date
newest »

message 1:
by
Brian
(new)
Apr 04, 2017 08:40AM

reply
|
flag




I find the Star Ratings on Goodreads and Amazon quite annoying. Almost all of the books range between 3.5 and 4.5, It is nothing but an average. So I have been playing with the statistics.
Total the 4s and 5s and divide that by the total of the 1s and 2s. That basically gives you a ratio of the people who liked a book to those who didn't. Using Goodread's statistics Komarr has the highest ratio I have found so far. Moby Dick got 2.29.
LOL
Barrayar got 40 and Shards of Honor got 18
There is a certain logic to this for a series. People are trying the first book in a series as an unknown. People who rate it and don't like it will lower the ratio. But they probably will not read more of the series. So later books will mostly have a biased audience of people who liked earlier books and the Star Ratio should rise.

I find the Star Ratings on Goodreads and Amazon quite annoying. Almost all of the books range between 3.5 and 4.5, It is nothing but an average. So I have been playing with t..."
Heh. You could run a sample test of your theory by rating the four Penric novellas. Being something of a series-in-miniature by now.
Ta, L.
Ta, L.

But I am biased against fantasy and the test should be with a different author.
LOL
How about Robert J. Sawyer?
#1. Far-Seer 14
#2. Fossil Hunter 25
#3. Foreigner 27
You beat the pants off of him. Naughty girl.
There will of course be series where things eventually go down hill where the author runs out of steam or producing junk for the money. The Gor series for example. Then Dune series taken over by different writer. I got tired of Weber after Flag in Exile.

I find the Star Ratings on Goodreads and Amazon quite annoying. Almost all of the books range between 3.5 and 4.5, It is nothing but an average. So I have been playing with t..."
If a book rates over a 2.5 I give it a chance. I also look at what people hated about it in their reviews. I have certain "absolutely not" triggers, like poorly written romantic leads or dystopians where society collapses in an unrealistic way (it's been 50 years since New York fell...and now we have castles!). But other than that, I rate pretty harshly myself sometimes so I give things a little benefit of the doubt.
I like your formula though for ratings. I'm gonna take it out in the wild and give it a spin. Thanks for sharing!

I am thinking about weighting the system. Some stats have more 5s then 4s and some more 4s then 5s, but my current system gives equal weight to both. So I am thinking of multiplying the 5s by 1.1 and 4s by 0.9 so if there are equal 4s and 5s there should not be much effect but if they are significantly different the score should change.
