On 11/12/2017 06:25 AM, dibya wrote: > Thanks for your response. The bestseller rank indeed means the Amazon > Bestseller Rank. I understand the reasons you mentioned about why this rank > might not be the perfect way to rank the books - and that public voting > might be better. However, I went ahead ahead with this choice for the > following reasons:
> 2. From my experience, it seems that the Amazon Bestseller Rank, with all > its faults, is still a very good indicator of quality. How do I know this? > Well, I visited several online threads about best Python books for > beginners, intermediate programmers, books on specializations like Data > Science, Security etc. and made a list of books that people wholeheartedly > recommend. Then I compared this list with what I get from the Bestseller > Rank. It's astonishing how well the two lists match up in almost every > case. My conclusion from this is the following: you can promote a book all > you want, but if you want strong sales, quality beats promotion. > > 3. The third point is about books not on Amazon. I do believe this applies > to only a small minority of Python books. According to my database, Amazon > currently lists 1005 paperback Python books published since 2009. This is a > quite complete database and all the popular books are definitely on there. > I do understand that I might have missed one or two books that are not on > Amazon, but that's a compromise that I was willing to make (because the > alternatives were far inferior is terms of cost-benefit). Not trying to argue, just making personal observations: I get nearly all of my tech books at the moment electronically directly from Packt. At some time in the past I had a company-paid subscription to Safari and those books arrived electronically as well. I'm assuming Amazon was never involved in any way in either set of transactions, unless there's some kind of API where you can inject a record of a "sale", though that's a particularly nebulous concept in the case of subscription services, which I think are a significant factor particularly in tech book sales - and probably cause "total downloads" to be not indicative of quality: if an individual book is effectively free, people grab it and then quickly ignore it if it's bad. Maybe Amazon's numbers are still representative on a proportional basis, I don't know. Anyway, there's nothing to prevent you from being marked as a wiki editor, which is now the case. _______________________________________________ pydotorg-www mailing list pydotorg-www@python.org https://mail.python.org/mailman/listinfo/pydotorg-www
