If you can make one heap of all your winnings,

And risk it on one turn of pitch-and-toss,

And lose, and start again at your beginnings

And never breathe a word about your loss…

Yours is the Earth and everything that’s in it…

— Rudyard Kipling, “If” —

While maintaining the Bad Arguments website, managing translations and publishing them on the website, handling foreign rights, fulfilling orders from readers and bookstores outside the United States, and taking care of all the moving parts involved in the eventual handover to the publisher, I began thinking of a follow-on project. This was in January 2014, less than three weeks after the book was first released. My desire to keep moving stemmed, again, from the terrible fear of falling into comfort and, worse, back into that debilitating state that had held my faculties captive for the longest time. I would use about 70% of the revenue from Bad Arguments

The project would be an illustrated novella about a clumsy tailor called Hans who goes into a hotel and then, through a set of strange circumstances, finds himself in another world, perched on top of a gigantic pitcher plant. Things aren’t what they seem in this world. There he befriends a woman named Joanne, and together they attempt to make sense of their new surroundings. To escape dangerous predators and navigate tortuous paths in thick forests, he has to exercise his faculties like never before. Hans eventually realizes that the world he is in sits in another world, which in turn sits in another one. And the whole thing is run by a former computer scientist who may or may not still be alive and in control of it all.

The artwork would borrow many cues from Wes Anderson’s style — rich colors, intricate details, symmetry, characters looking straight at the camera, and so on. I had recently seen The Grand Budapest Hotel three times within the span of a week, and so his style was very much on my mind. I planned for ten pieces plus a cover. I would release the novella in three parts, each a year apart. A staggered release would allow me to test the audience’s appetite for the format before dedicating more time and resources to the project. It would serve as a sort of pilot.

The project’s primary goal was to infuse fundamental concepts from computer science into a narrative and not point those lessons out to readers, the idea being that they would absorb them and then perhaps recall them later in life. It took eleven months to put that first part together. I wrote most of it from a cafe run by two sisters on Folsom Street. The end result looked fantastic. The website was minimalistic, with black text on a white background, and the artwork was linked in the margins. I had a wonderful quote from the executive director of the Mozilla Foundation — Mozilla’s education, teaching, and advocacy arm — that I added to the top of the website. Unlike Bad Arguments’ format, this one worked much better on mobile. I pitched the project to the publisher that had acquired Bad Arguments and they got back to me with an offer, which I said I would think about. Everything was looking great. The project would follow the same model as before, with a free version available to readers online and a print edition available for purchase. By now, it was December 2014.

Then I lost confidence in the project. A feeling washed over me that suggested that the format might not work as well as I had hoped. I was asking too much of readers. Subtext works great in fiction, but it’s not particularly useful for teaching. During that period I realized I had once felt that same disconnect — that same lack of tessellation, you might say — in a completely different field. I did freelance photojournalism for a while, and quickly learned that perhaps nine tenths of the success of a shot is due to the rapport that a photographer builds with a subject in the few seconds needed to take a photo. The transparency, the eye contact, the slight nod of the head on the photographer’s part, the smile, the shot, showing complete and utter respect for the subject’s world that I was encroaching on. Sticking a camera in someone’s face or trying to be overly discreet are bad moves. If I sense any breach in that seconds-long path towards a rapport, I back away. And that’s what I was feeling with Hans.

I released the website anyway, since I had spent about $15,000 on the project. It came out on March 15, 2015. On that day, I also revealed the names of the two forthcoming installments — Part II: Hans in the Belly of the Nihilistic Tuna Fish, and Part III: Hans in the Mantle of the Socially Awkward Deep Sea Octopus. That first installment ended up getting around 25,000 visitors over its useful lifetime, which confirmed my suspicions that it wasn’t something that would likely stick. But the project, despite having been cut short, paved the way for another idea not too long afterwards. What if I formalized a story space, I thought, and broke it up into scenes, analogies, artwork, and prose, then allowed readers to submit one or more of those elements and weave together their own stories? They might end up with a story that begins like Coates and ends like P. G. Wodehouse. How great would that be? I had read about a startup during that time that had just raised nine million dollars from Google Ventures. They sold personalized picture books, and they were having success. It seemed like a great way to make the format engaging. Three years later, I came across another startup at the Bay Area Book Festival, which was doing something similar.

The project would be grounded in three pedagogical tenets: one, analogies are a good way to learn new concepts; two, learning something using a variety of methods can help one better understand that thing; and three, stories are the oldest form of learning. A story would be an ordered set of scenes that defines an arc. A scene would be a set of three elements, each of which engages the reader’s attention in a different way. Given a story made up of three scenes, if readers contributed just six additional scenes to the story space, one could in principle create twenty-seven unique stories.

During that time, I met up with a friend who was in town for a few days, and she mentioned in passing that she knew the director of research at Google. Let’s call him Sam. Naturally, I sent him an email the next day. Sam wrote back a few hours later and suggested a lunch meeting the day after. I’ve found that a good rule of thumb is that people who plan to reply to you will likely reply within two days. We met over lunch that Wednesday. I was so nervous on the train ride to Mountain View that I experienced what I later learned was vertigo, where you feel like everything around you is spinning. That sensation eventually gives way to nausea. Right after I got off the train, I spent twenty minutes sitting on the chairs outside the Greek bakery on Castro Street. I was still feeling lightheaded by the end of it, but I couldn’t afford to sit there for much longer, so on that warm summer morning, I walked for forty-eight minutes from the Mountain View train station to Google’s campus.

Sam was gracious. He showed up in one of his iconic Hawaiian shirts and took me to lunch. The main takeaway for me from that meeting was Sam’s reaction to my explanation of how Hans was like Sophie’s World and The Phantom Tollbooth and Alice — Sure, those were all successful, he said, but how many just like them weren’t? Put differently, they were successful despite their educational elements. That short meeting with Sam would start a series of thoughts that accelerated a few months later during my — by now annual — solo drive across the country. It culminated with an idea that would combine the narrative aspect of Hans with the interactive aspect of the formalized story space concept and the exposition style of Bad Arguments.

I worked on that idea during each of the five nights I spent in hotels on my drive back from Florida to California, and by the time I got home, I had something that felt compelling. It was a set of fifteen scenes, sprinkled with malaprops. In each one, a character attempted to solve a problem in one of several ways. The reader would therefore see how various ways of solving the same problem fared relative to each other, and in doing so, learn not only about concrete fundamental algorithms and data structures, but also a general way of thinking about problems. During the meeting with Sam, the term algorithmic thinking had come up, which had reminded me of the person who likely coined it back in 2005 or thereabouts — a professor whose class I had once taken on formal methods. And so I christened the project Comparisons: Algorithmic Thinking in Everyday Life.

I wanted to try my hand at submitting to the big publishers and seeing what that experience would be like. I went to Alexander Book Company on Second Street and started looking through books that I figured had a similar audience to mine. I checked to see who the agents were for those books, since authors typically thank their agents in the acknowledgements. The first book that came to mind was What If? by Randall Munroe. I created a list of agents and headed home.

I spent a further two weeks finishing up a proposal for the idea that had come together on the road. I sent a query to the first person on my list. He was a literary agent born in Texas, now based in New York, and as luck would have it, he had gone to the same university as me, albeit to its school of literature. I made sure to mention that in my query. Given that it was a week before Christmas, I expected to have a few more weeks to go over the proposal — to flesh it out and clean it up. An auto-reply came back saying the agency was closed for the holidays and that I should hear back in around six weeks. Perfect.

The next day I got another email, this time from an associate saying that the agent I had contacted was interested. That was followed up by an email from the agent. “Can we talk on the phone?” I took the call and the agent mentioned that he had really liked the idea for the book. He went over what a literary agent’s role is — to have the author’s back during every stage of the publishing process. It would begin with spending time on the proposal to ensure it was as strong as it could be. Then deciding when to send it out, as there are book fairs all throughout the year, and it sometimes makes sense to tap the market during a particular season. Then managing the sale to a publisher. Then managing the agreement with that publisher. Boilerplate agreements are typically most favorable to a publisher, and an agent can work with the publisher to make sure it’s equally favorable to an author. The agent is then available throughout the year or so it may take to go from the manuscript to the finished book. If the publisher does something that’s not reasonable, the agent is there to sort it out. He or she may also help get endorsements, suggest shows the author can go on, make introductions, and so on. In return, the author gives up 15% of all royalties associated with the project. For the value a good agent is able to bring to a project, that 15% is well worth it.

He asked if I needed time to think things over, or to look at other agents. I said no. And so our relationship began. I signed a representation agreement, and we then moved onto polishing the proposal. It was early January 2016, and the plan was to finish the proposal before the London Book Fair in mid-March. The proposal would have a chapter breakdown, a section on why the book ought to do well, which would share various data points from Bad Arguments — visitors, sales in print, foreign rights, mailing list size, endorsements, and so on. It also had the first five chapters of the proposed project in full — prose and artwork.

It was going to be tough finishing five chapters in six weeks, but it was doable. Alejandro was on board with the tighter schedule too. I took two weeks off from work, which was nearly all the vacation time I had left, and within five weeks, the proposal was ready. I sent it to the agency. They made it better. The agent shared a list of a dozen imprints he thought would be a good fit for the project. He sent emails to all of them. The book was now out in the world. Seven imprints replied. We picked a day and the plan was for me to do back-to-back phone calls with all of them. I took that day off from work, and now had negative vacation days left.

The first call was at 8 a.m. Pacific Time with an imprint of Penguin Random House. We talked about algorithmic thinking as a transferable, timeless approach to teaching computer science. I mentioned that when people talk about coding, what they really mean is three things: programming languages, data structures, and algorithms. This book would focus on the latter two. They were interested in the life-hack approach to computing and wanted to frame the book that way. They liked that the book didn’t pander to readers. The imprint had a large variety of titles, everything from literary fiction to non-fiction to popular science to illustrated books.

The next call was with the trade division of a publisher based in Boston. They aren’t one of the Big Five, but are well-known for their educational books and textbooks. They published What If? and Thing Explainer, by the same author, as well as Umberto Eco, whom I’d heard of, but never read. I talked about the book’s value proposition of teaching computer science, but framing that teaching as a general-purpose thinking tool, a term I had borrowed from Daniel Dennett. They wanted to reframe the project by focusing more on computing and less on algorithms, which would have required a good amount of effort, and I was already out of vacation days. So that was a problem. I liked that they published only a few titles every year, and didn’t focus solely on bestsellers. They tried hard to make every book work.

The next call was with another imprint of Penguin Random House. I made the case for algorithms being like math in that they are both ubiquitous and hence worthy of attention. The ideal reader for the book I was proposing would be a non-specialist interested in computer science, or perhaps a parent. We talked about shortest path algorithms in global positioning satellite systems, recommendation engines on websites like Amazon and Netflix, and file compression algorithms. We also talked about the tradeoffs of explaining algorithms by analogy.

I chatted with an imprint of Hachette next. The highlight of that call was our conversation about Bad Arguments and how the community aspect of that project had evolved. My answer that much of it was due to sheer luck seemed to underwhelm the editor. I learned that the imprint has the highest number of bestsellers to published books. I had mentioned in the proposal that one of my aspirations was strong editorial direction, and that, along with strong marketing were two things they prided themselves on. They also mentioned that there were lots of things they felt they could do with the book online, which was wonderful, since that was my general thought for marketing the book.

Next was yet another imprint of Penguin Random House. I went into the call slightly partial towards them since they publish my favorite author, by far — J. M. Coetzee. They also happened to have published Stanislas Dehaene’s book Number Sense, which was one of the inspirations for Comparisons. I learned that they have a strong frontlist of non-fiction books and an equally strong backlist.

Next was an imprint of HarperCollins. They were passionate about narrative science, humor, and “big idea” projects. They had a relatively small list of books, which allowed them to give a lot of attention to each book. No book would get lost in the shuffle. The final call was with yet another imprint of Hachette whose list of books includes Malcolm Gladwell’s. This was a fantastic call. The editor was technical, and he and I shared an alma mater. He liked the book’s emphasis on there being more than one way to do something and its primary focus on efficiency. He mentioned that publishers are really good at presenting work to a general audience, rather than a niche one, which in non-fiction would mean taking an idea and showing how it is broadly applicable. Much like me, the editor had begun his career writing software and eventually came to publishing because he was fascinated by how transformative books could be, and he wanted to be a part of that.

All in all, calls like these are an excellent way for an author to get to know a publisher. Some agents like to be on these calls, but my agent preferred not to. And I have to say, in hindsight, it was the right decision. The conversations were much more intimate and less formal that way. One interesting thing to come out of this experience was my realization that there were too many imprints out there to keep track of, and that it wasn’t always easy to figure out which imprint belonged to which publisher. So I spent a few days putting together a visualization of the so-called Big Five trade book publishers in the United States and their imprints. It was for my own benefit, primarily, but it proved popular with a lot of people in the industry. Last I checked, the piece had been liked or shared 5,400 times on Facebook and visited by 52,000 visitors. Nine executive editors, editors, and readers have since contributed to it.

When more than one publisher is interested in a proposal, the book goes to auction, and the publishers have the opportunity to bid on it. If there’s more than one bid in the first round, the agent arranges a second round. Four publishers entered bids in our first round. Of those, one had a hard limit and wasn’t able to go any higher, so three publishers went into the second round. The bids were all competitive. I was walking by the San Francisco Public Library on Market Street, heading to Van Ness Avenue to get lunch when I got the call. Two of the publishers had nearly doubled their original bids.

My agent walked me through the three offers and reminded me of the publishing houses and the acquisition editors and what each of them would be able to bring to the project. We chose one. He said that particular publisher had asked for world rights, but he only wanted to sell them North American rights. He didn’t expect it to be a problem, but he would have to get on the phone and work it out with them. The audacity of the man. I wished I were a literary agent at that point. What an amazing gig it must be, to be in the business of making people happy. I skipped lunch that day.

The next morning, as I was pushing my daughter’s stroller to school, I got a one-line email that read “CALL ME.” I assumed the worst. The deal was off. They must have insisted on world rights. What a shame. I dropped my daughter off, stepped out into the courtyard in front of the school, and made the call. He told me the good news. The publisher had accepted our terms. The project was now unofficially signed by a major publisher. “Don’t go out buying a yacht just yet.” He explained that payments happen in installments. An author gets, say, a fourth on signing, a fourth after the manuscript is handed in, a fourth after the book is published, and a fourth a year after that.

I was indebted to my agent, and I still am. He had brought two years of experimentation to a beautiful place in a matter of months. There’s a daily newsletter called Publishers Lunch that I’m subscribed to. They share daily acquisition updates and have a scale of euphemistic categories indicating how lucrative a deal is. According to them, a nice deal is a book sold for between $1 and $49,000, a very nice deal is $50,000 to $99,000, a good deal is $100,000 to $250,000 and a significant deal is anything more than that. We sold the North American rights to one of the Big Five publishers. It was a great deal. Though we hadn’t gone with one of the imprints whose editor had made a strong and personal appeal for the project, I emailed that editor. I was grateful for his time and looked forward to our paths crossing again. I needed him to know that.

The first thing the publisher requested was to change the book’s title from Comparisons to Bad Choices. I had chosen the original title because it didn’t overpromise anything nor attempt to suggest any causal links. It was a boring, matter-of-fact title that described the book’s approach in a word. I’m partial to titles like that, generally. Bad Choices wasn’t too bad, I supposed, as it did remind me of Donald Knuth’s descriptors for algorithms — good and bad to denote efficient and inefficient ones. It was somewhat ironic that I had spent two years on this project in order to escape the clasp of the previous project, and now the two were prominently joined at the hip. After a week or so, I received four books in the mail from my editor. They were titles she felt I might enjoy reading, by an author whom I’d mentioned in our first call. She included a card on which she’d written a note. I’m notoriously unsentimental, but I framed that card.

Next, my agent recommended meeting with the publisher in person. I agreed, but there was just one problem. I don’t like flying. I hadn’t boarded a plane in years, with one exception, and I was medicated during that flight. But I didn’t tell him that. We agreed on a week in December, which coincided with Comic-Con, and I drove for six days from California to New York. I loved every minute of that drive. There’s something about driving solo, about being slowly consumed by the open road that is infinitely calming. It reminded me, for some reason, of that contraption from Kafka’s “In the Penal Colony,” with its heavy gears turning and clicking and banging over the course of many hours so that the person strapped to it may experience a spiritual awakening unlike any other. Only to then die. Not the best analogy, I suppose. I drove along one stretch in Nevada in pitch darkness to the soothing sounds of one of Dan Carlin’s Hardcore History podcasts. The trance was only broken by the neon sign of a Burger King where I stopped for a much-needed bite. The drive through Wyoming and Nebraska was absolutely breathtaking.

I walked into the publisher’s building at 375 Hudson Street on the day of the meeting, and I suppose irrespective of who you are or how far you’ve come in life, that kind of an experience is thrilling. Reading about an institution your whole life, seeing their logo on a good number of books on your shelf, and then finally walking through its doors. It was magical. I was grateful to have landed such an opportunity. At moments like that, you feel all your very best work is still ahead of you. We had lunch at a nice restaurant, then I sat down with the team — marketing, publicity, editor, executive editor, agent, and me — and talked about the project. To clarify two terms that might occasionally get confused, marketing focuses on promoting the book through channels like ads or reviews, whereas publicity ensures the book’s readers have an opportunity to hear from the author by means of interviews, radio shows, tours and so on. We brainstormed ideas for promoting the book, and fleshed out two concepts.

The first idea was a story about friendship, good intentions, and clumsiness. I had taken a printout of one of the characters in the book with me on the drive — Feynman, with a plate on his head; Plate Guy for short — and had taken photos of him at various stops: a field in Fernley, Nevada, a bookstore in Fort Wayne, Indiana, another bookstore in Salt Lake City called The King’s English that I really liked, a mom-and-pop cafe and diner in Cheyenne, Wyoming, and dozens of other locations. I had mocked up a storyboard, and it struck me as a fun little picture essay that readers might enjoy.

You pass through a lot of small towns on a cross-country drive, and I remember this one little town where I’d spotted a Dairy Queen that looked to be right out of the 1960s. I parked the car and was about to get out to take a picture of Plate Guy inside, doing something silly like ordering a burger or eating an Oreo Blizzard, when I noticed that everyone inside was staring at me through the pane. A middle-aged man was holding a burger with both hands, elbows on the table, a gaping mouth ready to receive the burger, and yet he was frozen in place, transfixed. It was like a scene out of a Stephen King novel. I got back in the car and sped off.

The second idea was a ninety-second clip inspired by the Minotaur passage in the book’s chapter on mazes. Three beavers have to escape a maze and avoid meeting an untimely demise. I created a clip with a proposed storyboard and set it to the song “Run Boy Run” by Woodkid. I reached out to a studio in Barcelona whose work I’d come across online. Their style seemed perfect — quirky and detail-oriented. They were fans of Bad Arguments and gave me a quote, which, sadly, was an order of magnitude higher than what I had budgeted. So that idea never materialized. I’m sure readers would have loved it.

We ended up releasing a series of weekly 600-word pieces in the five weeks leading up to publication day, and associating each piece with an animated version of a character from the book. That way, we could share them on platforms like Instagram and hopefully get some momentum going in anticipation of the book coming out.

The animations were subtle: eyes opening and closing, or looking from side to side. Photoshop actually has a Timeline window that allows you to create simple frame-based animations, and that’s what I used. Each piece would take a concept from the book and talk about it using topical examples. My favorite was the one on curation and bias. It reacted to several news stories at the time about how terrible algorithms are because they cloud our judgment and force us into filter bubbles. In isolation, this can be a valid point, I suppose. But when contrasted with, say, newspapers, which were being held up as a platform that was somehow less immune to such bias, it was a preposterous proposition.

By June 2016, the manuscript was ready. I sent it in. The book had been scheduled for an April 2017 release, so we had plenty of time. After about a month, my editor mailed me a marked-up copy of the manuscript with notes written all throughout the margins. It’s a fun experience, when you think you have something that’s hopefully compelling and here is this other person whose sole responsibility is to make it better. By August, the manuscript was ready. I shared the manuscript and illustrations with three of my coworkers — all software engineers or computer scientists — who had offered to go through it and provide feedback. This was something I hadn’t done with Bad Arguments. Within two weeks, I had heard back from all three. Their suggestions were fantastic, and I incorporated all of them into the manuscript. Once the manuscript was frozen, so to speak, it was handed off to a copyeditor. I was surprised to find that the copyeditors were all freelancers. I had always imagined the big houses having dedicated copyeditors who all wore thick glasses and sat in lavish offices full of sharpened pencils. I asked Benjamin Dreyer, Random House’s copy chief, about this once and he replied that when he’d joined back in the nineties, there were indeed three in-house copyeditors. When they retired, the publisher determined that it made economic and organizational sense to outsource that responsibility to freelancers.

Once the manuscript had been copyedited, things went quiet for a long time. That’s how traditional publishing works, it seems. Action comes in bursts. This period was difficult to bear because I had no control over the process and I had limited insight into how things were progressing. Had we fixed those last typos for sure before sending out the review copies? Were reviewers getting back to our requests for blurbs, and if not, how could we fix that? Was the marketing plan in place, and if not, how could we fix that? When you’re the one navigating, you can react to exogenous factors that impede your planned course and adapt midway. But when you’re one or more degrees removed from that process, you become a true believer. It’s a tradeoff that can work for an author who delegates with ease, though it may prove challenging for an author who is obsessive about a project’s every last detail or one who prefers to know everything about how a project is going, including bad news. It was around this time too that I heard from the agent that the publishing contract was finally ready for review. It had taken several months for the agency to painstakingly go through it line by line and request various edits.

Next came the sales to international publishers who were interested in the book. There was a good deal of transparency on the part of the agency about how things were going, which publishing houses were interested or had made offers, and which of the offers they recommended. In some cases the domestic agency works with a foreign agency to close deals. And in those cases the foreign agency takes another 10% cut, so total commissions can amount to 25%. The agency sold the UK rights to a 250-year-old publisher of which I was a huge fan. Its editor was completely in sync with the project. In the weeks and months that followed, the agency sold the rights to the Simplified Chinese, Complex Chinese, Korean, Japanese, Russian, and Turkish translations. The Japanese agreement was the most memorable one as it was only two pages long.

Foreign rights are never a given. Sometimes they happen, sometimes they don’t. Sometimes they come in one after the other at the beginning of a project; other times they can come in slowly over a longer period of time. Sometimes the agreements are structured on a royalty basis, other times it’s a lump sum for a single print run. With this project, we were lucky that six international publishers were excited about the concept. Two have been published to date — the Simplified Chinese and the Korean editions — with the rest planned for release soon.