Telco Astroturfing Tries To Bring Down Reviews Of Susan Crawford's Book | Hacker News Reader

Telco Astroturfing Tries To Bring Down Reviews Of Susan Crawford's Book

techdirt.com

132 pointscameron91331 comments

A talk on the book in question (as well as audio) can be found here: http://cyber.law.harvard.edu/interactive/events/2012/12/craw... and if you prefer YouTube, here: http://www.youtube.com/watch?v=7R4xhwy-1oI It's a pretty good overview of the American broadband infrastructure problem.

Really great talk, thanks for the link.

Pretty convincing analysis. Many of the one star reviews follow a similar template: "I am a [profession X here]" (a jeweler, a freelance video editor, a professional truck driver, a pharmacy technician, etc.). Weird all of them, usually from rural areas, go out of their way to declare what they do for a living. Not to mention a majority of these "customers"--this is their only review on Amazon at all.

Yeah, I was actually surprised at how generic those reviews were. I mean, I've been reading Amazon reviews for a long time, but this just takes things to a new level.

My personal favourite review:

http://www.amazon.com/review/R1150DW8W7EM1/ref=cm_cr_rdp_per...

This happens with any and everything remotely contentious on Amazon. It's a default soapbox/battleground [0] for the Internet. Though, this certainly appears more organized than the typical ranting.

It's an interesting problem.

I tend to think that Amazon is big enough now that they could require a purchase prior to review without facing a shortage of content. I'm guessing someone over there has calculated that the sale potential to creators and readers of unqualified reviews outweighs any negative impact they have.

Allowing angry reviews against one view is a great sales opportunity for material supporting the opposing view.

0: https://news.ycombinator.com/item?id=6092631

Next time I do gonzo PR, I shall pick a powerful enemy, create an astroturf campaign, and make it look like my enemy did it.

Isn't the author looking at the data to find patterns to then measure it by. I'm no expert, but it seems like he is trying to find patterns to support a preconceived opinion (which I don't necessarily disagree with).

I agree, but a very damning statistic is buried at the end of the article. The vast majority of these 1-star reviewers don't have other reviews on Amazon, which seems very odd, considering the book's subject matter is wonkish.

"The vast majority" is putting it lightly. All but one 1-star reviewer have zero other reviews. And the one 1-star reviewer with multiple reviews is Richard Bennett of ITIF, someone who is verifiably real and biased against the book.

A simple sanity check would be to run the exact same analysis on different books, and compare the book to the population.

It would of course not be definitive, but, for example, if I had seen that only 1% of the books on amazon have this many people mentioning professions on reviews, I'd get suspicious.

It certainly isn't a scientific study but it does bring attention to the expanding breadth of the platforms involved in public discourse. How do you solve the problem of a fabricated public? That aside from the business and technical problems faced by the likes of amazon, facebook, twitter etc. regarding faked personas. I personally find it troubling because it is just these kinds of issues which provide the best fodder for those who view anonymity as a liability.

agree w/ the theme but anonymous != fake.

we can be anonymous and very "real", e.g. a verified reviewer, w/ a public review history etc, but whatever identity he chooses to display to the group.

I personally prefer to call those "pseudonymous" rather than "anonymous". A track record is effectively an uncompressed name: the salient part of anonymity is precisely that it has no useful metadata for the expected audience.

That gets highly academic about what we really mean by anonymity and what purposes we want it to serve, though, which we can get into if you want.

Of course, I prefer to use my real name and for others to do so if possible. I want to push for authenticity and get the problems fixed if possible, but that is another mess.

One of the issues is that there is no such thing as a real name. There's the obvious difficulty: the "real" part, which can be defined any which way and by picking one reality can exclude other valid realities. But there's also the fact that not everyone has a name. See the third comment on http://www.kalzumeus.com/2010/06/17/falsehoods-programmers-b...

One of those possible realities is that "real name" is defined as "legal name" and you'll find a long history of people rejecting that: hacker nicknames, author pen names, rebels' noms de guerre. (A survey in film: The Matrix, "Goodbye, Mr. Anderson" "My name is Neo."; X-Men, "What's your name?" "John." "What's your real name, John?" "Pyro."; Pirates of the Carribean, "Captain. Captain Jack Sparrow.")

Authenticity is a wonderful sounding goal, but it still doesn't provide a very strong comprehension of your goals. So if you get past the problematic concept of a "real name", you still don't have enough to really design for it. For instance, it's arguable that "Yuhong Bao" is less authentic than writing it in Chinese. It's also arguable that it's more authentic.

What's in a name? A name is just a pointer. The functional difference between a real name and a fake name is, in the end, just taste.

This is not the point though. I am talking about how people often can't use their real name all the time because of problems and how I'd like the problems to be fixed if possible.

Since the problems are those of trust, you're really aiming for a perfect trust society. That's sort of hard. I'm working towards that myself.

Yea, I know that some problems can be difficult to solve.

Hmm...go to http://www.amazon.com/Captive-Audience-Telecom-Industry-Mono...

  149 Reviews
  5 star:	 (99)
  4 star:	 (8)
  3 star:	 (4)
  2 star:	 (6)
  1 star:	 (32)

Yes, there's some astroturfing going on. It seems to be at both ends of the scale though. The other possibility besides Telco astroturfing is that some of the negative reviews are left by people who dislike the Obama administration (which Crawford formerly worked for). A negative mention on a conservative blog might have attracted readers with a negative opinion to the Amazon.com page.

The most recent 46 5 star reviews were written after the Techdirt article was posted, and many of them refer to it:

JakeI: The ONE STARS reviews posted here are actually people who WORK for these TELECOM companies and want to keep making fortunes off of misguiding people. Ignore them, report them, and move on.

a reader: Just a note to the paid consultants who have been discovered due to the spike of one star reviews from brand new accounts, I'm buying the book thanks to you.

vge: Susan P. Crawford has written an excellent, 5-star quality book, don't listen to the 1-star telco astroturfers.

Sid Wright: An injustice has been done. This book was given 0 stars by fake reviewers when it actually deserves a real, honest score.

Of course, those 5 star "Astroturfing sucks" reviews are just as useless as the 1 star "I am a rural truck driver and..." reviews, but with the Techdirt article as a cause, at the very least we have an alternate explanation. I see no reason to assume we're dealing with a second astroturfing campaign.

I should add that even if those 46 comments are all legit-but-useless 5 star reviews, then there's still 53 left that were made before then, which I lack the time to look more deeply into. Your conclusion might be accurate for those.

The fake 5 star reviews now handily outnumber the fake 1 star reviews.

Yay?

Do any related laws require disclosure of paid for or fake reviews? Blogs must state financial interest in reviews they do now:

http://internet-law.lawyers.com/Truth-in-Blogging-New-FTC-Bl...

It would seem this type of action would be done (in this case) by lobbyists (paid for by lobbyists), I wonder if there are rules on lobbyists paying for fake reviews (since lobbyists do have some constraints on actions that others don't).

Why does there have to be a law against this? If Amazon wants to allow astroturfing, what say does the government have in it?

As long as the reviews aren't presented as being honest or real. If they are presented as honest or real, it sounds like fraud to me.

If I sell you a cinderblock in a box and tell you it's a computer, I think the government should have a say. Same situation if I sell you a book and tell you that Bob the fisherman from New Orleans thought it contained real original scholarship about Soviet central planning, and it turns out that his name isn't even Bob.

Sure, but that's not what is happening here. Someone not involved in the purchase or sale of the item is making statements about said item.

That's entirely different.

No, Amazon is the publisher. I think they're responsible.

What do you mean no? Amazon isn't writing the reviews, someone else is.

I am not saying there should be. I am not sure I really think the law for blogs is a good idea. I agree it isn't a great thing to do, but there are lots of things that are not great things to do - most of which are probably legal.

I was just asking if there was a law that prohibited it.

I wish for reviews + social. You know, somehow combining the power of the web with the good old process used to find out whether something is good or crap - asking the people you know.

A review by someone 2 links away from me is very unlikely to be spoofed and I reckon that even 3 links away from me (this is probably a large enough pool?) would be quite trustworthy.

Reply on news.ycombinator.com