I use goodreads a lot because I like to read books and want to find more books to read. This morning I was reading a thread on hackernews on Sarah Manavis’s NewStatesman article and it reminds me of how much I want, I yearn, for useful book recommendations. I think about this pretty much every time I open the goodreads page or app and hope that maybe this is the day they finally fix recommendations, so I figure I would capture some ideas I have for fixing their recommendations.

I Love Goodreads

First, I love Goodreads. My profile says I joined May 2012. But I think that’s just whenever I started this profile and I have multiple, old profiles that I can’t find. My memory says I started using them soon after launch, but maybe I’m misremembering. But I’ve used them for a long time because I like to read books. I love browsing bookstores and libraries looking for new books. I experience joy finding a 50 or 100 year old book that I didn’t know about. One of my favorite things is reading a book so good that I can’t go to bed, or want to walk an extra mile. But that’s pretty rare for me, I’m lucky if I get one of those a year.

I have goodreads on my link stack with the tag “purdah” that I think means it is part of the stuff I do that uniquely identifies me.

Any time I get a recommendation from a friend or article or wherever, I add it to my “to read” shelf on goodreads. That helps me when I’m trying to buy a new book. So at least I can read through the 245 books that at least a past me wanted to read. This works well and is better than a vanilla list.

I also like seeing what my friends read and especially what they like or dislike. My friends have different tastes so it’s not as simple as just reading everything they read, but it helps a little. It’s one of the few places that I think facebook helps because at least I get to find out facebook friends who also are on goodreads. Every once in a while I find a new book through this feed, maybe once every 2 years. This might be because I don’t have many friends on goodreads (only 42), but I think it’s also because different people have different tastes in books.

I also like that goodreads will tell me when an author I’ve read releases a new book, or is having a book chat or something. Although I wish they would tell me when they come to a local book signing.

The Problem As I See It

This gets me to what I think is wrong and want to fix. I have read and rated 802 books with my average rating of 3.97. This makes sense as I’m probably only going to read books I think I’ll like, but I think this is the average goodreads rating, not the average of my ratings. Looking at what I rated, I’ve only given out 3 one stars, 42 two stars and tons of five stars. I keep scrolling on and on.

The user interface for goodreads is pretty bad- information is hard to find, layout is wasted, search doesn’t work well, hard to tell ads from user content- but that’s not what I’m talking about here. The main problem is in finding books that I will like. Maybe a UX overhaul from a great designer would help, but not that much. A craigslist interface would be great if they could just solve this one problem…

What is the single book that I will love most in all the world?

I want to find the next amazing experience. Books are great because they are cheap and portable. It’s not like wine where the best is more than I’ll ever pay. Or travel where a first class flight can cost $20k. If I can find a book, I can read it. And maybe someone I know can read it too.

If I can mix in more wonderful books, my life will be better. My family’s life will be better. Etc etc. Maybe this is the answer to world enlightenment. Probably not, but it will be great, I think.

When I click on recommendations, goodreads shows me five books that I don’t want to read. And it’s weirdly bad. Some of it makes since why it’s bad. They recommend different versions of books I’ve read. They recommend two different versions of Lord of the Rings (one of my favorite books), but I guess they don’t know these are the same book. That’s just a waste of time and it’s weird that amazon isn’t smart enough to avoid duplicative books.

Screenshot of goodreads recommendations page

But then they recommend stuff like Stephen King’s The Long Walk. This is perplexing because I only have one book that goodreads knows I’ve read from King and I rated it two stars. I don’t have any of his books in my “to read” list. Why would they recommend this book? Two of my friends rated it four stars and one even wrote a recommendation. So maybe that’s why. I’m aware of this book and don’t want to read it. But other books like Shoe Dog by Phil Knight has no relationship to me. Why would they recommend this? I also know about this book and don’t want to read it. I wish there was some way to note books that I don’t want to read. Then at least they would stop showing me the same book over and over.

So it seems like goodreads is making pretty basic, rules-based recommendations. And they don’t help me pick new books.

My Perfect Wish

My idea for an 11-star experience 1 in finding new books is that Goodreads knows me even better than I know myself and constantly recommends the perfect book. They recommend books that teach me, that spark enlightenment, that I hate but teach me to appreciate other books, books for when I want to be happy, and books for when I need to be sad. I want something like Jane from Ender’s Game who experiences everything I know but is fixated on finding me a good book. So the output of a book sommelier whose passion is seeing me sit there and love a book.

That’s what I would like.

A Little Story About Jeff Bezos

Once I was lucky enough to go to TED. It was great on so many levels, I highly recommend it and 2/3 of the cost was tax deductible. It was full of great stuff. I got to wait in the registration line in front of Neil Gaiman and Amanda Palmer, that was neat.

Anyway, one night Neil Gaiman was doing in a midnight ghost story reading in some neat four story townhouse type building 2 that was totally empty and spooky and a great place for the event. The reading was on the top floor and there was a huge crowd waiting for the single elevator in this place. The lights were off and there weren’t any staff around so it was maybe 100 people waiting to go up three at a time. So I looked around the corner and took the stairs. About 2 or 3 other people had the same idea so we started racing up the stairs to get to the top and avoid the crowds.

Vancouver Club street view photo by Joe Fox

This building was awesome. It seemed 100 years old, rare in Vancouver, and as we went up the stairs we saw each floor. The stairwell opened up to a big room that could probably seat a 50 person wedding dinner. The lights were off, and someone had opened a few windows and the white, wispy curtains were blowing with a little breeze. Even though it didn’t feel like a breeze. I said “cool, haunted house” and stopped climbing the stairs to look around. The person next to me said “yeah, cool” and we started quickly running around checking out the empty room. We did this on the third floor too and me and this random person did quick 60 second explorations of this spooky house.

It all happened quickly. Maybe 3 minutes to go up the stairs, run around each floor, and get to the top. About halfway through I realized this random dude was Jeff Bezos. I thought that was neat. I’ve bought books from Amazon since 1995 and I’m pretty sure I telnetted into it to order some books (although maybe that was cdnow before they bought them), so it was cool to think that Jeff Bezos likes haunted houses and still likes exploring empty buildings.

We got to the top and went into the venue and were maybe the first 10 people. Comically got there maybe after two or three elevators full of people. We went to the front where Gaiman was setting up at the podium. I got a comfy chair right in the front about five feet in front of Gaiman. Jeff Bezos sat in the chair to my left. The chair on my right had Neil’s wife, Amanda Palmer, who was carrying a ukelele case. I didn’t know it was a ukulele case until she told me. She also said she might play, but she didn’t.

Gaiman finished setting up and came up to talk to Bezos. Bezos asked if Gaiman got the books he sent. Gaiman said he did and they were really great and he enjoyed reading them. It seemed like this was the most recent in a series of book shipments.

Aside from nerd joy from getting to witness this small talk of people who interest me, I thought, at the time and many times since,

How cool is it that Neil-freaking Gaiman is still finding new books that are great, and how cool would it be to get book recommendations from Jeff Bezos. I bet he has access to the best book recommendations in the world.

Also, TED has something called the “TED book club” where they mail you this box of books a month before the conference. This was the closest I’ve come to my dream. They maybe sent 10 and 5 are still on my bookshelf. But not the shelf behind me, so I don’t remember them specifically. I’ll look them up one day.

That’s probably better than Jane making recommendations. Amazon had just bought goodreads a year earlier and I was hopeful that any day, goodreads would roll out a wonderful recommendation engine any day now.

What I Want to Do to Fix

What I would like to do is to just get a data dump of mine and everyone on goodreads recommendations. I don’t need to know the identities, just to know what recommendations go with what individual. It would be neat to protect privacy and find individuals to potentially follow them on goodreads, but I think for the analysis I want to do, I don’t need to know the actual identities.

I think if I have this data, I could try to find people who rate a few hundred books like I have, then look at what books they’ve read the most, or books that they’ve rated highest. I think that would be a good signal that I might like it.

It would also be cool to see books they hated. Just to see uniqueness amongst the millions of readers.

It would also be cool to find my exact opposite. People who rate the absolute opposite from me, my antipode. 3

If I just had this data, I would do what goodreads isn’t doing for me. Maybe they’ve already tried and it’s impossible to achieve. Maybe it’s just one of those things where we have to increase opportunities and leverage serendipity.

They do have an api but it seems meant for integrators to do goodreads functions from an app or something. I can get my own recommendations, but there’s no way to get other people’s ratings, and unless I know the name of the book I can’t look up ratings. And that’s sort of the problem since I don’t know the name of books I don’t know. They do allow me to view a specific user’s info and even lets us compare shelves.

I wish they had a data dump somewhere. It’s strange to me that lots of users allow their ratings to be public (eg, this random friend user), so I can theoretically scrape all the reviews from just public users, but that’s a lot of work. I share my ratings because I want to help others. I assume that other users do so for the same reason so I wish Goodreads/Amazon would help facilitate this. Since userids are just 8 digits, I supposed I can use their API to read one user a second (their requested rate limit) and it would just take me and it would just take me 3.17 years, or maybe less since they may not have 99,999,999 users, but Statistica claims they had 90 million registered users on July 2019. 4

Housekeeping

Unanswered Questions

  • Low quality of reading random people’s reviews and comments

Future Blog Ideas

  • Describe the link stack idea.
  • Emergent awesomeness. When you can’t explicitly do something, you just have to increase the rate of random awesome things and appreciate them when they happen.

Stuff I Read While Working on This Post (That You Might Want to Read Too)

Footnotes

  1. https://mastersofscale.com/brian-chesky-handcrafted/ Masters of Scale podcast with Brian Chesky where he describes AirBNB process of trying to design experiences where users rate something 11 stars (out of 5). I find it funny that AirBNB has never given me an 11 star experience, but lots of 5 stars. And I think that’s the message here. 

  2. https://vancouverclub.ca/ I think it was The Vancouver Club, so imagine this place but at midnight, empty and abandoned. Wikipedia says it was founded in 1889. 

  3. https://en.wikipedia.org/wiki/Antipodes I would also like to see this antipode functionality for twitter. Who is the person with the exact opposite of me on Twitter. Would be neat to read their ideas. 

  4. https://www.statista.com/statistics/252986/number-of-registered-members-on-goodreadscom