How to interpret "Median/mean book obscurity" ?
Talk Frequently Asked Questions
Join LibraryThing to post.
This topic is currently marked as "dormant"—the last message is more than 90 days old. You can revive it by posting a reply.
1msodalis First Message
I do not understand what the "median/mean book obsurity" numbers indicate. For example, I have 474 books listed and my current obscurity rating is 2/10. How are these numbers calculated and what are they telling me?
2AsYouKnow_Bob
How are these numbers calculated and what are they telling me?
Hi, welcome to LT. The flip answer is that the "book obscurity" statistics don't tell you a whole lot.
Your books are checked against the database of all 130,000 LT users - and on average, each of your books is held by 10 LT users.
The median is another measure of your library:
if you rank your 474 books from 'most obscure' to 'most widely shared', then your median book (that is, #237) is shared by only two users. Half your books are more obscure than that, half will be more widely held than that.
Amongst the things that make this problematical is the 'book obscurity' stats depend upon matching up your book data with the rest of LT: if there's a typo in your info, or if you hold an unusual edition, it won't match up with other copies here.
Anecdotally, it's telling you that you have a relatively unusual collection. A library consisting only of best sellers would have 'book obscurity' stats well into the hundreds.
Hi, welcome to LT. The flip answer is that the "book obscurity" statistics don't tell you a whole lot.
Your books are checked against the database of all 130,000 LT users - and on average, each of your books is held by 10 LT users.
The median is another measure of your library:
if you rank your 474 books from 'most obscure' to 'most widely shared', then your median book (that is, #237) is shared by only two users. Half your books are more obscure than that, half will be more widely held than that.
Amongst the things that make this problematical is the 'book obscurity' stats depend upon matching up your book data with the rest of LT: if there's a typo in your info, or if you hold an unusual edition, it won't match up with other copies here.
Anecdotally, it's telling you that you have a relatively unusual collection. A library consisting only of best sellers would have 'book obscurity' stats well into the hundreds.
3SimonW11
It means your books are ridiculously obscure:^)
If we were to take your book and order them according to howmany people shared them the one oat the half way point would be shared with only one other person.This is the first the median figure only two people on LT hae that book. looking at your list i would say this is probably a work of mainly local interset you seem to have a lot of books that would normaly only be seen in a limited geographical area and hence hae few copies on a global site like this hence the average or Mean number of books calculated by totaling all the copies of books on LT that you have a copy of and dividing by that the number you yousrself own is also at ten extraordinarily low. I hope that made sense I am more than a little drunk.
Simon
If we were to take your book and order them according to howmany people shared them the one oat the half way point would be shared with only one other person.This is the first the median figure only two people on LT hae that book. looking at your list i would say this is probably a work of mainly local interset you seem to have a lot of books that would normaly only be seen in a limited geographical area and hence hae few copies on a global site like this hence the average or Mean number of books calculated by totaling all the copies of books on LT that you have a copy of and dividing by that the number you yousrself own is also at ten extraordinarily low. I hope that made sense I am more than a little drunk.
Simon
4msodalis
Thanks AsYouKnow_Bob and SimonW11 for the feedback. I believe I understand the concept now. I am fairly new on LT and so far have fewer than 20% of my library listed, and I'm still learning the basics. I do tend to have a lot of books that are unusual, obscure, narrow-focused, or technical, so I expect my "obscurity" numbers will remain fairly low. Although I do have some relatively popular books, I do not yet have my rarer and older books listed.
Another contributing factor to my low obscurity numbers could be because some of my editions do not match the publisher and date information provided so I edit those, and sometimes also edit book titles because because they do not match exactly or are misspelled. I have also found errors even with authors' names. Plus I sometimes include the edition number with the title and delete the information in parentheses when it is not actually part of the title and shows up in the publication information block.
Currently, I am also struggling with coming to grips with how to tag my books so that I can do searches that make sense to me. I'm in the process of reworking some of my initial tags and doing test searches to see if I am happy with them before getting too far along with my listings. Hopefully, doing this now will help save some work in the long run. Since I have several thousand books in my collection, it is important to me to do this reasonably well since it's getting hard to keep it all my head. :-)
Another contributing factor to my low obscurity numbers could be because some of my editions do not match the publisher and date information provided so I edit those, and sometimes also edit book titles because because they do not match exactly or are misspelled. I have also found errors even with authors' names. Plus I sometimes include the edition number with the title and delete the information in parentheses when it is not actually part of the title and shows up in the publication information block.
Currently, I am also struggling with coming to grips with how to tag my books so that I can do searches that make sense to me. I'm in the process of reworking some of my initial tags and doing test searches to see if I am happy with them before getting too far along with my listings. Hopefully, doing this now will help save some work in the long run. Since I have several thousand books in my collection, it is important to me to do this reasonably well since it's getting hard to keep it all my head. :-)
5AsYouKnow_Bob
Since I have several thousand books in my collection, it is important to me to do this reasonably well since it's getting hard to keep it all my head.
I hear you. (Believe me, I hear you....)
The first thing you learn on LT, is that, yes, the automated 'Amazon look-up' seems miraculous: enter an ISBN, and there's your book information.
The second thing you learn on LT is that the Amazon data is not reliable.
I hear you. (Believe me, I hear you....)
The first thing you learn on LT, is that, yes, the automated 'Amazon look-up' seems miraculous: enter an ISBN, and there's your book information.
The second thing you learn on LT is that the Amazon data is not reliable.
6SimonW11
If you are editing book titles you will be isolating works from their fellows. this can be combated by combineing, I suggest you read the combining group for a while first.
7regorfa First Message
Along those same lines, I'm new to library thing (just joined today)... how do I interpret the "Users with your books: Weighted | raw | recent" on the right in my profile?
8mangaroo
The numbers in parentheses are the number of cataloged books you have in common with the user.
* "Raw" simply lists them in order from most books in common to fewest.
* "Recent" does the same for books added by users in the past two weeks, so you can more easily identify new users with your interests.
* "Weighted" sorts on an algorithm that privileges two factors: the obscurity of the books you have in common and the percentage of shared books in the user's library. The numbers in parentheses (1/2) are the number of books you share with the user (1) followed by the total number of books in that user's LT catalog (2).
There's a cut-off point for all the lists, or users with commonly-held books (hello, Harry Potter series) would have thousands of names listed.
* "Raw" simply lists them in order from most books in common to fewest.
* "Recent" does the same for books added by users in the past two weeks, so you can more easily identify new users with your interests.
* "Weighted" sorts on an algorithm that privileges two factors: the obscurity of the books you have in common and the percentage of shared books in the user's library. The numbers in parentheses (1/2) are the number of books you share with the user (1) followed by the total number of books in that user's LT catalog (2).
There's a cut-off point for all the lists, or users with commonly-held books (hello, Harry Potter series) would have thousands of names listed.
