clock menu more-arrow no yes
Two stick figures, male and female, watch a third attempt to change a lightbulb. The third is hovering, with one foot each on a helicopter drone. Randall Munroe/Riverhead Books

Filed under:

‘I’m always thinking of really terrible ways to do things’: a talk with XKCD’s creator

How to make an optimal cool fact delivery system

Say the name Randall Munroe to your average internet literate, and they might not recognize it. But they’ll certainly recognize the hundreds upon hundreds of stick figures he’s drawn over the 14 years he’s been producing the often funny, sometimes heartfelt, frequently experimental webcomic xkcd.

XKCD has formed, if not the backbone, then at least some of the ribs of internet culture. It’s inspired the chessboard roller coaster picture and a method for creating easy-to-remember secure passwords, and Munroe won a Hugo award for a comic in the form of 3,099 individual images uploaded sequentially over the course of four months.

Munroe’s books are reflective of the same inventiveness. Thing Explainer offered explanations of complicated scientific ideas using only the 1,000 most frequently used words in the English language. What If? gave realistic answers to absurd questions, like “what would happen if one tried to funnel Niagara Falls through a straw?” His latest, How To, is the exact opposite: absurdly bad advice for mundane problems, like how many pounds of rocket thrust it would take to literally move your house, instead of, you know, moving your belongings to another house.

Polygon sat down with Munroe to talk about stick figures, science, and the internet — and wound up touching on ’90s kids, astronaut Chris Hadfield, and the Marvel Cinematic Universe.


Polygon: Where did the idea for How To come from? It has some similarities with your previous books, but it also has its own very specific hook.

Randall Munroe: Well, I’m always thinking of really terrible ways to do things. It’s not like I try to think of terrible ways to do things. It’s just that when I’m trying to figure out how to do something, I just think of a bunch of different ways and see which ones seem like they might be good ideas. But what’s really fun is that sometimes the ones that are pretty clearly bad ideas, it’s still a little bit interesting trying to figure out why exactly they’re bad. Sometimes it’s really entertaining just to imagine trying to carry out an idea.

A lot of the time I’ll learn something really cool in the process of figuring out whether an idea would work or not. So even if it doesn’t work for the thing I’m actually trying to use it for, I might learn to use a tool or learn about some cool stuff that will help me with an idea later on. And all that aside, even if it’s not useful for anything and the idea does turn out to be bad, it can just be fun.

Did you have a favorite how-to method that you discovered or investigated for the book?

I think that my very favorite thing that I discovered was when I was trying to do the chapter on how to make an emergency landing, which is one of the cases where I got to reach out to some really cool people and get them to give me their advice on how to do things. I got to reach out to Chris Hadfield, former commander of the International Space Station and an accomplished test pilot.

A stick figure comic from How To?. Flight officials ask the passengers on a plane if any of them know how to land an aircraft. One shrugs, saying “I guess I’ll never know until I try.” Randall Munroe/Riverhead Books

I spent a long time trying to come up with a list of the weirdest things that could happen to you while you were flying, where you would need to land. Like, What if you get your sleeve closed in the cockpit door and you have to throw stuff at the controls? Or, What if you’re on the outside of the plane? Can you land by crawling around and pulling on the right flaps?

I figured I was just going to ask him about these progressively weirder situations until finally he just hung up on me. But the joke was on me, because he just answered every question I asked with no hesitation. In some cases it was, No, there’s nothing you could do there;you would crash. But in a lot of cases, he was like, Oh sure, I just do this, pull this flap, climb out of the seat, do this, and then you can bring it down to land. People have done that before, there was an incident, and so-and-so ...

And he laid out everything in such a matter-of-fact way that — I was gonna write that as a regular chapter where I gave my thoughts on all this, but I wasn’t expecting him to not only have thoughts on it but have them all immediately. His answers were so punchy that I just printed that as an interview and illustrated it. In retrospect, my idea of trying to stump a test pilot by throwing surprise situations at him all of a sudden might’ve had some flaws.

A stick figure portrait of astronaut Col. Chris Hadfield from the book How To. He is holding a space suit helmet with the Canadian flag on it. In the background is a rocket ready to launch, and two clouds. Randall Munroe/Riverhead Books

Hm, yeah, that might be the entire job.

Exactly. And it might be that there are some situations in there where what he said wasn’t the optimal thing that you could do. But what he says is, This is the thing that would be the first instinct of one of the world’s foremost test pilots. So, it’s probably a pretty good place to start.

When I said that there are some similarities between How To and your other books, I meant that there’s a streak of explaining complicated things in layman’s terms to a broad audience. Do you see that as a theme in your work, that you seek to put into the world?

Yeah. I don’t know if it’s necessarily that I see it as a calling or a mission that I want to do. But it feels more like: When I was a little kid and I learned cool facts, I would try to tell people about them. And I noticed pretty quickly that a lot of time, people looked bored or would walk away, and I would have to chase after them telling them my cool facts. And so I’ve tried to, like — I’ll learn about something that I think is cool and I want to show other people why I thought it was cool.

A lot of the time, that means figuring out how to fill them in on all the context for it. Like, it might be a cool programming idea, but people who don’t know that programming language don’t know why it’s a neat thing. A lot of the time I’ll just find a really neat research paper, and then I want to summarize it for someone and to tell them why I thought it was so funny or so interesting. And then to do that, I have to explain what the research is about, so I just try to figure out ways to do that quickly, where people won’t get bored and wander away.

An optimal cool fact delivery system.

Yeah, exactly. And that’s good because I think it’s really good in general when you’re trying to communicate about science to understand that it’s not that people aren’t interested in science, or that they aren’t smart, or anything — it’s just, people have their own lives. They have stuff going on. Your fact really has to be cool and you have to figure out a way to get it to them quickly, or they are rightfully going to turn to one of the other 10 million things trying to get their attention. And that’s fair; you aren’t entitled to people’s attention. You have to make it interesting for them.

Do you come from a scientific background?

I did an undergraduate physics degree and then I worked for a little bit on a couple of different things at NASA [Langley] Research Center; I built robots there and stuff. But then I pretty quickly went into internet cartoons full time. People always mention it in my biography, but it was — I was not there for all that long. It wasn’t like I had a long career as a robot assistant and then finally found a second career.

A chart depicting the alternate orders Asimov could have put the three laws of robotics in, and the dire, or at least frustrating, situations that would result.
“The Three Laws of Robotics,” the 1,613th xkcd strip.
Randall Munroe

I relate to the idea of trying to very efficiently tell people about a cool thing. A lot of the focus of my job is distilling Marvel and DC Comics universes down to people who are not really well-versed in them. I am, naturally, of the opinion that there are a lot of cool things in there.

For sure! Quick question that I’ve always been curious about and don’t know the answer to: Of people who have seen at least one Marvel Cinematic Universe movie, what’s the median [total] number of [Marvel Cinematic] movies they’ve seen? Have most people who have seen one of them seen a lot of them? Or are there a lot of people who have just seen one of them but are not that into it?

Because every event or thing connected to the movie — every review or every comment thread — it’s disproportionally people who are really into it. But I wonder if it’s one of those things where those people are just a tiny slice of it, and the vast majority of people have seen fewer than four of them but still like them, you know.

Yeah, that’s a question where the answer is absolutely beyond the horizon of what anybody who is a journalist who covers that can see, which makes it a very interesting one.

I’ve always been curious. I don’t know an easy way to find it out. But I wonder about that sometimes whenever people talk about the Marvel movies. Because I’ve seen like five or six of the MCU movies, and I feel like every review or every comment I’ve ever read about them either presumes you haven’t seen any of them, or presumes you’ve seen all of them. And so I wonder if I’m a huge outlier.

A stick figure comic pointing out that it’s much easier to think of a movie you hate that everyone else likes, than it is to think of a movie you like that everyone else hates, from the webcomic xkcd.
“Unpopular Opinions,” the 2,184th xkcd comic.
Randall Munroe

But I don’t mean to take away from...

No, I think the completionist instinct runs very strong in nerd media. But! There’s a lot of comics philosophy out there about how the job of cartooning is about minimalism, which is kind of a fancy way of saying, “OK, you draw a stick figure comic.” And maybe that might start as I’m not Alex Ross, I’m not a monthly comic book artist; this is just the way I’m communicating my ideas. But at some point it becomes a very deliberate choice. I’m curious if you have ever felt like you were restricted by that form or wanted to break out of it?

I feel like there are a few places where I did where I might’ve felt like that and then did kind of expand out of it. One of them was when I discovered I guess what we call infographics, which I didn’t invent, but I didn’t really know. I was just like, I really like making these elaborate maps of things — but not always maps, just charts.

I would do these big data visualization-type projects, and I discovered that I could just do those and put them up on my comic in place of my regular comic, and people seem to like them. That was really cool for me, and that became one of my favorite parts of it.

Movie scientists do interesting-looking experiments and come to a swift and specific conclusion, while regular scientists perform a relatively mundane test and come to an obscure result. All the stick figure scientists are wearing lab coats and safety gog
“Science Montage,” the 683rd xkcd comic.
Randall Munroe

Similarly in the big chart genre, doing Thing Explainer was fun because it actually felt like going back to when I was a little kid and trying to come up with inventions. I would draw these elaborate diagrams of, like, my castle fortress with its booby traps, or the hover vehicle I wanted to build, or whatever. With Thing Explainer, I got to draw up blueprints of a bunch of stuff and go back to some of my — like, I took some technical drawing classes — but also just to those childhood drawings.

One stupid joke that I’ve always been a sucker for — and I now and then think of a of a new joke you could do with it, but I can’t do because of the constraints of my medium — is jokes involving people who tear away a suit and have a different suit under it, a different outfit. That’s just one of those silly sight gags that I’ve always been a sucker for, and I don’t know why. Now and then I’ll think of a new way that you could use a series of tearaway suits, and I can’t do those jokes because I draw stick figures and you can at most put one set of clothes on them. And even then it’s a little bit weird.

I think the initial instinct for folks who maybe have not gotten into comics hard enough to get the craft of the rhythm and panel placement — all of these things go into a comic; it’s not necessarily that you can draw the perfect eye, or whatever. What do you feel like you have like learned or improved about your craft? Is there, are there things that you’re like, I am so much better at drawing stick figures doing this than I used to be?

Honestly, all the examples I think of are things that are exactly the opposite, where I’ve been drawing them forever and I still can’t get it right. I’m sure there are things that have gotten easier and those things I don’t remember them because, you know, [remembering that stuff is hard]. But like, I keep having to do comics about airplanes or illustrations about them, and there’s something about the geometry of them that I’ve just never been able to draw.

So anytime I include an airplane, and especially one of those airliners with the wings that are sort of swept back — every time I draw one of those, it meant that I tried it 20 times and that one was, like, the best I could do. But yeah, there are a few things that I learned early on about how you have to put everything explaining a joke or explaining a comic into the comic itself.

When I started out, I thought, Oh, I’ll put up these news posts, like I see a lot of people do. And then sometimes I’ll do a comic and I can put out a thing there explaining, This is why this joke worked or This is based on this cool study you should read about. But I would find that that would really quickly turn into, like, Oh, you know, this comic doesn’t really make sense, but I can explain it in the news post so it’ll make more sense to people.

I had to learn and remember that nobody — to a close approximation — nobody reads anything except for the comic. So you have to put it there; with 99% of your readers, that’s the only shot you get at getting a thing across. I had to learn to give up on my urge to explain things to people, especially if I’m like, Oh, people didn’t understand this. Well, I’ll reach out to them and change all their minds individually. It’s like, Nope, you had your chance; you got to just do it better next time.

A stick figure man clings to the top of a passenger jet, saying “Hello?!”, in an illustration from the book How To. Randall Munroe/Riverhead Books

That harks back to such a specific era of webcomics, where there would be an accompanying news post.

Yeah, and I saw people doing [that] and was like, Oh, maybe that’s just how this works? But I resisted that, so I never really included anything like that with my comics after the very initial ones that I posted.

Whenever I talk to comics creators and let slip the age I was when I first discovered their work, I always feel the need to apologize, but ...

Oh, for sure, but I’m a huge sucker for those. I’ve done a bunch of comics about how weird it is to pick out things that age you in unexpected ways or in unexpected places. So it’s always fair when people want to spring those on me because I’ve sprung so many, like, Hey, you want to feel old? Here’s a thing.

I even have a chapter in my book that’s “How to Tell if You’re a ’90s Kid.” It’s all about like — there’s a really interesting section on analyzing what names you think are normal and how that really dates you. And how you can see fictional characters have names that are common among little kids, and then 20 years later, there are a bunch of adults with those same names.

And the reason for that is that TV writers and new parents are both doing their work at the same cultural moment and picking names. But the TV writers have a head start because they can give names to adults. You can see those cultural forces playing out. Like, for the last 10 years or so, there have been more babies named Brooklyn than Sarah, which blows my mind because I knew, like, 20 Sarahs.

And then I also talk about how nuclear testing leaves radioactive particles in your teeth. That tells you if you’re a Baby Boomer or not! So it was a fun chapter. But yeah, you were saying, when you talk to creators ...

A stick figure drawing of a tooth fairy approaching a person at a desk, from the book How To. The person refuses the fairy’s teeth, because they are from Baby Boomers. Randall Munroe/Riverhead Books

Right. What I was about to say was, I think I discovered xkcd when I was in college, so I’ve been reading xkcd a long time. And through being a professionally online person starting in about 2010, going through the Google Reader era and the Reddit-is-still something-that-normal-people-don’t-know-what-it-is era, and the Digg era — xkcd stood out for me for having a very optimistic idea of online culture and online spaces and online communities.

And in the early 2000s, it was not radical to feel that way about the internet. But this is all context for asking, do you have a take on the evolution of online communities? As someone who’s observed them and been around for so long? It feels like xkcd is broadly optimistic about the potential of the internet, but how has that played out for you personally?

It’s funny, because I’m definitely not particularly out there myself on a lot of these communities. I’m watching them from the sides. I think there are certainly some things that I was optimistic about that then turned into ... more like forces that were ambiguous at best. And then other things were things that I was really upset and worried about, or mad about, that now look like the only good ones. Like, Oh wow, if only we could have kept having it that good.

I mean, I think back to the ’90s: Bill Gates was the ultimate evil for crushing things with his software for business, and we all wrote Microsoft with a dollar sign and we were rebels and stuff. And now it feels like he’s sort of the only, like, tech billionaire who’s just trying to do good stuff? Like, he’s out there curing malaria or whatever, and then you look at what everyone else who has a bazillion dollars is doing and you’re like, Whoa, wait a minute. Maybe we were too hard on that guy. [laughs] Like, Wow, now I wish we could get back to when Bill Gates was the worst thing on the internet.

I think that what I’ve always been really optimistic about is people themselves. I really feel like everyone, I don’t know ... like, one thing I just really, really dislike about science and physics sometimes is when there’s a little bit of that condescending streak, and that idea that like, Oh, well we’re smarter than other people.

That just really rubs me the wrong way. When people are like, People are stupid. And it’s like, people are exactly average; people are people. And if you’re saying that what you’re really trying to say is, like, “I’m smart. I’m not like everyone,” [then], like, well, OK, so you’re full of yourself.

I’m optimistic about people in general basically being good. There’s some survey question that they ask like, Do you think most people can be trusted? And it in the mid-20th century, like 60, 70% of people said, Yeah, for sure. Most people can be trusted. And now it’s like 30% say that, and 60 or 70% say like, I dunno, you can’t be too careful.

And I’m as wary as the next person about any kind of random stranger on the internet, because you don’t know where they’re coming from. But fundamentally, I really believe most people can be trusted, most people are good. You just have to build a system that has the right tools to make it safe to interact with the people who aren’t, and tools to protect communities from being swayed by a few people who are really bad.

That’s the thing that I think I criticize the most often about how the internet has developed, which is that it lets small communities — people who are dedicated — manipulate larger communities to give the impression of a peer consensus around some topic. Or if someone’s targeted by a handful of people, it doesn’t take very many people to completely dominate every public forum, or comment thread about something on the internet, and give the sense that like, Oh wow, if I support this person, I’m in a small minority, because everyone seems to be mad at them.

And this is something where even if you’re relatively savvy ... like, for me it’s really still really hard not to have my impression swayed by what I see in the first couple of comments. I really do think we need our norms to catch up to where we can respond to this kind of stuff better, and we can build systems that are more robust in these ways. And I feel anxious and urgent about it.

It’s like in the era of email forwards, where there would be like a chain letter that’s like, If you don’t do this, you’re going to have terrible luck, or Bill Gates is going to send a dollar to everyone who forwarded this email. It was everywhere; it flooded email. It isn’t that we built something that stopped those emails. We didn’t have to rearchitect things, but we did need everyone’s email behavior to catch up.

I think we need a combination of both of those, and I feel frustrated that it doesn’t feel like it’s happening as quickly as it should, but I think it’s happening. So I dunno. I think that the more we talk about the kind of, for example, harassment people get on the internet — if they are [laughs, and then continues knowingly] are a certain kind of person, or express certain views, the way that it can leak into every part of the internet and completely destroy someone’s ability to exist in the digital space — [the better].

[About] how easy it is to launch those kinds of campaigns, and how easy it is for the rest of us to be swayed by them, not really realizing that what we’re seeing is not representative. I guess that comes back to — it’s like that Marvel Cinematic Universe thing.

A chart ranking tools that Thor could have been associated with from best to worst, including hammer, circular saw, nail gun, digital caliper, and others.
“Thor Tools,” the 2,097th xkcd comic.
Randall Munroe

We only ever hear from the superfans, but that means that it’s pretty easy for a group of people to decide to be superfans and really sway things for everyone else. And you know, sometimes that’s how activism gets done and that’s good. But I think that we really haven’t come to grips with exactly what that means for digital spaces. That’s the most pressing and scary thing for me. We need to figure that out. And whether it’s a technical thing or a norms thing, or probably both, I really wish that would hurry up. Before, you know, democracy collapses, or we all destroy each other over some dispute over a dress or whatever the heck.

I was just sitting here thinking that optimism isn’t the right word, but futurism is. Looking forward and trying to figure out what direction we should go in before we are already going in a direction is a better characterization of xkcd’s exploration of online spaces and the impact of technology.

It’s a thing that a lot of people are doing. I mean, it’s a thing that also everyone who’s trying to build a new internet service or who works at a tech company and is designing a new platform — they’re all also trying to figure out the next big thing. And there’s no end of media people who are trying to figure out why the next big thing might be bad or, Are scooters destroying our cities? or whatever.

I try not to be reflexively critical, and also try not to be just excited about every new tech thing without worrying about the consequences. It’s sort of like with How To: It’s not that want to find why ideas are bad or find why they’re good. I just want to take an idea and explore it, see where it goes, and understand that it may turn out to be really bad — and be a little bit humble about my own ability to tell which is which.