The Sword and Laser discussion

All that being said, do many people actually use the ratings to pick which books to read? I generally only look at them after I've already picked or read something and the most it ever causes me to do is say "What the fuck were these people thinking?".

My one objection to those goodreads subjective statements about the star ratings is that 3 stars, being the median selection when you can only choose from 5 answers (there's no zero star option), is a neutral opinion, not "i liked it." 3 stars is replacement-level entertainment product (i wasn't mad that i spent time reading this but neither was i pleased) and i wouldn't recommend it to others unless i had reason to think they'd enjoy it much more than i did.

Phil wrote: "All that being said, do many people actually use the ratings to pick which books to read? I generally only look at them after I've already picked or read something and the most it ever causes me to..."

I seldom check out random strangers' reviews before I start a book, but I have picked up books (or prioritized them) based on Goodreads reviews from my friends list.

Brendan wrote: "My one objection to those goodreads subjective statements about the star ratings is that 3 stars, being the median selection when you can only choose from 5 answers (there's no zero star option), i..."

I have a 0 star option. An exclusive folder named 'Abandoned' where I put books that I couldn't finish.

Sean wrote: "What you're describing is a four star book for me. An author should have to work at least twice as hard to earn that last star. c..."


And that's why the star system is useless. You have a different set of criteria for that last star than others yet the system doesn't differentiate - a star is a star to GR.

"Ready Player One is a really good example. It's super dumb and wildly implausible, existing purely as a vehicle for 80s pop culture nostalgia. But as I've said a few times, hating the book for that reason is like hating cotton candy for not being nutritious..."

mmm, I don't HATE the book, but it's a 2 star to me - it's OK. Decent. Not a re-read. Precisely because it leans so heavily on 80s nostalgia and is otherwise a lazy, templated story. Unlike other books like that, I'm more vocal about RP1 because it gets praised so much for being awesome by people who don't differentiate between a book that's truly awesome and one that just tickles them. There's nothing wrong with the latter, but let's recognize if for what it is, a book that one really loves if one is a reader that shares that 80s love.

I think most people should average above a 3 because we tend to read books by authors we like.
For some of us (or perhaps just me) one issue with the 5 star system is that it maps to the A-B-C-D-F grading the US uses. At least this is true for me and so to me a 3 is a C is... OK. The lowest decent passing grade. A 2 star is a D which means you're a poor student, er, book. Yet GR classifies it as OK.

PS: Go ahead and find the definitions of the star ratings on GR. If you KNOW the link, don't play. But if you don't... you won't easily find it. Now, how silly is it for a review site to not make it easy for people to find the definition of the ratings system?

Rick wrote: Go ahead and find the definitions of the star ratings on GR. If you KNOW the link, don't play. But if you don't... you won't easily find it. Now, how silly is it for a review site to not make it easy for people to find the definition of the ratings system? "

Is it not available every time you hover your mouse over a star in a review? Maybe it's different if you're using the app.

Rick wrote: "PS: Go ahead and find the definitions of the star ratings on GR. If you KNOW the link, don't play. But if you don't... you won't easily find it. Now, how silly is it for a review site to not make it easy for people to find the definition of the ratings system?"

... It's super easy to find the definitions of the star ratings, just hover your mouse over the stars when you're about to rate (I already mentioned these in my earlier comment):

1 (did not like it)
2 (it was OK)
3 (liked it)
4 (really liked it)
5 (it was amazing)

Well, I've never noticed that... and for a very good reason. The small definition popups up after a short delay. I don't (and I would bet most don't) pause while hovering over the stars. I move my cursor there and click.

Also... why not have these definitions in Help, etc?

PS: hover behaviors don't work on touch devices since there's no way to detect a hover there.

Generally I find good reviewers by their negative leaning reviews on more obscure indy books that I agree with, then compare to see if they hate everything or they tend to like the same things I do, and where they disagree do I agree with their disagreement. When I follow these people it's normally very helpful in going okay X liked this a lot and Y had some problems but I'm okay with that lets go, and of course if nobody I'm following has reviewed it that's a pretty big red flag in of itself unless it's in the <100 reviews land.

I lean pretty heavily on 3s and 4s with 77% of all my reviews being one of those 2 numbers.

AndrewP wrote: "I have a 0 star option. An exclusive folder named 'Abandoned' where I put books that I couldn't finish. "

That raises the question of should there be a "DNF" (Did not finish) option as an alternative to stars?

I, personally find it frustrating to see review that begin with something like "I only made it 15 pages into the book..." and then proceed to rate the book 1 or 2 stars.
I don't feel like its fair to the author or the process if you rate an entire booked based off your experience with 1% of it.
While you can certainly tell if you may not enjoy that book within that amount of time, I don't think you can tell if its good or not.

You can create a DNF shelf as an exclusive shelf and that way abandoned books won't go on your read shelf. It's better than deleting them, so you don't make the mistake of picking them up again because you forgot you already abandoned them.

Jenny (Reading Envy) wrote: "You can create a DNF shelf as an exclusive shelf and that way abandoned books won't go on your read shelf ..."

Thank you so much for this comment! I had no idea you could add a shelf that behaved similarly to the three default options. I was just agonising over how I would file the book I'm on the verge on lemming. I want to record the attempt made, but not count it toward my 2017 reading challenge.

On the topic of the usefulness of the ratings - the numbers themselves are always going to be subjective as everyone approaches it slightly differently. I would never read or not read based on the average score alone, but I often use the written reviews to find out more when a book catches my eye.

And yes, I do wish I could award half stars to differentiate a bit more finely, but I'm not naive enough to think that anyone besides myself would notice.

THIS!

Trike wrote: " hating the book for that reason is like hating cotton candy for not being nutritious. So for me it's a 3-star book, because it fulfilled its cotton candy mandate well.

I just watched Kong: Skull Island last night and I liked it for the same reason: it's a giant monster movie that showed giant monsters.

Rick wrote: "For some of us (or perhaps just me) one issue with the 5 star system is that it maps to the A-B-C-D-F grading the US uses. At least this is true for me and so to me a 3 is a C is... OK."

If you equate it to grades, 3-stars is 60%, which is a D- if we're being generous.

Colin wrote: "Jenny (Reading Envy) wrote: "You can create a DNF shelf as an exclusive shelf and that way abandoned books won't go on your read shelf ..."

Thank you so much for this comment! I had no idea you could add a shelf that behaved similarly to the three default options. I was just agonising over how I would file the book I'm on the verge on lemming. I want to record the attempt made, but not count it toward my 2017 reading challenge...."

The reading challenges don't care what shelf a book is on, only books with a date read within the year will count.

Deborah wrote: "The reading challenges don't care what shelf a book is on, only books with a date read within the year will count."

If you abandon a book why would you mark it as read?

AndrewP wrote: "Deborah wrote: "The reading challenges don't care what shelf a book is on, only books with a date read within the year will count."

If you abandon a book why would you mark it as read?"

Right, the exclusive dnf shelf solves this problem.

You have to also NOT give it a date when you shelve it under DNF/abandoned/whatever you use, so if you've just added the shelf, you'll want to go back and clear the date-read on those books in addition to changing the shelf.

Does the "start reading date" impact this, or just the finish date?

Just the finish date in my experience, but you could probably try it out with a test book, and then delete it once you see any effect (if any).

The xkcd star system seems distressing similar to my own.

My use of the rating system is a little different from most. I only expend the effort to look up and rate my favorites and even then only occasionally. So my ratings are skewed heavily to 5 stars, but only include a tiny percentage of my reading. I'm hoping the 5s send more people to those books.

Mark wrote: "The xkcd star system seems distressing similar to my own."

But they're talking about aggregate ratings, which runs into the problem of griefers down-voting books because the author's an "SJW".

The last 3 posts are all examples of why stars are meaningless unless you know the reviewer and what they mean by those ratings.

Aaron wrote: "My use of the rating system is a little different from most. I only expend the effort to look up and rate my favorites and even then only occasionally. So my ratings are skewed heavily to 5 stars, ..."

The thing is, I'd write off your reviews because "he rates everything 5 stars." Remember, no one reading your reviews or looking at your profile knows that you read a lot of things that, if you rated them, would rate 1-4 stars. We only see someone who seems to love everything.

Seems like Goodreads (or one of you smart programmery types) could write a script that searched the entire database to find people whose ratings were most similar. Basically taking the "compare books" feature to the next level by adding a search.

So if a person had rated 100 books, it would find those readers whose ratings were most similar for those same books. I guess kind of like a "book dating" comparison. If two people's tastes matched up for 91 out of those hundred books, it would return a high match value.

Then if one of those people had then rated another 300 books, you could reasonably assume that the books they liked might be ones you would like, as well. That way it automates finding reviewers whose taste matches your own.

Of course, people being people, two completely opposing viewpoints might settle on the same rating for different reasons. In Sean's example, a conservative might rate a book as 1 star because the author is an SJW, while a gay black woman might rate it 1 star because the author's attempt at inclusivity rang hollow and tone deaf to her, while a neutral reader merely felt the writing was inadequate.

So you'd still have to make that judgement call. But the heavy lifting of finding people with similar tastes could be done via program that sifted ratings into degrees of compatibility.

The problem, trike, is the 3 posts above my last one and similar... people don't MEAN the same thing by the same rating. So, yeah, it wouldn't be hard to do what you say. It's interesting idea, but it's what you highlight here "In Sean's example, a conservative might rate a book as 1 star because the author is an SJW, while a gay black woman might rate it 1 star because the author's attempt at inclusivity rang hollow and tone deaf to her, while a neutral reader merely felt the writing was inadequate."

Now, if there are enough points of comparison, like 100, your idea would work pretty well I think. Which means it would work best for people who rate a lot of books.

hmm... I might play with this...

There are five topics I refuse to discuss online anymore:

1. Politics
2. Religion
3. The correct pronunciation of "gif"
4. Any discussion about genre
5. Rating systems

I'll refer anyone who is actually interested to an analysis of Goodreads ratings I did 3 years ago. https://www.goodreads.com/topic/show/...

tl;dr version:
1. Find people with similar tastes and see what they think.
2. When you read something, tell others what you think with a review. Good or bad. Be the reviewer you'd like to read.
3. No matter what you do, some of the books you read will be crap. Accept it, let others know what you think, and move on. There are still plenty of good books out there to read.

Fine but 1) isn't simple, 2) doesn't bear on ratings although I agree entirely, 3) isn't the point.

I wonder what would happen if the rating system was binary. Thumbs up = 10, Thumbs down = 0. Up means "I liked it" and down means "I didn't like it".

YEs, we'd lose gradations of how much an individual rater liked or didn't like a book but as an aggregate score I think it would be more honest and if someone wants to discuss how much or little they thought of a book, they can write a review.

Rick wrote: "I wonder what would happen if the rating system was binary. Thumbs up = 10, Thumbs down = 0. Up means "I liked it" and down means "I didn't like it". "

I like it, but it needs a "meh" rating.

Then a slightly better than "meh" rating
and a slightly worse than "meh" rating"

oh wait :-?

;-)

The problem with that is, there are 10 kinds of people who would like that system, those who use binary and those who don't.

John (Taloni) wrote: "The problem with that is, there are 10 kinds of people who would like that system, those who use binary and those who don't."

I lol'd.

Rick wrote: "The thing is, I'd write off your reviews because "he rates everything 5 stars." Remember, no one reading your reviews or looking at your profile knows that you read a lot of things that, if you rated them, would rate 1-4 stars. We only see someone who seems to love everything."

How would you know I rate everything 5 unless you go to the extra work to look at my profile? Checking each reviewer's profile is more work than many people put into checking a book. I suppose I should add my star system explanation to mine for those who do.

My rating use is mostly for people who know me and those who are less critical. I also rarely write reviews, but I've begun to change that with recent ratings/reviews.
On the flip side, I rarely look at star ratings or read the reviews of strangers. Most reviews I have read don't tell me enough about how the reviewer's tastes compare to mine (the most important element) and we all have different opinions about what constitutes spoilers.

Goodreads could add a field to profiles to allow people to explain their star system and have that be a hover pop-up in the review to cover this.

Aaron wrote: "Goodreads could add a field to profiles to allow people to explain their star system and have that be a hover pop-up in the review to cover this."

I actually see some people do this in their "About Me" section.

Aaron wrote: "
How would you know I rate everything 5 unless you go to the extra work to look at my profile? Checking each reviewer's profile is more work than many people put into checking a book. I suppose I should add my star system explanation to mine for those who do...."
Granted, I'd have to look at your profile. That's assumed. Someone who just looks at one of your reviews won't know - but someone looking to add people as friends so they can see others opinions likely will look at your profile.

The thing is, that I personally discount almost all 5 star and 1 star reviews. VERY few books are really perfect or close to it, so I tend to view those people as either overenthusiastic or the opposite.

"My rating use is mostly for people who know me and those who are less critical."

But we're talking about the utility of ratings to the community. As for 'less critical' the entire point of star ratings is to separate the outstanding from the very good from the good from the bad. Praising everything as 5 stars removes that. If we just want a Recommended/Not Recommended world we'd be better off with a thumbs up/down rating as Thumbs Up doesn't also say 'This is an amazingly good book" in the way that a 5 star rating does, just that the reviewer feels it's worth your time and money.

"Goodreads could add a field to profiles to allow people to explain their star system and have that be a hover pop-up in the review to cover this. "

Or, you and others could use the stars as they're defined vs trying to redefine the meaning. But people DO NOT DO THAT. No matter how many times one pleads with reviewers to use the definitions provided, people do not do it. The situation is also worsened because other prominent sites (Amazon, for example) use different definitions. If you're used to what stars mean on Amazon, you'll use them 'wrong' by the lights of the GR definitions. That's why any graduated rating system open to the public is useless. With a stable of professional reviewers you can enforce the consistent usage of a system. With a public site like this, you cannot.

As for the binary comments above... Yes. Also no.

David wrote: "Aaron wrote: "Goodreads could add a field to profiles to allow people to explain their star system and have that be a hover pop-up in the review to cover this."

I actually see some people do this ..."
The other issue with this is that no matter how one person uses the system, they're all averaged into the single, aggregate score we see on a book's page and those differences in meaning are lost.

I don't know if people use BoardGameGeek much, but that's another site with specific meanings for the number ratings for games (which you'll see when you go to rate a game). The ratings are supposed to indicate how eagerly you'd play this game again, which caused some pseudo-controversy when people gave 10s to legacy games like Pandemic Legacy (which is a game that can only be played so many times and thus CANNOT be replayed after a time).

People are always going to use the stars for their own purposes (I know some Goodreads folks will apparently give 1 star to book that they're looking forward to, just to add them to their shelves, instead of just hitting "want to read"--so I see lots of random 1 stars that literally don't make sense, but they bring down books' ratings, which makes me feel sorry for authors who have few other ratings on a book, because their books will start looking unattractive (how many people have second thoughts when they see a book with a 2 star average?)

Well Goodreads IS Amazon now, right?

But as you said in the original post, I too focus on the reviews of people I follow because either we have similar tastes or I trust their reviews to be informative and appropriately critical.

David wrote: "I know some Goodreads folks will apparently give 1 star to book that they're looking forward to, just to add them to their shelves, instead of just hitting "want to read"--so I see lots of random 1 stars that literally don't make sense, but they bring down books' ratings, which makes me feel sorry for authors who have few other ratings on a book, because their books will start looking unattractive..."


BING! This is why stars are bad too.

GR also allows people to rate books that aren't out for whatever reason. I wish they'd at least make it not possible to rate books that aren't published unless you check a "I have an ARC" box and even after published, unless you have checked off that you've read the book (people could lie of course but they couldn't just click a rating).

Hi,

just want to add an author's viewpoint.. albeit a humorous one.

Here is a review by Patrick Rothfuss on his unpublished book which has so many 5 stars. https://www.goodreads.com/review/show...

Stuart

Rick wrote: "GR also allows people to rate books that aren't out for whatever reason."

A potential issue with this is that most book pages are either A) created by an Amazon algorithms and rarely ever corrects it even when the release date changes or B) Goodreads librarians (like my humble self) could, by accident or design, change the release dates for books.

Becoming a GR librarian isn't hard (a superlibrarian is where the money's at, though), and there's currently nothing in place to lock in a release date.

Plus, release dates aren't always exact even when they're "right"--I remember last year my library got Fran Wilde's 2nd book & had it ready for me--a full week before the actual release date.

I'd certainly like the superlibrarians to have a rating-clearing function especially for book pages like The Doors of Stone which has over 2000 ratings and over 300 reviews. REALLY?

BUT: As long as the Goodreads staff sees the site and themselves as a social media site (which they do), they are NOT going to put that much effort into making accurate catalog pages for the books.

I've mostly made my peace with it (by avoiding books that attract that attention), but it's annoying, and you're NEVER going to convince Goodreads to change the way they do things, not when 90% of people are fine with it as is, the bums. :-)

David wrote: "Rick wrote: "GR also allows people to rate books that aren't out for whatever reason."

A potential issue with this is that most book pages are either A) created by an Amazon algorithms and rarely ..."

OK, but we definitely DO know when a book has been published, yes? if so, then you disallow ratings before it's published, allow them after. Knowing when a book *will be* published isn't required.

Most of this could be alleviated (for people who care) if I could follow reviewers and see the rating of the book from just those people. Something like this:

GR: 3.94/Your Friends: 3.67

If none of my friends have rated it:

GR: 3.94/Your Friends: --

Rick wrote: "I wonder what would happen if the rating system was binary. Thumbs up = 10, Thumbs down = 0. Up..."

Netflix started doing this a few months ago when it comes to outward-facing reviews, but they also kept the 5 star rating system for personal ratings. I think that was the best solution for an imperfect system: everyone grades on a pass/fail while for your own purposes you can distinguish between similar movies.

So for me Citizen Kane and Tremors are both Thumbs Up movies, but I'd give the former 5 stars and the latter 4 stars. Both great movies that I will watch at the drop of a hat, but Welles' opus is just so great on so many levels that I need to distinguish between them for my own use.

The other downside of an odd-numbered rating system (which is usually 5, as it is here) is that over time ratings tend toward the center option.

In corporate reviews, the better option is to have a 6-slot rating, which forces people to choose something other than average. Because a lot of folks are just trying to be nice by not rating something too harshly, so they default to 3/5. If the ratings need to be actionable, then that's unhelpful.

Rick wrote: "OK, but we definitely DO know when a book has been published, yes?"

It can differ between markets, too--I've had some books on my TBR list that haven't come out in the US even though it's been out in the UK for example for 5+ years.

I don't know how many reviewers you follow or people you're friends with, but I definitely see the ratings of people I follow or are friends with below the main book information (at least on the website). We don't get an easy-to-see average of their rating, but I can see at a glance that, "Oh, 2 people I follow gave that book 1 star, and 3 others gave it 4, and here's a couple reviews from them, great."

David wrote: "I definitely see the ratings of people I follow or are friends with below the main book information (at least on the website). We don't get an easy-to-see average of their rating, but I can see at a glance that, "Oh, 2 people I follow gave that book 1 star, and 3 others gave it 4, and here's a couple reviews from them, great."
..."


Oh yeah I can see that. And it's helpful. I was talking about an average score, but unless a bunch of friends had rated it...

Rick wrote: "OK, but we definitely DO know when a book has been published, yes? if so, then you disallow ratings before it's published, allow them after. Knowing when a book *will be* published isn't required."

True, but aren't actual early reviews still valid? ARCs, NetGalley, other early review copies?

I don't think that date is necessarily the way to filter for 'valid' reviews.

Rick wrote: "
The thing is, that I personally discount almost all 5 star and 1 star reviews. VERY few books are really perfect or close to it, so I tend to view those people as either overenthusiastic or the opposite.
...
Or, you and others could use the stars as they're defined vs trying to redefine the meaning."

The official Goodreads rating system as defined by hovering over a number of stars is:
1 did not like it
2 it was ok
3 liked it
4 really liked it
5 it was amazing

Therefore, rating my favorite books as "it was amazing" is the correct use of the rating system and those who define 5 stars as "perfect or close to it" are the ones redefining the rating system. Something doesn't have to be perfect to be amazing. Perfect would be the unattainable stand of 6 stars.

People use Goodreads in so many different ways. Even here in our Sword & Laser groups, there are some active members who barely have any books in their profile at all... and meanwhile I've got every single book I've ever read from June 2007 on, and an additional 800+ books from before then that I know I've read but don't have a finish date for. I do rate everything I've read, though I rarely review.

We've got all sorts here for all sorts of reasons, and it makes Goodreads very diverse and annoying and fun and aggravating, so I'll have to "deal" as it were.

(Just FYI, in case people aren't aware--if you have some ideas or feedback, there is a Goodreads Feedback Group--I've submitted threads and contributed before, and I think something like Rick's "friend/follow average at a glance" might not be a bad addition: https://www.goodreads.com/group/show/...

I don't think anyone will ever convince Goodreads to take itself more seriously as a catalog vs. a social media site, but we can nibble around the edges.)

Rick wrote: "The other issue with this is that no matter how one person uses the system, they're all averaged into the single, aggregate score we see on a book's page and those differences in meaning are lost. "

There's a Firefox plugin that fixes that by changing the star-average into a percentile. Instead of telling you that a book has a 4.1 star average, it tells you that it only ranks above 40% of the books on Goodreads.

ladymurmur wrote: "True, but aren't actual early reviews still valid? ARCs, NetGalley, other early review copies?
..."

Which is why I said that people should have to check an "I have an ARC" checkbox to review before it's released (was in an earlier comment). That wouldn't stop everyone but the extra speed bump will filter some of those people who click stars for whatever reason when they haven't read the book.

Aaron wrote: "Rick wrote: "
The thing is, that I personally discount almost all 5 star and 1 star reviews. VERY few books are really perfect or close to it, so I tend to view those people as either overenthusias..."

I simply don't trust reviews by people who think most of what they read is amazing. I tend to grade on a curve, i.e. about 10-20% of things I read can possibly *really* be amazing. That forces me to think, when I read a book I really liked, "Is this as good as other things I've 5 starred? Is this really one of the best things I've read in the genre? Or is it really just very good?" Now some of these people could be like you where they only rate the stuff they love and don't rate the 1-3 star books.... but remember we only know a reviewer by what we see here.

Also, don't distort my arguments. The point about not redefining the stars was a direct reply to your assertion that "Goodreads could add a field to profiles to allow people to explain their star system and have that be a hover pop-up in the review to cover this. " which is why I quoted that part of your comment.