Camille LaGuire made a good comment on the Passive Voice. The post was another one about the never-ending paid-review scandal, but LaGuire points out that "simple group behavior" can trip up an algorithm, too:
Let’s say there is a large forum frequented by authors who are all interested in promoting their books, along with some book bloggers who are into the same culture....
They all review more than your average reader. And everybody who reads their books and interacts with them on blogs or elsewhere hears again and again how important reviews are to authors, so they also have a “bubble” in their reviewing behavior. They also all submit to the same book bloggers. And they all have an overlapping readership, and even though they avoid mutual reviewing… the authors and their fans tend to read a lot of books from other authors in the same forum. And so their reviews are clustered in the same pool.
From the algorithm’s standpoint, it sure looks like a mutual admiration society, and in some ways it is. It’s not intentional, but people are using leverage to get an unnatural number of reviews, and the reviews are created with a different pre-conscious agenda than most reviews are.
And this pattern shows up really obviously in an algorithm....
Your best bet is to not to work against what the goals of the algorithm are. The goals of the algorithm is to NOT favor one book over another but to make every book equally available to the people who would most want it. Therefore, the best way to work with the algorithm is to work on good labeling, appropriate covers, titles, blurbs — and, of course, good content.
Or alternatively, you can just keep coming up with new leverage strategies when Amazon cuts off the old ones. That’s perfectly legit. Just don’t be all surprised when Amazon cuts those off too.
I liked this because I think there's a temptation for indies to revert to the clubby sort of reviewing that marks a lot of traditional publishing--after all, that's what we know and what appears to have worked for them.
But the clubbiness of that world actually limits the usefulness of those reviews--a lot of people don't bother with, say, The New York Times book reviews because they know that paper only reviews certain kinds of books, so if you like, say, potboilers or romances or erotica, you'll never find anything useful there. Amazon works as a retail outlet because it's good at getting things in front of people that they actually want--it doesn't worry about who's in the club, it just offers up the goods. And readers have clearly responded quite favorably to that, which benefits us all.