Hacker News Reader

Comment by matsemann | Hacker News Reader

> There is literally a poorly formatted pirated version of my book on the Amazon Kindle store accumulating 1 and 2 star reviews for its poor formatting that count against all editions.

So not only do they commingle physical items so you never know what you get. They also allow other people's products to affect your rating.

View on HN

That struck me as especially egregious; even if the e-book-version was legitimate, why on earth would anyone think that merging these reviews is a good idea?

Some tech books are, in my opinion, terrible for e-book without adjustments and vice versa (e.g., description on left page and corresponding picture on right page might work quite well in a paper book but not in an ebook where you see only one page at a time); why would these ever be merged?

Is this a normal thing on Amazon? Speaking as someone who never ordered anything from Amazon as it barely services my country

I don't work at Amazon, but worked at a similar (European, national) retailer:

You really have to walk the line between ratings-per-SKU and aggregate ratings sometimes.

In an ideal world, the reading experience of a Kindle book is the same as the physical book. So someone might argue that ratings should be aggregated across all version of the product. Would you expect different ratings for hardcover/paperback versions of books? Probably not, at least assuming people review the content of the book, not its appearance.

Similarly, do you really want people to leave ratings on a black shirt in size L, and that rating not apply to all sizes and colors?

I assume someone at Amazon thinks about this, and the defaults (fashion => merge, books => merge, electronics => donotmerge) etc. do OK until complaints come in. There's no profit in not being heavy-handed.

I expect a review of a book to be a review of the book. The contents of the book are only one aspect of the book. The digital or physical form of the book is no less important to me when I am looking to buy the book, particularly when the digital or physical form of the book might very well be irredeemable trash (rendering the quality of the contents irrelevant.)

> Would you expect different ratings for hardcover/paperback versions of books? Probably not, at least assuming people review the content of the book, not its appearance.

I would expect different ratings, because when you buy a book it is not just the content you are buying. The reason people look at reviews is to decide whether or not to buy, and the physical quality of the book is important for that.

If the hardback is fine but the paperback has pages out of order and that fall out when you turn them, then people who bought the paperback are going to (quite justifiably) be leaving reviews with the lowest rating possible. This doesn't help me at all when I'm considering buying the hardback.

Well, people can explain their low ratings. "A real page turner but the paperback printing and binding were awful." The problem is that, if you rate at the SKU level, you now have different reviews for every size and color of shirt. "The green color isn't at all like what was on the product page."

The reality is that for most books (pretty much all fiction, most trade paperbacks without a lot of illustrations/code/etc.), Kindle format is just fine even on a e-paper reader.

> "The green color isn't at all like what was on the product page."

That's a valuable review though; if a green shirt is not as advertised I want to know that before I buy it.

Yes, but it's pretty much as valuable lumped in with the other shirt reviews whereas if you split out reviews for each individual SKU, many of the items will end up not having reviews at all.

Amazon could simply have aggregated star totals and reviews, with the ability to select narrower subsets.

If I'm looking at a book and worried about quality differences between formats I could ask ask for just reviews from verified paperback purchasers, then from verified hardback purchasers, and then from verified Kindle purchasers.

They already allow filtering by verified purchasers. Just add the format filtering, and when viewing through filters show the aggregate star stats for the resulting reviews.

This would be close to trivial for them to do and greatly increase the usefulness of reviews with no effect on people who are OK with the current aggregate approach.

If a particular variation has no review, I want to know that before I buy it. What good is a bunch of 5 star reviews for a red shirt if the lime green shirt I want to buy is actually the color of baby shit?

> In an ideal world, the reading experience of a Kindle book is the same as the physical book. So someone might argue that ratings should be aggregated across all version of the product. Would you expect different ratings for hardcover/paperback versions of books? Probably not, at least assuming people review the content of the book, not its appearance.

Right, the issue though is that there is a much larger gulf between the eBook and print book than there is between the paperback and hardcover. And the delivery mechanism can greatly affect your enjoyment of the book. It's about more than just appearance. Source code in particular, does not often work well on a Kindle's tiny screen.

Although maybe it does if you read it on a full-size tablet. In general, I mostly avoid buying books I know aren't 95% text in Kindle format unless they're very cheap or are from a known publisher for things like guidebooks. They can work well but it's definitely hit or miss. (I also just don't like ebooks for certain types of references where I want to stick post-its etc.)

I personally fully agree!

But I imagine a PM at Amazon was presented with the question "Should ratings of all versions of books be merged?".

Unless they really really cared (or personally had experienced dramatic differences between versions of books before), they probably said "Yeah sure".

(I'm also biased a bit in that I assume that implementing this more fine-grained would cost more than $0).

It seems like there's room for compromise. You could simply assign a lower weight to reviews for a different size/format. Or you could split ratings entirely if (and only if) the review averages are significantly different.

To play devil's advocate, they should be merged because they're mostly the same product. If I'm contemplating buying a paperback novel, I might read criticism from reviewers who only had the hardback edition. Maybe occasionally this will make me buy some books with poor-quality printing and binding, but mostly I'm interested in the content of the book. And if I did care about the presentation, I might be better off reading reviews for different paperback books by the same publisher.

In the same way, an e-book mostly has the same content as its paper source, and most of the time the publication process will be of an acceptable quality and what I care about is the content.

These assumptions all break down once other publishers enter the market though. If the work is public domain, then finding reviews of the general content is likely much easier and less useful than a review of how this edition has been put together.

From the earliest days, Amazon apparently made the decision not to separate reviews of content from reviews of the delivery of that content, e.g. award-winning films with a poorly-made DVD.

And, yes, it can be an issue with Kindle. A lot of books with illustrations, sidebars, footnotes, etc. (i.e. not simply free-flowing text) get designed for print and are sometimes just automatically converted to ebooks without much care. It's a bit of a dice roll.

I once bought a Kindle version of a book that was originally published before Kindle existed, and the publisher never produced an ebook version. I bought the Kindle version because the print version was very hard to find, and the copies available were outrageously expensive, in some cases 10 times the original price of the hardcopy book. The Kindle version, which I assumed at the time had been automatically produced by Amazon, was horrible, completely unreadable. I've never bought another Kindle version of an older textbook.

You used to be able to at least hit the link to see all reviews. The all reviews filtering options included one to filter by format. This helped a lot.

I don't see that anymore.

Also, on the reviews themselves it used to state what format a given review was for. I don't see that anymore either.

So...it looks like Amazon deliberately decided to make it harder to tell if the Kindle edition is good or not.

Review filtering is still available for many products, see my note at https://news.ycombinator.com/item?id=24400210 for how to get there.

Unfortunately it's not available for this book. (You can filter by star rating but not by edition.)

Reply on news.ycombinator.com