1JoeB1934
LT provides a powerful statistic when doing tag analysis of a book. This stat is the prevalence measurement for a specific book on a specific tag. For example: one of my best book reads is Isola by Allegra Goodman. This book is a powerful example of Women's Fiction. This book has 537 readers and only 1 reader applied the tag to this book. which gives a .19% prevalence.
If you look at the tags cloud for the book in LT you see that the maximum tag from anyone is 63 for 'historical fiction' and 3 for 'women's fiction'. I suggest that, considering how few readers bother to tag a book that the prevalence is more like 5%.
My proposal is that prevalence would be more accurate if the base wasn't the number of readers, but the number of readers who actually tagged the book. We don't know that latter number but the tag that garnered the largest count for the book is one measure.
Another candidate which is easily found is the number of reviews on the book. In this case there were 34 reviews, which could be considered.
I looked at another book of this sub-genre Go as a River. It had very similar stats: members 864, reviews 44 and tag count 3. I suggest that these prevalence calculations would be more relevant. Under the current method the prevalences are all under 1% which is considered the cut-off.
If you look at the tags cloud for the book in LT you see that the maximum tag from anyone is 63 for 'historical fiction' and 3 for 'women's fiction'. I suggest that, considering how few readers bother to tag a book that the prevalence is more like 5%.
My proposal is that prevalence would be more accurate if the base wasn't the number of readers, but the number of readers who actually tagged the book. We don't know that latter number but the tag that garnered the largest count for the book is one measure.
Another candidate which is easily found is the number of reviews on the book. In this case there were 34 reviews, which could be considered.
I looked at another book of this sub-genre Go as a River. It had very similar stats: members 864, reviews 44 and tag count 3. I suggest that these prevalence calculations would be more relevant. Under the current method the prevalences are all under 1% which is considered the cut-off.
2MarthaJeanne
Isola is the correct touchstone.
3JoeB1934
>2 MarthaJeanne: You are correct, but I have forgotten how to correct the touchstone in my post.
4MarthaJeanne
>3 JoeB1934: Open edit for your post. Go to the touchstone area on the right. Click on (others) and chose the one you want.
6jjwilson61
You're proposing a site improvement, so why wouldn't the developers be able to determine the number of users that have tagged a work at least once?
7MarthaJeanne
This message has been deleted by its author.
8bnielsen
>1 JoeB1934: Where do you see the "powerful statistic when doing tag analysis of a book. This stat is the prevalence measurement for a specific book on a specific tag." ? I can see the 537 (now 538) and 3 tagged with "Women's Fiction" but is the prevalence .19% something you calculated or something that's visible somewhere on LT?
I admit to not having visited every corner of the site so I'd like a link :-)
I admit to not having visited every corner of the site so I'd like a link :-)
9keristars
I'm a little confused at why "Women's Fiction" (3 uses) and "women's fiction" (1 use) are being treated as different tags in the list, too, when the only difference seems to be a matter of capitalization, and they both point to the same tag page.
10bnielsen
>9 keristars: If you look closer at them, you'll see "Women's Fiction " and "women's fiction" i.e. there is some odd spacing efter Fiction in the first one.
11keristars
>10 bnielsen: Aha! And here I thought extra spaces before/after were stripped.
Thanks for the solution :)
Thanks for the solution :)
12bnielsen
>11 keristars: Investigating a bit more it looks like a non breaking space %C2%A0, so it is not an ordinary space which would normally be stripped.
13paradoxosalpha
>9 keristars:, >10 bnielsen:
There is tag combination to clean up that sort of thing, isn't there?
https://www.librarything.com/log_helpers.php?view=tags
There is tag combination to clean up that sort of thing, isn't there?
https://www.librarything.com/log_helpers.php?view=tags
14JoeB1934
>8 bnielsen: My mistake! Go to the 'Mirror' for your library. Bring up the display of all tags in your library. Click on any tag name of interest, especially the bolder ones. Up comes a list of all the books having that tag by readers who tagged that book.
A nice display of book name, author and other stats. Including a tag count, the number of readers and a prevalence number which is the ratio tag count to total readers.
My issue is that usually the fraction of 'readers' that actually tag is a small fraction. Any method that comes up with a plausible number of taggers on a book would result in a more useful prevalence stat.
A nice display of book name, author and other stats. Including a tag count, the number of readers and a prevalence number which is the ratio tag count to total readers.
My issue is that usually the fraction of 'readers' that actually tag is a small fraction. Any method that comes up with a plausible number of taggers on a book would result in a more useful prevalence stat.
15gilroy
This isn't a research library. This is just a cataloging site. Tags are personal.
I'm not sure what a a more useful prevalence stat from outside your library should matter?
I'm not sure what a a more useful prevalence stat from outside your library should matter?
16keristars
I think the assumption about the most-used tag (or number of reviews) being a useful denominator for prevalence calculation is leaning far too heavily on more common works.
It utterly fails on this book - https://www.librarything.com/work/14954583/classification/295939040 (Three Vassar Girls in Russia and Turkey). Abigail and I are the only members cataloging the book thus far, but we don't have any overlapping tags. Similar, yes, but not identical.
I chose that example because it's so clearly a case where the proposed prevalence rate would be entirely wrong. There are lots of works with similar stats, maybe a few more copies or fewer tags.
(Also, the use of "readers" to refer to who has added a book and tagged (or not!) is mistaken here, too. Not everyone who has added a book has read it!)
It utterly fails on this book - https://www.librarything.com/work/14954583/classification/295939040 (Three Vassar Girls in Russia and Turkey). Abigail and I are the only members cataloging the book thus far, but we don't have any overlapping tags. Similar, yes, but not identical.
I chose that example because it's so clearly a case where the proposed prevalence rate would be entirely wrong. There are lots of works with similar stats, maybe a few more copies or fewer tags.
(Also, the use of "readers" to refer to who has added a book and tagged (or not!) is mistaken here, too. Not everyone who has added a book has read it!)
17MarthaJeanne
Since there is no set list of possible tags, lots of books have similar tags that cannot be combined. Different people make different decisions on what to tag. I'm not sure that 'prevalence' is a very useful measure.
I'm looking at one book's tags. gay (70), LGBT (102), LGBTQ (162), LGBTQ+ (83), ..., m/m (23) m/m romance (27), mm (23)...
Does it really matter which of these got more uses out of how many people entering/tagging?
I'm looking at one book's tags. gay (70), LGBT (102), LGBTQ (162), LGBTQ+ (83), ..., m/m (23) m/m romance (27), mm (23)...
Does it really matter which of these got more uses out of how many people entering/tagging?
18bnielsen
>17 MarthaJeanne: Does it really matter which of these got more uses out of how many people entering/tagging?
I guess not. But @JoeB1934 is good at asking odd questions and once in a while it stirs up something useful, so that's why I was curious :-)
I guess not. But @JoeB1934 is good at asking odd questions and once in a while it stirs up something useful, so that's why I was curious :-)
19bnielsen
>14 JoeB1934: Thanks. Note to self: https://www.librarything.com/stats/bnielsen/tagmirror and click on say Comic to see the prevalence number.
I can see that some of the small numbers lose a lot of precision i.e. 0.04% contains less information than 0.41% or 4.11%. Is that the problem you want to do something about? I.e. displaying more digits would help? I think >16 keristars: has a point that your proposed calculation would also give weird numbers. I.e. a lot of 100% on tags like this-book-contains-no-canned-fish that are only used once by a single user :-)
I can see that some of the small numbers lose a lot of precision i.e. 0.04% contains less information than 0.41% or 4.11%. Is that the problem you want to do something about? I.e. displaying more digits would help? I think >16 keristars: has a point that your proposed calculation would also give weird numbers. I.e. a lot of 100% on tags like this-book-contains-no-canned-fish that are only used once by a single user :-)
20Charon07
>15 gilroy: The prevalence rate is already a feature—calculating it is not the new proposal. As I understand it, what JoeB1934 is proposing is using a different denominator to calculate it. If I’m not mistaken, it’s currently calculated using the total number of members who’ve cataloged the book as the denominator. Joe has offered several examples of denominators that would produce a more meaningful number. I agree that a more meaningful number would be total number of members who have tagged the book or the total number of tags.
21JoeB1934
>20 Charon07: THANK You for your coming to the rescue on the idea. The person who should be considering this issue is Tim, himself. He created the Mirror concept and the somewhat complex of bold and size names of the tags is very important.
He created the prevalence concept and the note about a book is below the 1% cutoff.
I consider the Mirror feature one of the crown jewels of LT. Goodreads has a TOP Shelves approach, and the LT concept is more powerful.
He created the prevalence concept and the note about a book is below the 1% cutoff.
I consider the Mirror feature one of the crown jewels of LT. Goodreads has a TOP Shelves approach, and the LT concept is more powerful.
22JoeB1934
>15 gilroy: Your approach to this issue isn't a surprise to me at all.
My interest is precisely as a research project. I am 91 yrs old and doing a look back at my total reading history to understand why certain books really clicked with me.
Do you understand at all why you like some books, but not others?
I have in my library books I read back in the 1970's. At least some of them as my memories aren't always retrievable.
My interest is precisely as a research project. I am 91 yrs old and doing a look back at my total reading history to understand why certain books really clicked with me.
Do you understand at all why you like some books, but not others?
I have in my library books I read back in the 1970's. At least some of them as my memories aren't always retrievable.
23gilroy
>22 JoeB1934: Do I know why some books click and some don't? No.
Why? Because it isn't something that I see as quantifiable. It's not something that can be drilled down on as a number or a tag or a data point.
And it changes over time. A book I tried and struggled with as a teen, I have a better understanding of as an adult, so it sits better.
And how other people chose to tag, to quantify their library doesn't always equal how I would quantify my library.
Why? Because it isn't something that I see as quantifiable. It's not something that can be drilled down on as a number or a tag or a data point.
And it changes over time. A book I tried and struggled with as a teen, I have a better understanding of as an adult, so it sits better.
And how other people chose to tag, to quantify their library doesn't always equal how I would quantify my library.
24keristars
>21 JoeB1934: What is "the note about a book is below the 1% cutoff" referring to?
"Why did i like this book?" is an interesting project, but honestly I'm not sure if tags can help there. I just have too many books where the objective tags are the same, and there isn't much subjective tagging. But you are clearly finding it more useful than me. :)
Anyway, the Tag Mirror says:
"An ordinary tag cloud shows how a member tags their books. Tag mirror shows how the whole LibraryThing community is tagging the same books."
I suppose prevalence for that could be interpreted as either "tag rate per number of copies" or "tag rate per number of taggers". The latter does seem more interesting to me, especially if there are hundreds of copies without even a "fiction" or "TBR" tag.
But it would have to be based on number of copies with tags, not any of the proposed shortcut solutions. Is that easy to calculate, or would it require a lot of processing time?
"Why did i like this book?" is an interesting project, but honestly I'm not sure if tags can help there. I just have too many books where the objective tags are the same, and there isn't much subjective tagging. But you are clearly finding it more useful than me. :)
Anyway, the Tag Mirror says:
"An ordinary tag cloud shows how a member tags their books. Tag mirror shows how the whole LibraryThing community is tagging the same books."
I suppose prevalence for that could be interpreted as either "tag rate per number of copies" or "tag rate per number of taggers". The latter does seem more interesting to me, especially if there are hundreds of copies without even a "fiction" or "TBR" tag.
But it would have to be based on number of copies with tags, not any of the proposed shortcut solutions. Is that easy to calculate, or would it require a lot of processing time?
25JoeB1934
>24 keristars: You raise several interesting points here. First, what if the the objective tags are all the same. For example, I prefer literary, historical books with suspense. I have read hundreds such books, so what can I learn from tagging? The answer is in the non-main, so-called minor tags.
I have learned over the years of messing with this issue that the more legitimate tags on a book the odds of my liking it goes up. An analogy would be what differentiates one good meal from another when the entre is the same thing. It is the accompaniments, sauces, side dishes, the chef provided details that caused the bell to ring.
The best book I have read recently is Bel Conte and it possesses around 20 substantial minor tags.
The reason I am interested in this prevalence rate is tied to my interest in delving accurately into the minor tags.
As to how the prevalence rate can be improved is a serious question of number crunching. When I look at the tag cloud for a book I look for the tag that has the maximum count. That is some sort of baseline but I don't know how it could be formulated.
Something that I think is a candidate is the number of reviews on the book. Presumably anyone who writes a review has placed their tags on the book. This number is known for every book and it would easily be used.
I am not proposing either of these solutions as all I wanted was that someone in management would see this issue warrants consideration.
I have learned over the years of messing with this issue that the more legitimate tags on a book the odds of my liking it goes up. An analogy would be what differentiates one good meal from another when the entre is the same thing. It is the accompaniments, sauces, side dishes, the chef provided details that caused the bell to ring.
The best book I have read recently is Bel Conte and it possesses around 20 substantial minor tags.
The reason I am interested in this prevalence rate is tied to my interest in delving accurately into the minor tags.
As to how the prevalence rate can be improved is a serious question of number crunching. When I look at the tag cloud for a book I look for the tag that has the maximum count. That is some sort of baseline but I don't know how it could be formulated.
Something that I think is a candidate is the number of reviews on the book. Presumably anyone who writes a review has placed their tags on the book. This number is known for every book and it would easily be used.
I am not proposing either of these solutions as all I wanted was that someone in management would see this issue warrants consideration.
26keristars
Presumably anyone who writes a review has placed their tags on the book.
I would warn that you can't presume anything about LT member behavior. I'm positive there are reviews with no tags attached to the member's copy.
But I was referring to minor tags as being the same, too, for some of the books I have. It's when I get into the unique tags that I start seeing differences - and those wouldn't show up in the tag mirror, unless I'm pointing it at a very narrow selection
And again, using the tag with the highest usage breaks very quickly. In the example I linked above, my tag "traveling abroad" has a 100% prevalence under that formula, but it is on one of two copies. The other copy prefers "Americans abroad", which would also have a 100% prevalence.
There are a lot of books where the highest usage is only 1 or 2! I see them all the time when doing combining or tracking down books in my research field.
That's why I made the suggestion to improve it by using number of copies with tags as the denominator. It's the only formula that would work as an alternative to total copies.
I just have no idea how feasible it is to do the calculations. Tags are resource intensive, is my understanding for why they aren't instantaneously updated on works but are monthly/quarterly/whatever.
I would warn that you can't presume anything about LT member behavior. I'm positive there are reviews with no tags attached to the member's copy.
But I was referring to minor tags as being the same, too, for some of the books I have. It's when I get into the unique tags that I start seeing differences - and those wouldn't show up in the tag mirror, unless I'm pointing it at a very narrow selection
And again, using the tag with the highest usage breaks very quickly. In the example I linked above, my tag "traveling abroad" has a 100% prevalence under that formula, but it is on one of two copies. The other copy prefers "Americans abroad", which would also have a 100% prevalence.
There are a lot of books where the highest usage is only 1 or 2! I see them all the time when doing combining or tracking down books in my research field.
That's why I made the suggestion to improve it by using number of copies with tags as the denominator. It's the only formula that would work as an alternative to total copies.
I just have no idea how feasible it is to do the calculations. Tags are resource intensive, is my understanding for why they aren't instantaneously updated on works but are monthly/quarterly/whatever.
27JoeB1934
>26 keristars: I can see where you are coming from, and you could well be right in your circumstance. In my case the main tags are ones many readers use, like mystery, suspense, science fiction etc.
For some of these reasons I have become disappointed in reader generated tags where there aren't any guidelines, or ground rules.
As a result, my current product of genre analysis is based on the site FICTIONDB which has a defined set of genres/sub-genres/themes and so on. Every book in the very large system has genre assignments that meet their definitions.
I can analyze my LT library using the Mirror and just ignore the prevalence issue. I can also analyze my same library in FICTIONDB and see what a consistent set of genre definitions does for me. I am in the midst of doing these comparisons.
For some of these reasons I have become disappointed in reader generated tags where there aren't any guidelines, or ground rules.
As a result, my current product of genre analysis is based on the site FICTIONDB which has a defined set of genres/sub-genres/themes and so on. Every book in the very large system has genre assignments that meet their definitions.
I can analyze my LT library using the Mirror and just ignore the prevalence issue. I can also analyze my same library in FICTIONDB and see what a consistent set of genre definitions does for me. I am in the midst of doing these comparisons.
28lilithcat
>25 JoeB1934:
Presumably anyone who writes a review has placed their tags on the book
I wouldn't assume that. I just looked at a random book in my library that had six reviews. Half of them had no tags at all. The only tag the other three shared was "fiction". Interestingly, they all had a tag that was similar, but not the same, in meaning: "gay", "LGBT", and "LGBTQ".
Presumably anyone who writes a review has placed their tags on the book
I wouldn't assume that. I just looked at a random book in my library that had six reviews. Half of them had no tags at all. The only tag the other three shared was "fiction". Interestingly, they all had a tag that was similar, but not the same, in meaning: "gay", "LGBT", and "LGBTQ".
29GraceCollection
The way I see it, 'prevalence' can mean a few quite different things:
- How often this tag is used on this book, out of all copies of this book (which from my understanding, is the current calculation)
- How often this tag is used on this book, out of the copies of this book which have any tags at all (from my understanding, this is what JoeB is asking for, and certainly is more helpful than the above)
- How often this tag is used on this book, out of all times this tag has been used (Although it would likely be a very small percentage number, if it were a ratio or some other weighted measurement, I think this might be even more helpful)
- How often this tag is used on this book, compared to all tags on all books owned by anyone who has ever used this tag (probably too difficult to compute, but certainly would be interesting)
- How often this tag is used on this book, compared to all tags on all books (not likely to be very useful in my opinion)
I know JoeB was part of this discussion, for but any who may have missed it ( https://www.librarything.com/topic/368106 ), I think significance might be information that would be more interesting to be available in more places, like on the pop-up box for a particular tag in the tag mirror, or on the tags for a particular book:
"Significance is defined for a tag as:
Count of times on all your works
divided by
SQRT( count of times across all works on LibraryThing + 1) + 10"
(quoting @timspalding)
- How often this tag is used on this book, out of all copies of this book (which from my understanding, is the current calculation)
- How often this tag is used on this book, out of the copies of this book which have any tags at all (from my understanding, this is what JoeB is asking for, and certainly is more helpful than the above)
- How often this tag is used on this book, out of all times this tag has been used (Although it would likely be a very small percentage number, if it were a ratio or some other weighted measurement, I think this might be even more helpful)
- How often this tag is used on this book, compared to all tags on all books owned by anyone who has ever used this tag (probably too difficult to compute, but certainly would be interesting)
- How often this tag is used on this book, compared to all tags on all books (not likely to be very useful in my opinion)
I know JoeB was part of this discussion, for but any who may have missed it ( https://www.librarything.com/topic/368106 ), I think significance might be information that would be more interesting to be available in more places, like on the pop-up box for a particular tag in the tag mirror, or on the tags for a particular book:
"Significance is defined for a tag as:
Count of times on all your works
divided by
SQRT( count of times across all works on LibraryThing + 1) + 10"
(quoting @timspalding)
30bnielsen
>27 JoeB1934: I took a (short) look at FICTIONDB (and the various ad's :-)
https://www.fictiondb.com/title/broken~karin-fossum~233040.htm
Displays
Main Genre: Law Enforcement
and under Genres:
General Fiction
Suspense
and under Sub-Genres
Law Enforcement
so it seems that FICTIONDB has three notions of Genre?
And a given book can be in a lot of Sub-Genres
https://www.fictiondb.com/title/game-on~janet-evanovich~2834522.htm
https://www.fictiondb.com/title/broken~karin-fossum~233040.htm
Displays
Main Genre: Law Enforcement
and under Genres:
General Fiction
Suspense
and under Sub-Genres
Law Enforcement
so it seems that FICTIONDB has three notions of Genre?
And a given book can be in a lot of Sub-Genres
https://www.fictiondb.com/title/game-on~janet-evanovich~2834522.htm
31JoeB1934
>30 bnielsen: The full set of tags can be seen by looking at 'Search' where they provide Genres/Sub-Genres/Themes/People-Creatures/Age Level/Time Period/Other.
Within each grouping there are many/many tags. Under 'Other' is one of my favorites 'Award Winner' which indicates whether the book has received attention from the system.
It is an indicator of 'recognition' for the book to me. Also, the main book page provides the rating stats.
Within each grouping there are many/many tags. Under 'Other' is one of my favorites 'Award Winner' which indicates whether the book has received attention from the system.
It is an indicator of 'recognition' for the book to me. Also, the main book page provides the rating stats.
32JoeB1934
All I'm asking for is a reasonable improvement to the prevalence indicator that has been in LT for a very long time. If it doesn't work for you then ignore it.
33timspalding
Sorry. Can someone explain exactly what they want, where, specifically?
34GraceCollection
>33 timspalding: Per >14 JoeB1934:, this is an RSI about the 'prevalence' information in the Tag Mirror:
From my understanding, this number is currently calculated as how often this tag is used on this work, out of all copies of this work. I believe what JoeB is asking for is that this be changed to calculate how often this tag is used on this work, out of only the copies of this work which have any tags at all.
In other words, if only 2% of the people who own a certain work have bothered adding tags to it, the 'prevalence' of any tags will be super low, even if every one of that 2% has used, say, 'horror'.
I chimed in in >29 GraceCollection: that I think adding 'significance' as it was defined by you in this thread: https://www.librarything.com/topic/368106 would also be very helpful information to have.
My mistake! Go to the 'Mirror' for your library. Bring up the display of all tags in your library. Click on any tag name of interest, especially the bolder ones. Up comes a list of all the books having that tag by readers who tagged that book.
A nice display of book name, author and other stats. Including a tag count, the number of readers and a prevalence number which is the ratio tag count to total readers.
My issue is that usually the fraction of 'readers' that actually tag is a small fraction. Any method that comes up with a plausible number of taggers on a book would result in a more useful prevalence stat.
From my understanding, this number is currently calculated as how often this tag is used on this work, out of all copies of this work. I believe what JoeB is asking for is that this be changed to calculate how often this tag is used on this work, out of only the copies of this work which have any tags at all.
In other words, if only 2% of the people who own a certain work have bothered adding tags to it, the 'prevalence' of any tags will be super low, even if every one of that 2% has used, say, 'horror'.
I chimed in in >29 GraceCollection: that I think adding 'significance' as it was defined by you in this thread: https://www.librarything.com/topic/368106 would also be very helpful information to have.
35timspalding
Okay, I've pushed a change. Honestly, I don't think you'll like it, but I think it's for the best.
First, why?
1. I understand your concern.
2. We don't easily have the number of members who've tagged a work. Yes, it's trivial to get it, but not when there are thousands of taggers, so we'd have to precompute it. Unfortunately we have more than 50 million works. We do a lot of calculations; I don't want to add this one for just this.
3. We DO have the number of tags on a work, but
4. That number is cached and may be up to weeks out of date. It can be used, but it can't be the only factor.
5. I don't the prevalence number adds much; it wasn't the same as the sort order, which was confusing. Less is more, especially when more is confusing.
6. The whole thing used old design, with tables. It needed a refresh.
The changes are:
1. The sort calculation (which was similar to the prevalence calculation) has changed slightly.
2. Prevalence has been hidden.
3. The design has been updated and improved, using a different CSS structure.
For what it's worth, the sort order is based on the following formula:
First, why?
1. I understand your concern.
2. We don't easily have the number of members who've tagged a work. Yes, it's trivial to get it, but not when there are thousands of taggers, so we'd have to precompute it. Unfortunately we have more than 50 million works. We do a lot of calculations; I don't want to add this one for just this.
3. We DO have the number of tags on a work, but
4. That number is cached and may be up to weeks out of date. It can be used, but it can't be the only factor.
5. I don't the prevalence number adds much; it wasn't the same as the sort order, which was confusing. Less is more, especially when more is confusing.
6. The whole thing used old design, with tables. It needed a refresh.
The changes are:
1. The sort calculation (which was similar to the prevalence calculation) has changed slightly.
2. Prevalence has been hidden.
3. The design has been updated and improved, using a different CSS structure.
For what it's worth, the sort order is based on the following formula:
36JoeB1934
Tim, I was the member that triggered this thread way back in Nov 2025 and had pretty much moved on to my main analytical activities in 2025 that triggered the initial question. Fortunately, I had starred this thread, and I was surprised when your post turned up.
I didn't exactly know what the sorting algorithm would accomplish, but willing to give it a try. I originally assumed that, like most changes in LT, it was a proposal for the future release. However, I thought why not got to the Mirror and see.
Amazed that it was there already! I don't know if anyone else has commented, but personally I think the approach is terrific.
I don't own many books so I can build a TBR any day I want by using whatever procedure I am interested in. I had built a TBR using my preferred sub-genre analysis at around 800 books to start 2026 with.
Here is where the utility of the sort came into play. If I open the Mirror on a favorite sub-genre (tag), I see a ranking of the books in my TBR very close to what I expected based on the reasons I put the books in my TBR.
What if I was a more typical member that had a TBR built at random over the years and wanted in 2026 to pick some books to read. This Mirror method of sorting the TBR would be a lifesaver.
I realize that I am way more interested in using the Mirror than most any other member, but this new sort is excellent.
I didn't exactly know what the sorting algorithm would accomplish, but willing to give it a try. I originally assumed that, like most changes in LT, it was a proposal for the future release. However, I thought why not got to the Mirror and see.
Amazed that it was there already! I don't know if anyone else has commented, but personally I think the approach is terrific.
I don't own many books so I can build a TBR any day I want by using whatever procedure I am interested in. I had built a TBR using my preferred sub-genre analysis at around 800 books to start 2026 with.
Here is where the utility of the sort came into play. If I open the Mirror on a favorite sub-genre (tag), I see a ranking of the books in my TBR very close to what I expected based on the reasons I put the books in my TBR.
What if I was a more typical member that had a TBR built at random over the years and wanted in 2026 to pick some books to read. This Mirror method of sorting the TBR would be a lifesaver.
I realize that I am way more interested in using the Mirror than most any other member, but this new sort is excellent.
