A novel about video games became a surprise best seller | Hacker News Reader

A novel about video games became a surprise best seller

nytimes.com

78 pointsbenbreen64 comments

"When her literary agent, Douglas Stewart, sent out the manuscript in January 2021, he felt in some ways as if he were introducing her for the first time. “Despite the fact that Gabrielle had several big successes before, there are people who had never heard of her before this,” he said. A frenzied bidding war with 10 publishers broke out, and Knopf won, paying a seven-figure advance. Shortly after, an auction for screen rights drew 25 bidders, and Paramount Pictures bought them for $2 million."

How does this work? An unknown author suddenly gets bid on for loads of money before anyone bought the book?

"Knopf, which initially printed 60,000 copies, has reprinted the book 21 times to keep up with runaway sales. "

Company paid more than a million and prints only 60000 copies? How do they intend to make the money back?

This makes no sense to me.

Also, nowadays I don't believe in grass roots anything, what I believe in is clever marketing on social media.

According to the article, the book has sold almost a million copies since last year. So Knopf has definitely made their money back.

The size of the initial print run only tells you that they were still working out the marketing campaign. Doing smaller runs first lets you experiment with parameters like cover and format and media campaign timing.

It only makes sense to do a massive initial print run if the demand is both guaranteed and transient, e.g. a book that’s part of the national discourse right now but will be forgotten in six months.

Thanks. Now it makes sense to me to make a small run first before printing more. Doesn't explain why they immediately paid the author so much money. Are best sellers made by publishers by investing money in marketing it (like what this article is doing)?

This is how the industry works. People make bets based on their estimation of how a book is going to fly...and nobody really knows for sure.

Andew Wiley (a well known literary agent) also makes the observation that the bigger the advance the more assiduously the publisher will work to justify it.

In Hollywood this is called "buying your gross". Marketing costs are a major portion of revenue for a blockbuster.

This is how the publishing industry works. Authors get a guaranteed up-front advance and then also a backend residual dependent on performance. In fact this is how many industries work; as an engineer I get an up-front guaranteed salary, but then I also get bonuses and stock refreshers in arrears as reward for good performance.

As for why the industry works like this, this is the income that professional writers live off of. Most of them aren't coming out with even one book a year, and they need some form of guaranteed income to live off of. The publishing company is fulfilling a role here not too dissimilar from VCs, potentially losing money on most bets that don't win big but then hitting the jackpot on something like this novel. It's risk aggregation; what works for a profile of many bets is not good when you're the author with one bet.

Does that explain why every surplus store winds up with so many copies of politician's autobiographies right after they leave office for good? I'll never forget as a kid seeing aisles & aisles of nothing but Bill Clinton's "My life" for $2 per copy or something like that.

Political books are a different animal than other books. You'll find political donors buying a huge number of copies to propel the book to the top of the charts to make a political statement. (my ideas can't be crazy if 50K people bought my book on the first day!)

Politicians use the books for fundraising and can end up with a lot of surplus books. I suspect that the donors that purchased Clinton's book needed to free up some garage / warehouse space and dumped them when they were no long a useful political tool.

She might be initially unknown to some publishers, but her introduction by her agent is really effective: ”She is a published author, who sold hundreds of thousands of copies and won some awards with her previous books”

That’s a lot different from a real unknown that had never published a book.

Yeah like -she wasnt' famous and now we made her so or whatever

Read the entire article. This is her ninth published novel, and some of her previous novels have sold hundreds of thousands of copies. And I've heard of at least one of them, Elsewhere. She's not unknown in literary circles.

To save you a click: Tomorrow, and Tomorrow, and Tomorrow by Gabrielle Zevin

https://en.m.wikipedia.org/wiki/Tomorrow,_and_Tomorrow,_and_...

The book was heavily pushed on tiktok because it hit the platforms target demo perfectly with the angsty young characters set in a video game backdrop that the average tiktok age group could understand.

Personally I was hoping it would have been more like the backstory episode of the Mythic Quest TV show and was disappointed. It was an average story at best with characters that never really matured, but that doesn't matter because I'm not the target audience. The target will connect with the characters and come away thinking it was a fantastic story and will gobble up any movie/tv show that comes out of it. Good job, author and marketing.

Fwiw I read it for a work book club and liked it, though it did feel pretty similar to Masters of Doom and Soul of a New Machine, both of which I've also read recently. Those two were cited as influences in this book, are non-fiction, and I found them more interesting (maybe because they were non-fiction).

Masters of doom is a highly researched memoir of the gaming company iD software and also gets into some fairly technical details of how the first 3-D game engines worked (ray casting, binary space partitions, etc)

It's probably one of my favorite memoirs, I can't imagine this book which seems to have a more focus on interpersonal relational fictional drama being in any way similar.

Masters of Doom has plenty of interpersonal drama in it as well. Hell, John Carmack and John Romero are the two main characters, and most of the other id Software early employees end up getting fired. I suggest you re-read Masters of Doom (or at least read this synopsis https://en.wikipedia.org/wiki/Masters_of_Doom#Content ), also read Tomorrow^3, and you'll see the similarities.

This book was really disappointing. The characters were all underdeveloped, remaining angsty melodramatic teenagers until the end despite them all being adults. It also tries to deal with some heavy themes but fails due to the immature characters. There is a lot of exposition to keep the plot moving, which is simply bad writing. Show don’t tell!

When I was a teenager I sometimes didn’t want a character that grew and matured. I wanted a character that validated my angsty melodrama.

I guess if the target demographic of this book is young angst driven high schoolers then that makes sense.

> In March 2023, game designer Brenda Romero told The Washington Post that the game Solution represents a substantial, uncredited appropriation of Romero's own game, Train...

> “I have no doubt that Train is the best game I’ll probably make,” Romero said. “It is the one thing I will have to show for dedicating my life to games. And somebody decided that was just fair game.”

Don't look at the wikipedia page for Train unless you want a huge spoiler. Train has a VERY specific twist. It sounds like that was what was used for the novel, and Brenda Romero is right, especially as other games were credited.

And it's disgusting and cruel. No, the ends don't justify the means.

And Romero stole the idea from a famous Orson Scott Card book anyway.

What does ‘it’ refer to in your first sentence?

https://archive.ph/YoiQ5

I enjoyed the book but it read like something by a well researched outsider, not someone with first hand game dev experience. E.g, references to 3.25" disks.

Haven't read it (yet); 3.25", not 3.5" or 5.25"?

Did she mean 3" discs for Amstrads?

I know what you mean, but this didn’t actually ruin it for me. As a video game player doesn’t work in the industry I actually learnt a few things about games design and process. I suspect anything written by an industry outsider is going to feel like a bit clunky when read by an insider, especially when the author is undoubtedly optimising for accessibility to a wider audience.

It could very well be that the things you learned are in fact wrong (haven't read the book but talked to a few other game devs who did, who all pointed some issues with how it portrays the process of game development)

Don't get a good sellable story get in the way of the truth

I suppose it's the equivalent of thinking Tom Clancy novels give you truly factual insight on militaries:)

They also code SO HARD that they burn through graphics cards multiple times. =)

It's a really interesting story, but it could've used a fact checker to get a few of the details right.

Oh, and the MC dude gets really jacked at one point - but it's never referenced past a sentence or two in the middle of the book =)

(Read it for a book club and we dissected this for hours)

> They also code SO HARD that they burn through graphics cards multiple times. =)

Can you elaborate on this a bit more? I honestly don't understand what that even means.

Since you mentioned that you've read the book, would you also recommend it?

I'm currently dabbling bits in some indie game dev, so I'm wondering if I'd be the target audience.

> Can you elaborate on this a bit more? I honestly don't understand what that even means.

Small manufacturing issues that usually don't get noticed, can fail a card if you put it through 100% load for a long period of time. Cryptocurrency miners regularly would downclock cards for cost reasons (energy/$) and because consumer GPUs also aren't made for sustained 100% load (across days).

GPUs are plenty of capable running 24/7 without issues for a long time.

Plenty of cloud providers have consumer grade GPUs.

Cryptominers don't have an issue running it at 100% either.

They have heat issues and power efficiency issues not GPU Chips breaking.

The only instance I know are if you use a very particlar load type which would normally never happen like 24/7 hitting one register only or moving data only through the bus without anything else.

Those issues are rare.

> GPUs are plenty of capable running 24/7 without issues for a long time.

Running, yes. But running 24/7 at 100% load for days without any sort of break or underclocking it is a surefire way of making it fail quickly. If you know any people doing cryptocurrency mining, ask them :)

The guy you're replying to knows when those issues pop up and has relevant experience here. I'd listen to him. I too have experienced hardware failure at 100% load for months straight, and have resorted to underclocking for power efficiency and longevity reasons. The performance/cost/longevity curves are not linear. Just take a little bit off the top and you do much better.

Ethereum miners often underclock the Core and overclock the Memory in their GPU because it’s such a memory intensive operation.

I didn't understand it either, because it literally doesn't make sense. The author used it as a vehicle to show how hard the protagonists were working in a few different occasions.

It's a story of two people who like each other, but refuse to communicate (because they're both broken in their own ways). It's set on a backdrop of gaming history over a few decades, maybe 80/90's to current day-ish.

There are a few anachronisms and weird tech issues (they do crap with computers that wouldn't be possible at the time), but as a whole I'd recommend it.

It's far-fetched and only happens if you don't cap the framerate but it does happen:

https://www.gearrice.com/update/blizzards-newly-introduced-d...

https://www.denofgeek.com/games/new-world-bricked-gpu-causes...

Neither of these are in the context of coding though.

I haven't read the article, nor the book, while likely a mistake, 3.25inch disks did exist

https://en.wikipedia.org/wiki/Floppy_disk_variants#Drivette

I know the games and companies involved were "alternate reality", I'd expect slightly smaller disks aren't so far-fetched. One does wonder if the writer knew that at the time.

I have ever seen the cover of her earlier novel "Elsewhere" somewhere. Didn't know about her otherwise.

Turns out she has multiple novels the plots of which are quite interesting to me.

I am definitely going to try out this novel, Margarettown, and The Storied Life of A.J. Fikry. And maybe even Elsewhere, too.

Gold Rush! by Sierra still does exist but you can't easily buy it. It will work in scummvm or you could use an emulator. Unlike books it is much harder to access and use older computer programs.

"Gold Rush! Classic", $2.99 on Steam. Runs in DOSBox just fine.

That appears to be based on a re-release after the rights reverted to the original developers, rather than the original build published by Sierra.

Yeah, this part was kind of puzzling :

> But when she tried to play the adventure game Gold Rush, she discovered that the version she had played obsessively as a kid no longer existed.

What do they mean by that ? That she failed to ask around ??

Well reviewed, i was just about to start reading it :)

3 authors i interviewed recommended it too… https://shepherd.com/book/tomorrow-and-tomorrow-and-tomorrow

The recent Tetris movie seems somewhat similar ? It was surprisingly pretty good !

Hijacking this to ask for help on my own side-project.

I want to write a story to help young adults (i.e. 10-16 years old) understand how programs work, and the difference between traditional programs and machine learning.

My idea is to tell it from the perspective of AIs in a computer game of Monopoly. An ML player is introduced, initially clueless, but improves between iterations until it dominates and destroys the game. The game is upgraded with an ML rulebook that returns everything to order.

But I've never written a story before, and I don't know much about how ML. I'd love to get suggestions, advice, feedback, etc.

It's an interesting concept, although I would argue that Monopoly is not the best choice of a game for this. Outside of 'do I buy this property', auction bidding, should I mortgage this property, or 'when/where do I buy a house', there's no decisions to be made in the game. And about half the turns (especially mid-late game) are just 'roll the dice, move, land on someone else's property, give them the money I owe'. I understand that Monopoly is super popular, but there's probably a game that would fit your story better.

And this is already a computer game version you're talking about. Is there a reason why you want it to be a board game adaptation instead of a video game?

Also there's a chance if you focus too much on one popular and existing and in-print game it might not fall under fair use and you could possibly get a cease and desist from Hasbro.

One of the reasons why you see more shows around chess and not Monopoly, chess is in the public domain and no one is going to get sued over using it.

As for learning about ML, I'd suggest taking an online course or watching some Youtube videos so you understand some details about it. The videos could inspire some ideas for plot points in the story as well.

There's tons of videos/books/articles on writing also, but really for that I'd just dive in and get started, post some of it somewhere, get some feedback, and do some editing/rewriting. Maybe even get some recommendations from ChatGPT or NovelAI or something.

If you're willing to critique other people's work, there's a really good website to get feedback called Critique Circle that has a quid pro quo model (you critique other people's works, get some credits, then spend them to solicit critiques for your own work). Worth checking out. Reading your story out loud is also a good trick for making sure the story sounds smooth and flows well.

Good luck!

Thank you, that's brilliant feedback.

Good point about the licensing, I hadn't thought of that. I guess I wanted to choose a game where I don't need to spend half the book explaining the game mechanics just to get the meat of the story. I thought of Monopoly because there are some interesting loopholes an ML might try, like offering the same trade 10000 times. I suppose I could think about how this might work as a character in Age of Empires or Minecraft.

I'm doing a fair bit of research on ML at the moment, though I'm mindful of not using that to procrastinate on actually writing the story. I do want the ML concepts to be necessary to moving the story forward, not just a gimmicky buzzword dropped in to make it sound technical. I'm really not sure how I'm going to do that, though - if I make it too complex, it will probably a poorly written confusing deep dive on calculus.

Thanks for the recommendation of Critique Circle, I'll check that out.

> I've never written a story before

You need to read and write more fiction if you want to tell an effective story. Just a description of the differences isn't a story. A story usually has a plot, character development, and worldbuilding in addition to themes. You can get away with not much character development or plot, but doing so in middle-grade/YA is not really viable. A hard sci-fi book with cardboard characters can work if it's pitched at nerdy, tech-literate adults. Likewise, a book without a cohesive plot but with good characters, themes, and emotional resonance is usually pitched at adults and is enjoyed by a fraction of the audience of popular literature. A good middle-grade/YA book needs to be accessible to a population which is not well-read and that needs the scaffolding a more conventional story structure provides.

Also if you've never written before, you should know that the younger your audience, the harder it is to write for the unexperienced author. Writing for an audience you're a part of (e.g sci-fi fans, romance fans, etc.) is much easier than writing for one you're not a part of. For children in particular, you need to have at least some idea of child development, appropriate vocabulary and topics, and what a child wants out of a story. 10-16 is a really wide range; I'd recommend either shooting for middle-grade (~9-12 YO) OR proper YA (14/15+ YO). There is a lot of difference between a 10 year old and a 16 year old. Most 16 year olds are not going to want to read things that are 'for babies' or that is too simple for them.

Things to consider while writing:

- Why should the readers care about these AIs? I think using them as POV characters is an interesting idea, but you need to develop the AIs and give them goals and things they care about. These things can be alien, but there needs to be something there that makes readers identify with/sympathize with them. Do they want to play Monopoly, for example? Or those age groups are often wrestling with realizing that the adults in their lives aren't infallible - maybe one of the AIs realizes it's poorly programmed for its objective or even (in the case of the ML AI) comes to mutually incompatible conclusions and has to come to terms with that before it self-destructs.

- What are the stakes? Why do we as readers care about this Monopoly type game?

- How do the characters develop over time/what lessons do they learn? Fiction isn't non-fiction - just lecturing and stating your point is boring and sloppy. (I'm looking at you, recent Orson Scott Card). You develop your themes and make your points through the lens of character development. This is less straight-forward but fiction is powerful because doing it in this way allows the lesson to embed in both the thinking brain and the emotional brain.

"Zevin, who lives in Los Angeles with Canosa and their dogs, Frank, a pug mix, and Leia, a dachshund mix, is writing the screenplay for “Tomorrow, and Tomorrow, and Tomorrow,” a project that presents new narrative challenges."

Is this level of detail about her dogs necessary? Is this a journalistic thing?

I guess the dogs are part of the narrative challenges.

It's a coded way to say they don't have children.

I don’t know the meaning of the detail either, but just because you don’t get why some detail was included does not mean it was unnecessary. It’s possible, even probable, that the detail has meaning to a certain subset of the article’s audience and that the detail aides the story.

Money follows money. Truth holds the faltering hand of injustice. And so on...

This was one of my favorite books of the year. Reminded me very much of "The Amazing Adventures of Kavalier & Clay," which is my #1 of all time.

What is the criteria for “best seller”? Seems like a participation trophy by NYT standards.

Well if you read past the headline you'll see that the book has sold over a million copies worldwide, which is very good in the literary world. This is certainly not a participation trophy.

Queue the HBO series in 3, 2, 1...

>In 2021, Paramount Pictures and Temple Hill Entertainment purchased the film rights for Tomorrow, and Tomorrow, and Tomorrow for $2 million. The film will be produced by Marty Bowen, Wyck Godfrey, and Isaac Klausner,with Zevin writing the script.

It’s “cue” for this usage, I believe. “Queue” refers to a group of people in an ordered line, or the act of creating that ordered line. To cue something is to prompt it - as a cue card prompts an actor, for example.

Reply on news.ycombinator.com