The murky math of the New York Times bestsellers list | Hacker News Reader

The murky math of the New York Times bestsellers list

thehustle.co

90 pointsAnon8453 comments

What’s wild here isn’t that the list is totally bogus, but how little money it takes to game the system. People spend $150k on an MFA in creative writing when $60k will buy a consulting service to land you on the NYT best seller list.

be$t$eller$ li$t

Startups do the same thing. They spent too much on R&D when they should be spending the money on marketing.

Paul Graham's Submarine article talks about this: http://www.paulgraham.com/submarine.html

Tbf you can get it for a lot less than 150k

I have started looking at Goodreads ratings as a good proxy for well written books but there are a lot of bizarrely high rated pop culture books so it’s not perfect. Another strategy is to look at books shortlisted for the major awards. That’s how I came across North Woods by Daniel Mason (Pulitzer short list) which is by far the book I enjoyed most this year.

https://www.goodreads.com/book/show/71872930

>I have started looking at Goodreads ratings as a good proxy for well written books

Based on my experience looking at Goodreads reviews, I think you'd have just as much luck with a random number generator.

I think it is better for ruling things out rather than shortlisting. A rating below 4.0 is a guaranteed bad read (IMHO) but above 4.0 is not guaranteed of a good read.

Many books I've liked have <4.0 on Goodreads. But you're kind of right that I will usually take that as a bad sign that needs to be outweighed by some other reason to read it (like a personal recommendation or whatever).

A lot of my absolutely favorite books are rated below 4.0. A few examples:

1. White Noise at 3.86

2. Catch-22 at 3.99

3. Dhalgren at 3.77

4. At Swim-Two-Birds at 3.86

5. Midnight's Children at 3.98

6. Омон Ра at 3.90

7. Never Let Me Go at 3.84

Drawing the line at 4 rules out some of the most interesting books.

In my experience, you really have to fall below 3.5 to get consistently bad books, and even then it's not 100% guaranteed.

Woolf’s To the Lighthouse and Orlando are under a 4. But under a 4 is guaranteed bad? LOLWUT.

And that’s just the first author I checked.

Vonnegut’s Deadeye Dick (less of a crowd pleaser than his handful of biggest books, but certainly not in the bottom tier of his works) is under a 4. The A Man Without a Country collection, which is more of interest as a curiosity than because much of it’s good, is over a 4.

Yeah, you may as well flip a coin.

Anything assigned in a high school or college English class gets a huge ding, because people who don't want to read it are forced to and, predictably , hate it.

Generalised rating systems aren't really that great. Given how general they are.

Imagine if a book is advanced Sci Fi, as in the concepts are built on exiting and other well known scifi books. The normal reader won't understand and think it's terrible. The actual target audience, Sci Fi book nerds, would love it.

I find it most useful for the similar book recommendations. Just to create an easy to make list that I can then go through to pick out what I want to read.

There was an article on HN recently, "Apex Books of Goodreads" https://news.ycombinator.com/item?id=37970307, in which the author uses a Pareto frontier to attempt to account for the recency bias. It was not entirely successful, as the result includes four of the Harry Potter series and five Calvin and Hobbes collections, including the highest-rated The Complete Calvin and Hobbes, from 2005, and second is Brandon Sanderson's 2014 novel Words of Radiance (The Stormlight Archive, #2)

Nothing wrong with including Calvin and Hobbes five Times.

Bill Watterson is indeed a treasure. Are his works equal to that of Tolstoy, Gabriel García Márquez, Murasaki Shikibu, Wu Cheng'en, or Attar of Nishapur?

the way to use goodreads is to follow people you like and see what they read.

Off by a factor of 100 here:

> Making the list is a herculean task. Consider the number of books, more than 3m, published every year. And then consider that the number of slots on the NYT’s coveted lists is ~6240. That gives authors a coolly aspirational 0.00208% chance to snag one.

For anyone else who's as bad as me at mental arithmetic:

3,000,000 ÷ 6240 = 0.00208 = 0.208%

No it’s 6240 / 3 000 000

Ah, right.

The murky math of someone trying to write about the New York Times bestsellers list and getting it wrong by a factor of 100

I pay zero attention to this list. If it’s not actually a list of best-selling status, it’s just another recommendation from the NYT. I don’t need a “Wirecutter for books”, since I already have plenty on my list.

I know that among Dutch writers, it is a public secret that one does not have an English translation because the US (and rest of Anglosphere) is an interesting market, but it's a requirement for getting on the NYT bestseller list. This means worldwide visibilty and worldwide sales. I never managed to get hard numbers from anyone, but reading between the lines, Dutch and American book markets appear to be on the same order of magnitude, not necessarily in favour of the US.

788 million us, vs 44 million, taking top google hits.

I've wondered about that as an author, actually:

Given how little Americans read books, might there non-English-speaking countries where there are more book readers than the US?

It would be simple to get a BAD translation of my books into Spanish, say, but a good translation would cost me a lot. And I wouldn't know how good it is without showing it to several other native Spanish-speakers.

I was surprised that some author said that on the Dutch TV show "Lubach" earlier this year. He stated that the (absolute) number of copies sold to get in the top list of the Dutch market is about the same as the number of copies that gets you in a NYT list.

He explained this by saying "Americans don't read much" and you seem to imply the same. But it appears that in the US about 300,000 titles are published per year (2013 data) versus 16,000 in the Netherlands. So is this an apples-to-apples comparison ? (source https://en.m.wikipedia.org/wiki/Books_published_per_country_...)

It wasn't about the size in terms of titles, but total market in currency, AFAIK.

Plenty of "internet marketers" sell each other a way to make it into the list cheaply, and it seems to work given that their bs does make it..

I like (current) bestseller lists: they usually contain all books not worth reading.

Looks just as manipulated as view counts on YouTube or like counts on social media. Sad, really.

> Looks just as manipulated as view counts on YouTube

Can you elaborate? I thought they are mostly accurate nowadays.

or University Rankings, similar murky maths all the way to outright fraud

See https://en.wikipedia.org/wiki/Criticism_of_college_and_unive... or https://en.wikipedia.org/wiki/Times_Higher_Education_World_U...

The numbers are wrong, 6000/3000000 is 0.2% not 0.002%. I stopped reading after this.

PACs have been doing these bulk buys for years to get people on the list. As the article mentioned, Mitt Romney did this.

Mike Pompeo did it as well. https://www.forbes.com/sites/zacheverson/2023/02/21/mike-pom...

Mike Pence https://www.forbes.com/sites/zacheverson/2023/01/09/mike-pen...

Members of the Trump family did this multiple times. https://www.nytimes.com/2019/11/21/books/donald-trump-jr-tri...

It's common practice for political parties and PACs to do these bulk buys and then sell them or give them away during events or conventions.

Curious why you chose only conservative politicians. Do progressive politicians refrain from the practice?

Jim Wright (D-TX), then Speaker of the House, got some grief over bulk purchases of his memoirs. The Wikipedia article on him mentions this.

I had to stop reading at "Latinx". Stop trying to make latinx happen, nobody wants it.

https://www.pewresearch.org/hispanic/2020/08/11/about-one-in...

You're not their audience. Mission accomplished.

It works as a tribal marker.

Using it shows you're willing to humiliate yourself to show tribal loyalty.

"Please don't pick the most provocative thing in an article or post to complain about in the thread. Find something interesting to respond to instead."

https://news.ycombinator.com/newsguidelines.html

A total work of fiction basically, like almost everything else written in the NYT.

The NYT doesn't get everything right, but they still have a strong reputation for factual reporting. when they do get something wrong, they generally print a retraction and it is a big deal. I don't think it is fair or factual to say that almost everything they write is fiction.

The problem is they consistently make mistakes in the rush to get the story out first. I suppose that's not much different than many other publications, but the NYT is the most visible. It tends to erode their "strong reputation for factual reporting," and for large swaths of the nation (mostly conservative), they're little better than a tabloid.

Here[1] is a detailed, thoughtful article demonstrating specifically that many outlets — including the NYT — make extremely significant errors increasingly often, using a recent major mistake as an example.

And not only do they generally not publish retractions, but nobody reads retractions. Willingly erring in specific, predictable ways repeatedly and then occasionally issuing retractions is a great way to manipulate the perception of readers, which is one of their goals.

Almost everything they write is fiction: they only report facts in the approximate area of the truth that support their preferred frame. This combination of omission and slight of hand is enough to craft stories that are, in the final analysis, completely false.

[1]: https://www.silentlunch.net/p/did-the-entire-media-industry-...

Reply on news.ycombinator.com