Case study: What Steemers can learn about "system gaming" from Amazon's Kindle Unlimited.

in steem •  2 years ago

I'm a self published author who's exited the KU system, Amazon's "all you can read" subscription service. All you can read, that is, of authors who participate and allow their books to be borrowed. Hugh Howey made some serious bank as a participant, for example. The catch was, your title had to be exclusive to Amazon for 90 days, and auto-renewed that exclusivity every 90 unless you actively opted out.

In the end, I found that I was in competition with more scammers than authors.

And Amazon was reptilian in its response to the master game players... and when it did respond, it often just made matters worse for us "real authors." 

The fund is a monthly amount, divided by the number of borrows. So, say around $11m divided by the original "honest" borrows could make an author a tidy sum, and it was worth it to help Amazon destroy its competition. 

When the program started, every book received a flat borrow fee, if it was more than 10% read. Your Kindle data is fed back to Amazon, surprise, where your borrow was flagged as "read" when the reader reached that point. The original funding for the program was a gravy train - $1.50 or so a borrow.

Great news, right? 

Well, until someone had the bright idea of publishing 10 page books. The minute someone opened it, guess what? The author got $1.50. So those of us with 200 or 300 page books were in competition with those who wrote ten page books, and we saw our $1.50 become $1.40, $1.30...

So Amazon changed the system (with two weeks' notice to authors, who suddenly had to scramble to fly blind for two months, wondering how drastically their income would change). Instead, now authors would be credited only for "pages read." Sounds good again, right? Fair and honest, even if it was only around a half penny per page. A 300 page book  could still get $1.50 if it was fully read.

Until the scammers found a new way to drain the fund.

Someone had the bright idea of publishing 10,000 page books (Amazon's outer limit), and putting a link on the first page that said "CLICK HERE TO WIN A FREE KINDLE!" That "click" took you to page 10,000. And... Amazon had not anticipated this, and had nothing in its software to determine actual pages read, only the last page read. So, someone who published a "paranormal romance" wherein the first ten pages were paranormal romance and the rest was padded out with clickbait on the South Beach Diet or Time Management, suddenly got a borrow payout of... $50 per borrow. 

Many of us were left wondering how on earth our per page borrow rate kept going down each month, and why in January of this year, it took a nose dive. Then, one of the scammers publicly bragged about how he made $60,000 in one month by pubbing inflated page books.

I left the program, despite being a selfpub ebook romance author, and Amazon having about 75% of that market, because I couldn't take the stress, or the anger, of seeing people who were better at scamming consistently earn more than I could ... and, more importantly, get away with it. They got paid, and Amazon's only change has been to limit book size to... 3,000 pages. And to take down scam books, but... only after they're reported by authors or readers.

My point is...

It's critical that Steemit self-police, and,

Those who find plagiarism should be rewarded with bounties, commensurate with the $ saved for the community.

In this post by @tuck-fheman:

and his follow up post here:

He showed that he did a lot of work to reveal that a profitable post (allegedly original content, first person heroin addiction story) that had been a scam.

But, because someone else posted a callout on the scam later, that person earned $5k whereas @tuck only got $80.

A Modest  Proposal

Using #scam as a tag that a moderator, or the community, monitors for scam output. And, based on timestamping (objective) and accuracy (subjective), the whales and or the community reward the person who did the hard work of unveiling a system abuser.

Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE STEEM!
Sort Order:  

This sounds like a great proposal.. I have seen a lot of articles that were literal reposts from news sites..
I have no problem with people posting links.. Then it is at least clear that they just want to redirect you to some other page.


Yeah, it was easier on Amazon to find scam books than it would be here to discover plagiarism/scamming; most of them came from a handful of people who also padded the book openings (which you could see via the Look Inside feature on the product page) with the same garbled English "disclaimer." All the more reason I think to reward people who do the hard work of catching those who are really gaming the system hard for $$$.

This post should have, by all rights, been much more popular.


Yeah it's discouraging.

Thanks for explaining what sounds like the disaster of Kindle Unlimited. Game theory has been used extensively to incentivise users of this platform to protect and grow its value. Not so much on content - which I would argue is much more difficult. But being blockchain based, I think it is going to be much more feasible to build the sort of contracts required to minimise content abuse. I think the #scam idea is terrific.


Thanks, it's definitely a poorly designed and managed program. Is the game theory background in the white paper? Or is there somewhere I could read more about it? My hope is that the harder it is to fool the system, the faster quick buck artists will move on.


The white paper doesn't even mention game theory, but if you know some game theory you can see its implementation. Game theory is a full subject on its own, but there are plenty of resources on the web to find. A recent set of posts on this site also provide an introduction: