Pod Insights Unexplainable How to beat roulette About this episode You’ll need your best friend, a computer in your shoe, and a working knowledge of physics.
Guest: Doyne Farmer, Director of Complexity Economics at Oxford
For show transcripts, go to vox.com/unxtranscripts
For more, go to vox.com/unexplainable
And please email us! unexplainable@vox.com
We read every email.
Support Unexplainable (and get ad-free episodes) by becoming a Vox Member today: vox.com/members
Help us plan for the future of Unexplainable by filling out a brief survey: voxmedia.com/survey. Thank you!
Learn more about your ad choices. Visit podcastchoices.com/adchoices
Transcript Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.
0:01 Noam Hassenfeld? Anyone who listens to Unexplainable knows that music is really important to me. So I want an amazing sound system, but I also want something that's easy, that just works. Sonos delivers on all fronts. The sound is incredible, and my Sonos system follows me wherever I go. I can mix and match speakers in any room in my apartment, louder in the living room, quieter in the kitchen, and the Sonos app makes it all super easy. Visit sonos.com and build your system. Sonos: Love the sound, feel the system.
0:39 Speaker 01 I'm Anna Garten, and each week on Happy Hour, I invite new and old friends for a drink and a fun conversation at my kitchen table in New York City. On this week's episode, I'm joined by Chloe Mal. We talk all about her taking on the top job at Vogue, finding her personal style, and her favorite lunch order in the iconic Conde Nast cafeteria. New episodes of Happy Hour are available every Wednesday. You can watch on YouTube or listen wherever you get your podcasts. Cheers.
1:17 Speaker 02 All right, numbers are going up.
1:19 Speaker 03 Okay.
1:19 Speaker 02 We're recording.
1:20 Speaker 03 Good.
1:22 Speaker 02 All right, so to kick things off, how did you first become a scientist?
1:27 Speaker 03 Well, you know, I dropped out of graduate school for a while to beat the game of roulette, and that's how I became a scientist, I guess.
1:40 Speaker 02 Wow.
1:44 Speaker 02 Today, Doyne Farmer is a physicist, an economist. His work helped define the field of chaos theory. But before all that, nearly 50 years ago, he was gaming casinos. I gotta, I gotta ask about the roulette. Whose idea was it to apply physics at the casino?
2:04 Speaker 03 It was my friend Norman Packard, who I'd known since we were about nine years old.
2:10 Speaker 02 Doan and Norman had been Boy Scouts together in southern New Mexico. Silver City is on the edge of millions of acres of national forest land. It's also a university town. And both boys, self-described science kids, were inspired to go to grad school for physics. So by the summer of 75, Doan and Norman were very smart and very poor students.
2:37 Speaker 03 We were always trying to find a nice way to make money during our summers.
2:41 Speaker 02 Doan was making more money playing poker than at his summer job with the Forest Service in rural Montana. Reading The Complete Guide to Winning at Poker was enough to give him an edge and a taste for winning money. Norman, on the other hand, spent his summer in Vegas playing blackjack.
3:00 Speaker 03 But then while he was playing blackjack and staring at roulette wheels, he got this idea that we were physicists. We should be able to beat roulette.
3:10 Speaker 02 Roulette is a game that feels at home in James Bond movies or Casablanca, where tuxes and ball gowns gather around the table to bet on the epitome of chance. As the spinning ball falls onto a spinning wheel, it can land on one of 38 numbers. You can bet on the ball falling on red or black, an even number or an odd one, or cryptic combinations of numbers. But ultimately, it's simple.
3:43 Speaker 02 Predict where the ball is going to land. And if you guess right, the payout can be as much as 35 to one. Way higher returns than blackjack or even a really good hand of poker. The thing is, though, there's not really any strategy. There's no opponents to bluff, no cards to count, just pure random chance. But Norman saw potential. What, what was his pitch? Like, why, why should physicists be able to beat roulette? I mean, wasn't there like an Einstein quote that no one could win at roulette unless you steal the money from the table?
4:20 Speaker 03 Yes, that's right. But there's something Einstein didn't think of, but he should have actually, which is that roulette is a, it's just physics. There's nobody making any decisions. It's just physics.
4:34 Speaker 02 The casino croupier, not a dealer, spins the wheel one way and then spins the ball the other way. And then it's a little more complicated maybe than playing catch in the yard. But Norman and Doan's reasoning was largely the same. If you know where the ball is, what direction it's going, how fast it's going, you can have a pretty good idea of where it'll end up. And yet roulette is this random number generator. Even Einstein supposedly didn't think there was any rhyme or reason to it.
5:07 Speaker 03 And that's because tiny differences in the way the ball is thrown can cause big differences in where it lands.
5:13 Speaker 02 But there's an opening because casinos keep the table open for bets after the croupier throws the ball. Maybe 10, 15 seconds of golden opportunity. That sweet spot after the throw that makes the ball unpredictable, but before it's had a chance to land. Doan and Norman had to figure out where the ball was going to land in those precious few seconds when roulette was just a predictable physics problem.
5:49 Speaker 02 So how did you figure out how to beat roulette?
5:54 Speaker 03 Well, you know, we started by thinking, okay, it's a rolling ball on circular track, but it's slowing down. What causes it to slow down? We realized it's air resistance. So then we went to the library and looked up air resistance and thought about that. And so we wrote out an equation that described a rolling ball under circular track with air resistance and solved it.
6:17 Speaker 02 Okay.
6:18 Speaker 03 And then we bought a roulette wheel and we started doing experiments to see if...
6:22 Speaker 02 Wait, where'd you set up the roulette?
6:24 Speaker 03 Well, the roulette wheel was set up first in my dorm apartment. I was a graduate student advisor in the dorm. And then we would take it into the physics department in the middle of the night where, you know, I had a master key and the janitor was in on it. So he didn't tell the faculty we were doing this in the middle of the night.
6:47 Speaker 02 Doan and Norman spent their nights doing experiments on the roulette table and solving their equations on bulky 1970s computers. But they weren't going to make any money in an empty physics department at midnight. They needed to take their strategy to the casino and they needed to bring their computers with them. So naturally, they built the first ever wearable, concealable, digital computer.
7:18 Speaker 02 The idea of a personal computer, let alone one that you could carry around with you, was pretty out there at the time. Doan and Norman couldn't just order up parts and start tinkering. They had to set up a shell company to buy the components they needed. They used the same processor that went into their original Apple computer and even ran in the same circles as Steve Jobs and Steve Wozniak. Though Doan and Norman couldn't quite imagine what regular people would do with personal computers, they were completely obsessed with roulette.
7:55 Speaker 03 Doing experiments, building computers, solving the equations, writing programs that would solve the equations on the computer we built.
8:04 Speaker 02 The project, as they called it, grew from just Doan and Norman to about 30 people. They found that they could use their equations and computers to roughly predict where the ball was going to land. And, you know, even if the ball bounced around a little as it fell off the track, it wasn't going to bounce clear over to the other side of the wheel. They could narrow it down enough to cover the bets. They had it. The dream was to win enough money to fund a science commune where they could live and study and work together outside of the pressures of academia. But first, they had to hit the casino floor.
8:52 Noam Hassenfeld? Support for the show comes from Sonos. Hey, Sonos, give me some music.
8:58 Noam Hassenfeld? All right. One of life's great mysteries. You're listening to music in the living room, but you gotta grab something from the kitchen. You don't want to take the speaker with you, but you also don't want to miss that beat drop. And this is where I introduce you to the magic of Sonos. Sonos sent me a couple of speakers to try out, so I've got to move two here in the living room. And then I walk down the hall. I get in here to the kitchen, and I got a brand new Sonos Play right here on the counter. I can open the fridge, grab a seltzer.
9:41 Noam Hassenfeld? And yeah, I didn't miss a second. And I can make each speaker louder or softer in the Sonos app, or if I'm using AirPlay, so I can make it louder in the kitchen without blasting whoever is still in the living room. It's all just easy, and it sounds amazing. These speakers are crisp, they've got great bass, and I'm especially impressed with the new Sonos Play. It packs this like truly awesome sound in a much smaller size. So, I don't know, if I want to head outside for a minute. It's super easy. And yeah, pretty nice out. Visit sonos.com and build your system. Sonos, love the sound, feel the system.
10:32 Speaker 04 Support for the show comes from Quince. Fall is my absolute favorite season. I love fall colors, I love fall leather, I love the specific light you sometimes get on a fall afternoon, and I really love fall clothes, which Quince has. This fall, Quince is offering pieces with premium materials like 100% Mongolian cashmere, organic cotton, and merino wool. The styles are timeless, versatile, and designed to become the dependable pieces you reach for day after day. Their 100% Mongolian cashmere sweaters start at just $60, giving you the softness and quality you'd expect from a luxury brand without the luxury price tag.
11:15 Speaker 04 I think my favorite thing about Quince clothes is that they're just good wardrobe staples. I got three black shirts from them, each in a slightly different style, and they are just correct. They fit the way they are supposed to, they look nice, they feel nice, and they go well with the rest of my closet. Find your next fall favorites at Quince. Download the Quince app for app-exclusive offers or go to quince.com. Get free shipping on your order and 365-day returns, now available in Canada and the UK, too. And when Quince asks you where you heard about them, please let them know it was unexplainable. It's a great way to support the show.
11:56 Speaker 05 Are you tired of hearing about business, tech, and finance from the bros? We are the Tiger Sisters, a top business podcast with the Vox Media Podcast Network. Tiger Sisters releases every Monday, and we decode money, power, and love. And who doesn't love all those things? We are the internet's Wall Street and Silicon Valley big sisters. Tune in every Monday for late-night sister talk meets boardroom strategy. Check out Tiger Sisters on YouTube, Spotify, or wherever you get your podcasts. Welcome to the Tiger fam.
12:30 Speaker 06 Do you wish to place another bet? No, no, I guess not. Drive 22 tonight.
12:38 Speaker 02 So you had the strategy, you had the math. Walk me through, through the heist. What was, if I'm imagining like an Ocean's Eleven theme song playing?
12:47 Speaker 02 What was the heist that you guys?
12:48 Speaker 03 Yeah, so we'd start out by, in our apartment, we would suit up, which meant putting on the computer.
12:56 Speaker 02 Alright.
12:57 Speaker 03 You know, the computer was strapped under one arm. There was a 12 AA batteries strapped under the other arm. Back in those days, it took a lot of power to run a computer, and we could only run for about an hour and a half.
13:08 Speaker 02 Okay.
13:09 Speaker 03 So you'd strap on the battery pack. It was all sewn into this belt, like weight lifters would use. And then we had switches in each of our shoes with wires running up.
13:21 Speaker 02 Like up your leg?
13:22 Speaker 03 Up our legs. And we had a kind of radio transmitter around our shoulders. And we had little buzzers. So the buzzers, they were like thumpers. So they'd give us a low-frequency thump, medium-frequency thump, high-frequency thump.
13:38 Speaker 02 On your chest?
13:39 Speaker 03 Uh, actually, the men typically put them kind of under their belts. And the women had them sewed into a bra, so they would give them a vibration in their bra.
13:50 Speaker 02 Okay.
13:50 Speaker 03 So we would suit up in the apartment, and suiting up already was kind of tricky because you had all these wires and things you had to plug together. And then we would get in our car. It was usually my 1966 Dodge van, and drive to the casino.
14:10 Speaker 02 Dohen and the team made 11 trips to Nevada and all. They'd spill out of Dohen's van, adjusting the wires in their belts and bras, and head into the casino.
14:22 Speaker 03 And then we would look around for roulette wheels with a little bit of tilt. Turns out, only a tiny bit of tilt would make the dynamics much more predictable. Because it caused the ball to be more likely to land on one side than the other.
14:36 Speaker 02 And then one of them, usually Doan, approached the table. He'd act stupid and loud, call himself Clem, and bet small stakes. But all the while, Doan was using his big toes to click the switches inside of his shoes to program the computer tucked into his weightlifting belt. Using all this, he'd measure the speed of the ball and the rotor turning the wheel.
15:02 Speaker 03 Depending on the friction and the rotor, it might slow down faster, or the ball's a little more dense. Or even if the barometric pressure's changed, the ball will accelerate faster or slower.
15:15 Speaker 02 Once he had a feel for that particular tilted wheel, with that specifically dense ball on a fair-weather desert night, Doan and his computer were ready.
15:27 Speaker 03 I would signal the other player, and I would just be betting red or black. And I'd say, "Switch to betting even odd." So the other player would know, "Ah, now he's ready." And then the other player would take their place on the opposite side of the table. And my computer would relay to their computer when the ball was about six seconds ahead of falling off. It would tell them what part of the roulette wheel was the best part to bet on. And they would have memorized the pattern. So if they got a buzz that meant, say, seven, they'd know, "Okay, that's the seventh octave, and the numbers in there are 14, 32, and 3."
16:11 Speaker 03 And they'd lay down bets on those numbers, which were all in about the same part of the wheel, and that's how it worked.
16:20 Speaker 02 Yeah, and this is all on taps in the shoes and thumps under the belt or in the bra. And that's how you're communicating, trying to stealthily do all these calculations.
16:31 Speaker 03 That's exactly right.
16:33 Speaker 02 This system, it was complicated.
16:36 Speaker 03 We always had a lot of problems with hardware. For every hour we spent playing, we probably spent five hours fixing our equipment.
16:44 Speaker 02 And it didn't always go as planned.
16:47 Speaker 03 It could drive you crazy because the ball would land just short, just right next to the number you made your bet on. You could be making good predictions and still end up losing just from bad luck, which is part of why we were always nervous about having a big enough bankroll to outlast the bad luck.
17:07 Speaker 02 But trip after trip, it paid off.
17:10 Speaker 03 We make predictions and they were far from perfect, but they were still good enough to give us about a 20% edge over the house. So in other words, every time we laid down a dollar, on average, we got a dollar 20 back.
17:22 Speaker 02 Nice.
17:23 Speaker 03 That was pretty good, but we were always a little cowardly about scaling up to big, big stakes because every time we started to do that, we'd fill a lot of casino heat.
17:34 Speaker 02 Hmm, interesting. Feels like a lot of trouble to go through for a little extra pocket money if you consider all the like man hours put in.
17:43 Speaker 03 It was, it was, but it was a lot of fun, I must say.
17:52 Speaker 03 It was a engineering project where we had to do the basic science. We had to do the math. We had to do experiments. We had to build a device. We had to get the device to work in the field. My thesis advisor said, actually, you could just write your thesis about this if you wanted. And I decided not to. I didn't think that really would qualify me to do much else, but it did actually in the sense that it taught me a lot about predictions.
18:22 Speaker 02 About a decade later, Norman, Doan, and some friends founded The Prediction Company in 1991. They kept up the grad school vibes, setting up their company in a barely furnished adobe house in Santa Fe, and basically sharing one business suit between them to wear to meetings with investors. Just like their days playing roulette, Doan and Norman harnessed the underlying science, invented the technology they needed, and profited by finding predictable patterns where others saw randomness.
18:57 Speaker 03 We knew a lot. We'd learned an awful lot from that first enterprise, and we put a lot of the things we'd learned into practice, beating financial markets.
19:09 Speaker 02 At a time when trading floors still ran on brokers in colored blazers shouting out orders and throwing up hand signs, The Prediction Company used computers, mountains of data, and even some early machine learning to find these predictable patterns and automate trades.
19:28 Speaker 03 In stock markets, in bond markets, in commodities markets, foreign exchange markets.
19:34 Speaker 02 Basically, they started one of the very first quantitative hedge funds.
19:39 Speaker 03 Exactly. That's what we did, inspired by, in many respects, our time in the casino. And we did succeed in making a decent amount of money. We made a pretty good, we made quite a, a pretty good amount of money doing that.
20:01 Speaker 02 If you want to hear more about Doan's work, teasing out those patterns he keeps seeing in the economy, check out our episode, "Embracing Economic Chaos."
20:13 Speaker 02 This episode was produced by me, Meredith Hoddinott. I also run the show. We had editing from Jorge Just, sound design and mixing from Christian Ayala, music from Noam Hassenfeld, and fact checking from Melissa Hirsch. Julia Longoria is looking a little green, and Byrd Pinkerton got to the airport. She waited till 12 and made her way outside, and at the far end of the abandoned runway, she saw a pigeon. Ah, she thought, that must be Aaron Bird. But next to him, she saw something else hiding in the tall grass. It couldn't be a platypus?
20:54 Speaker 02 Thanks, as always, to Brian Resnick. Hey, O'Brien. Thank you so much for creating the show. And if you have thoughts about the show, you should send us an email. Brian, specifically, you should email us, but also any other listeners. Really excited to hear what you think about the show. We're at unexplainable@vox.com. You can also leave us a review or a rating wherever you listen. Really helps bring new listeners into the fold. Tell the world about Unexplainable. It's gonna be their new favorite podcast. They'll thank you for it. You can also support us and all of Vox's journalism by joining our
21:33 Speaker 02 membership program. You can go to vox.com/members to sign up. And if you do, you'll be helping make this place run. You also get unlimited access to all the phenomenal reporting on vox.com. You'll get exclusive newsletters, and you get all of our podcasts ad-free. Unexplainable is a part of the Vox Media Podcast Network. We will see you next week.
Previous Is climate change really making hurricanes worse? Next The musical structure of the universe Page updated November 3, 2026 · transcript and summary are machine-generated