than I do. The quote itself says "suspected". I would go as far as to say that believing this was racially motivated (rather than just a scummy move) based on this "evidence" is weird, but not surprising.
Interesting. I would have guessed that Chantal is a 90% black name, but I was wrong. It is only slightly more common for black women, and there are many whites with the name too. This may be a regional distinction.
- The subgenre this all happened in has a very different makeup than western publishing broadly, and
- The authors targeted were not all 'people of colour'.
I agree she was likely secondarily racially motivated in some way - maybe due to trust in that subgenre, or the view in some area of the internet that 'people of colour' cannot be questioned. Or maybe she's just a racist killing two birds with one stone. She seems a serial liar, so I doubt we'll ever know for sure.
But now we have the internet. The assholes can see you, touch you, talk to you, even decide what you are allowed to say.
This is a mark against the internet.
The Block & Mute are what makes them go away.
I would say the direction is the opposite. Assholes dont seek people out, people seek out the assholes.
In the real world, people dont walk up to strangers and ask them to disclose their most offensive opinion. However, on the internet people will scroll through pages of agreeable comments looking for something disagreeable.
The same is basically true for reviews too. The physical world default is you dont know what people think. The internet allows people to know what others think, including their dumb opinions, swings, and trends.
I thought that I was the only person who does this.
Therefore, given a forum full of people, the chance that you will offend somebody is increased. It becomes less "say something offensive" and more "say something merely noteworthy".
We get a “The nail that sticks out gets hammered down” situation. On the scale of the whole internet.
The internet becomes a huge engine for imposing conformity.
Maybe some people with a good understanding and able to produce a quality criticism about a book.
We could even nickname them "critics"?