Amazon Killed The Book Reviewer Star

[kennyc]

Amazon Killed The Book Reviewer Star
Gregory Ferenstein

Authors no longer have to impress stodgy English majors to get their book a quality review: new research from the Harvard Business Review shows that the aggregate rating of Amazon reviewers are every bit as good as professional book critics.

Professional book critics, on the other hand, suffer from nepotism: critics give more favorable reviews to their colleagues, authors who agree with their ideological slant, and if the book has been given an award by other critics. The result, implies this new research, is that Amazon has democratized the book reviewing process, with consumer reviewers less beholden to special interests and more representative of the book-reading masses. Perhaps most importantly, it rebuts critics who have claimed that Amazon is nothing more than a cauldron of corrupt and uneducated opinions.

Despite the strict editorial firewall between writers and commercial interests, “reviewers may not always have the incentive to provide objective reviews,” explain Professors Dobrescu, Luca and Motta in a new study of the professional book review industry. Newspapers and magazines are 25% more likely to offer a review of an author who has written for their publication before; unsurprisingly, the reviews are slightly more positive. Moreover, professional reviews suffer from self-congratulatory institutional nepotism: novice authors get slammed more often than established ones, especially if they haven’t won any awards.

The new research provides ample firepower against academic critics of consumer reviews, who say that Amazon is a circus of corrupt and uneducated reviewers.

“The democratization of reviewing is synonymous with the decay of reviewing,” lamented Professor of English Morris Dickstein, “The professional reviewer, who has a literary identity, who had to meet some editor’s exacting standard, has effectively been replaced by the Amazon reviewer, the paying customer, at times ingenious, assiduous, and highly motivated, more often banal, obtuse, and blankly opinionated.”

Others have implied that Amazon contains far worse than uncritical literary buffoons; Cornell professor Trevor Pinch, discovered systemic corruption within the ranks of top 1,000 Amazon reviewers, many of whom are given perks for good reviews or abstaining from bad ones.

But, if Amazon really is a literary cesspool, why did Dobrescu and his colleagues find that consumer reviews were nearly identical, on average, to professional critics, (under conditions when professionals would not be biased)? The likely explanation is what social scientists call the “wisdom of crowds.” A randomly selected consumer reviewer is no match for a professional reviewer, but the average opinion of all laymen is less biased than an expert.
....

more here:
http://techcrunch.com/2012/05/15/ama...reviewer-star/

[fjtorres]

In short: relying on Amazon reviewers (hijinks and all) is *no worse* than relying on annointed "critics". And possibly better.
(shrug)
Not unreasonable...
...as long as the book has a reasonable number of reviews.

That is one of the "virtues" of the Prime Lending Library and promotional freebies: increased reads that might lead to added reviews and thus greater legitimacy for future buyers.

[din155]

So following the reviews by masses is no worse than the opinion of few independent reviewers. All that was missing was the medium for masses to provide their reviews. Now if there was a way to pool these reviews from various sources and present them at a single place, we might get close to an unbiased review for most of the products.

din

[xg4bx]

"Others have implied that Amazon contains far worse than uncritical literary buffoons; Cornell professor Trevor Pinch, discovered systemic corruption within the ranks of top 1,000 Amazon reviewers, many of whom are given perks for good reviews or abstaining from bad ones."

2 words: harriet klausner.

[Ninjalawyer]

Welcome to the crowdscourced future where a crowd of random strangers can be, in aggregate, much better than any single expert. It's interesting that the Amazon crowdsourced reviews are as good as expert reviews in conditions where the experts couldn't be biased; does that suggest that overall the Amazon reviews would be better as their would be more chance for experts to be biased in the real world?

I'm a big believer in crowdsourcing. Wikipedia is the go to example (offering a breadth and depth of topic coverage that no expert created encyclopedia could hope to match), but there's also Goldcorp. that made millions by putting its geological data up for public view and asking people to take a look. People often complain about Wikipedia's accuracy, but I've never had an issue, and it's not like the physical encyclopedias I had as a kid were particularly error (or blatant racism) free.

[GlenBarrington]

in their influence by people within their respective industries. I can assure you that people in Canton, Ohio, or Peoria, Arizona don't pay nearly the level of attention to reviews, good or bad, nearly as closely as the media outlets fantasize.

The opinion of friends and neighbors have always counted for far more.

[Joykins]

Did the article mention the good-reviews-for-hire that are rife in niche/indie/self-pub books? The wisdom of crowds there just aren't.

[Fbone]

So did the Harvard study somehow find a way to remove all the "corrupt" reviews? And then match the rest with the professional critics?

Many titles have almost all (>90%) fake reviews.

[AnemicOak]

Originally Posted by Fbone

Many titles have almost all (>90%) fake reviews.

Yep, I don't know about >90%, but authors can buy positive reviews from people on places like Fiverr (not just at Amazon).

[kennyc]

Originally Posted by Fbone

So did the Harvard study somehow find a way to remove all the "corrupt" reviews? And then match the rest with the professional critics?

Many titles have almost all (>90%) fake reviews.

And you have proof of that claim?

[kennyc]

Originally Posted by Joykins

Did the article mention the good-reviews-for-hire that are rife in niche/indie/self-pub books? The wisdom of crowds there just aren't.

As well as the spammed 1 star reviews?

[carpetmojo]

Originally Posted by Ninjalawyer

Welcome to the crowdscourced future where a crowd of random strangers can be, in aggregate, much better than any single expert....................I'm a big believer in crowdsourcing.

And so you jolly well should, Ninja - bein' a lawyer !

It's not far off the same principle as trail by jury, really.......

[Fbone]

Originally Posted by kennyc

And you have proof of that claim?

Almost every indie/small press book I sample.

Here's a random example where the same person wrote multiple reviews under different names for this title. Whether intentional or not it still affects the ratings.
http://www.amazon.com/Chop-Darby-Sta...7185703&sr=1-1

Added:

Another example. If a book averages 3 stars, it may be an indication of 1 and 5 star reviews. Some reviews may be authentic and independent but the impression is not good.
http://www.amazon.com/Into-the-Basem...7185974&sr=1-1

[kennyc]

I'm not looking for examples, you claimed

"Many titles have almost all (>90%) fake reviews. "

I don't believe it unless you have facts to support that statement.

And first define what "Many" means.

[HarryT]

Originally Posted by Ninjalawyer

Welcome to the crowdscourced future where a crowd of random strangers can be, in aggregate, much better than any single expert.

That depends on the specific situation. I would submit to you that it would perhaps be unwise to have papers submitted to scientific journals peer-reviewed by a crowd of random strangers, for example.