> it's pretty clear that a list like this is going to underrepresent books where an amazon link is less likely
But the list was explicitly generated from Amazon links... wasn't that always going to be the case?
But the list was explicitly generated from Amazon links... wasn't that always going to be the case?
Tagging the links was a conscious decision, but I do think that the approach of looking for Amazon links provides more insight than any other domain/online store.
It's not super clear from the website whether this is something you did primarily to learn about data analysis, learn about HN's book recommendations, or to make money. Any of those is fine! But the motivation is likely to influence the outcome (even unintentionally) and it's something that could be made more clear.