[davidfor]

Originally Posted by Caine

In my opinion, these stats are useless. Like David I don't like the representation, it's poorly chosen, as evidenced by the attached image.

I didn't say that. I do like the look of the stats. I was mainly questioning the use completing the circle to the lengths. They need something extra for when the book is more than 500 pages.

How does the concept of pages apply to an e-book reader anyway? The number of pages in the print edition might be useful in this regard even though sometimes there are huge differences in font size and line spacing over multiple editions. Maybe there's a magic number of words assumed for a page? A page used to consist of roughly 250 words back when novels were written on typewriters. Today it's somewhere between 250 and 300 words for your regular paperback novel.

They are probably using the Adobe page count algorithm. Their ereaders do this for epubs and it is an option for kepubs. And of course it is innacurate and depends on so many other things. But, so long as they use the same algorithm across their site, it gives a good relative size. And is more understandable to most people, myself included. That's how I have compared length of novels for as long as I've been reading. I can see word count is probably better but I don't have a feel for what it means yet.

I've configured my Kobo in a way that I see roughly the same amount of content as when reading a traditional mass market paperback novel. I know people who use large fonts, a ridiculous amount of line spacing, or huge margins. The same applies to the reading time. I tend to read fiction three to four times faster than nonfiction and almost everybody I know reads much slower than I do.

I also doubt the algorithm is excluding all kinds of front/back matter like huge appendices (roughly 60 pages in Martin's A Dance with Dragons) from these stats, throwing them further off. Statistically not relevant? Maybe.

There are not that many books showing stats, at the moment. Probably depends on the data received from the customers.

To my mind, these are approximations and not stats. For what it's worth, I'm sure I read Ice Station at least twice as fast as I did Half a King. And I was on vacation when I read the latter last year with plenty of time.

I assumed the reading time is based on some words per minute or pages per minute for a mythical average reader. It would be nice to think they are calculating it based on their user base, or even better, based on my stats, but I doubt it.

And why would the backmatter be excluded? The appendices are parts of the books and get read, so they should be included. Excluding "Bonus" material like the first chapter of the next book in the series would be a good, but that needs some sort of indication in the book if the calculation is to be automated. But again, that is still part of the book, so if the time is for reading "cover to cover", it should be included.