Quote:
Originally Posted by BWinmill
Excluding certain books may be a good idea if you're compiling a list for their general readership. Simply put, you don't want one segment of the market to dominate the list if they are particularly devout in their reading.
|
Why? Isn't this contrary to transparency and meritocracy, assuming this is the kind of list that I want?
For example, for the USA Today's Best-Seller list (for Fantasy and Science Fiction), I'm not happy that the first six slots are occupied by a Song of Ice and Fire-related books (
http://books.usatoday.com/list/index?f=10,0,122), but if I want to know what the actually best-sellers are, that's the list I look forward to.
If I wanted an editorialized list, I'll look for an editorial list (over the years, people have written several "best books of -insert year here-".
I have qualms when a "best-seller" list isn't actually based on best-sellers. Or at the very least, explains their methodology to a reasonable extent so I know where and what the biases are.
Quote:
|
Likewise, you don't want people to know your sources otherwise it would be possible for people to create an artificial run on those booksellers to boost the profile of a book.
|
Ummmm, that's the stance of the NY Times, but not the stance of the other data gatherers. For example, here's the breakdown for the USA Today List:
http://www.usatoday.com/life/books/b...oks-list_n.htm
Can it be "gamed"? Yes, but it's admittedly difficult to do, and what I meant by being explained to a "reasonable" extent. (They're not 100% transparent for example by giving the specific branches, outlets, etc., but there's enough data so you know where their sources are.)
And the problem with "let's obscure it so that it can't be gamed" is like presenting a survey and citing your demographics. (And that's not to say all data should be open; Amazon for example does not disclose their figures, which is fine for them as a company.)