The New York Times' Bestseller lists have always been a bit of a mystery to me. I thought it would make more sense to take a longer-term look than just a single week's rankings.
Using the New York Times API I gathered data for eight weeks' worth of trade fiction paperback bestsellers. From there I loaded the book images into a script that would determine the most common colors from the pixels of the book cover. These colors are used in a gradient to make each line identifiable.
Choosing the width of the lines was difficult. I wanted to show the lines growing in width over time, so they are scaled to reflect the number of weeks on the list at the time of ranking. However books range from one week on the list (a single circle) to two weeks, up to 308 weeks for The Alchemist. I picked a scale that worked for the range of most of the books, letting longer-running books just use the maximum width.
There are some errors with the data, where "weeks on list" comes through as zero (which makes the lines on the graph very thin) and sometimes cover images are not listed (in which case I show a "no image available" image).
Roll over the lines in the graph to see the details of each bestseller, and its rankings over time. The small graphs below show each book's individual rankings over time.
Load the most recent eight weeks' worth of data from a different list by clicking "load a different list" and choosing one.