The Economics of Everyday Things - 83. Game Show Winnings

Episode Date: March 10, 2025

How do TV producers decide how much money to give away? A little psychology and a lot of math. Zachary Crockett phones a friend. SOURCES:Bowen Kerins, math teacher and former contestant on Who Wants to Be A Millionaire.Arthur Smith, CEO of A. Smith & Co. Productions and author of "Reach: Hard Lessons and Learned Truths from a Lifetime in Television."Aaron Solomon, television producer. RESOURCES:"‘The Price Is Right’ Celebrates 10,000 Episodes With Extra Big Wins for Contestants That Top Off at $100,000!" by Rosemary Rossi (Variety, 2025)."Why haven’t more game show prizes been adjusted for inflation?" by Janet Nguyen (Marketplace, 2023)."Deal or No Deal? Decision Making under Risk in a Large-Payoff Game Show," by Thierry Post, Martijn J. van den Assem, Guido Baltussen, and Richard Thaler (American Economic Review, 2008)."Why Game Shows Have Economists Glued to Their TVs," by Charles Florelle (Wall Street Journal, 2006). EXTRAS:"Bowen Kerins on Who Wants to Be a Millionaire" (2000).

Back in 2000, Bowen Karens found himself sitting in the hot seat on the game show Who Wants to Be a Millionaire. At the time, he was a 24-year-old math teacher earning around $45,000 a year. And as the lights dimmed and the intense music echoed through the studio, he tried to mentally prepare for a series of 15 trivia questions that could change his life. Let's play who wants to be a millionaire. Here we go. For $100, which of the following vehicles has no wheels?

It really is like you think it is. For $100, which of the following vehicles has no wheels? It really is like you think it is. All around you in every direction, it's a theater in the round thing. And there's the music and there's cameras. It's a strange, surreal experience filled with adrenaline and fear. My motivation was to try and clear my head of everything, except the fact that there's a multiple choice question sitting in front of me.

What Karens didn't know at the time is that the show was carefully designed to anticipate exactly how he would react to all of those stimuli. Because in the world of game shows, little is left to chance. Behind the scenes, producers have simulated the gameplay thousands of times. At every juncture, they know the odds that a typical contestant will get the next question right. And they've carefully baked all of the risk into their budget. You want it to be a struggle for someone to win a million dollars because otherwise it's not interesting. But then again, you have to give out a million dollars eventually because otherwise the audience will be disappointed.

For the Freakonomics Radio Network, this is the economics of everyday things. I'm Zachary Crockett. Today, game show winnings. You've likely seen some kind of game show on TV. There are shows where contestants have to answer trivia questions, complete physical feats, battle family members, or spin a giant wheel. It's a vast and complex part of the TV universe and nobody knows it better than Arthur Smith. than Arthur Smith. I'm the executive producer of Hell's Kitchen with Gordon Ramsay, American Ninja Warrior, The Floor is Lava, Human vs. Hamster, The Titan Games, Mental Samurai,

Prose vs. Joe's. I'll stop. I'll stop. Smith has been in the entertainment business for 40 years and has been involved in more than 200 shows. He says that game shows can have a very successful business model. Game shows as a genre, they're very cost effective. When game shows are produced, you're usually producing multiple episodes a day. And once you build your set and once you have your staff on, the production time is less than any other genre. So there's usually enough room to give money away. The total budget for a game show might range anywhere from

$100,000 per episode for a small daytime show to around a million dollars per episode for a prime time network show. And early in the process of developing the show, the producers and the broadcaster have to decide for a primetime network show. And early in the process of developing the show, the producers and the broadcaster have to decide how much of that budget to earmark for prizes. There's this exchange that you're having with the network. You create the format, and then there's this back and forth

and they go, you're giving away too much money, or you're not giving away enough money, or cut your lighting budget because you need to give away more money. The types of prizes shows choose to give away can vary. Some daytime game shows use material goods as prizes. The Price is Right, for instance, has given away more than 9,000 cars since debuting in 1972. Networks used to get these items for free in exchange for giving brands publicity on national television.

But these days, that's rarely the case. There may be some small discount. You know, hey, if you give away 30 cars, we'll cut the rate a little bit for you. But it's no longer the case where that's a cheaper way to get by than actually having to pay cash. Aaron Solomon is a game show producer who's worked on shows like Weakest Link and The Hundred Thousand Dollar Pyramid. He says in today's game show environment, cash is king.

I think the drama that they're looking for on network primetime shows is people winning life-changing money. It's hard to do that when you're winning a microwave or a jet ski as opposed to cold hard cash. If it's a to do that when you're winning a microwave or a jet ski, as opposed to cold hard cash. If it's a big network that has a big budget and they're looking to make a big splash, oftentimes dangling a million dollar or at least a high six figure grand prize is part

of that. In many of these game shows, the prize is the main attraction and the source of the drama. There are certain formats that really are dependent on prizing. Some examples of that would be Deal or No Deal, where the banker is offering you $150,000 to walk away, but you have to decide whether it's worth it to play it safe or risk it because there's still a $500,000 suitcase or a million dollar suitcase out there. Those kind of shows really depend on having a large cash jackpot to dangle because

the drama and the decision making process is not nearly as suspenseful if you're talking about smaller dollar amounts. But a bigger prize doesn't always make a show successful. In 1999, shortly after ABC premiered Who Wants to Be a Millionaire, Fox tried to one-up them with its own show called Greed. And one of the things they wanted to do was dangle a $2 million cash prize,

thinking that they would one-up ABC with their paltry $1 million grand prize. And what they found out was even though Greed was a modest hit for them, it didn't get double the ratings. And in fact, it didn't exceed the ratings of Millionaire. With high profile letdowns like Greed, it maybe isn't surprising that game shows have gotten less generous with their prizes. Solomon says that back in the 2000s, the prize budget worked out to around $150,000 or so per episode for a big primetime show.

Now, he says, it can sometimes be closer to $60,000. Of course, those are averages across an entire season, which are skewed by the extreme outcomes. But even the grand prizes have become stingier. Who Wants to Be a Millionaire hasn't permanently raised its top prize in over 20 years. A million dollars in 1999, when the show premiered, is the equivalent of only around $530,000 today. I think it's just simple economics. is the equivalent of only around $530,000 today. I think it's just simple economics.

We've got so many networks that are on right now. Viewership trends are changing. The younger generations are watching their phones and short form programming instead of broadcast television. If ad dollars are down, then everything about the production needs to also decrease, including prize budgets. Once a game show's budget for cash prizes is set by a network, a producer like Solomon will know roughly how much money he can afford to give away in each episode.

If this is a 10-episode primetime run and they're giving us a total of, let's say, a million and a half dollars, then we know that on average we're trying to shoot for a payout of $150,000 per episode. If people go over that, the network will end up having to pay for that. So it's in everyone's interest to make sure the format protects against that. With a game show like Who Wants to Be a Millionaire, the economics seem a bit unpredictable. In theory, every contestant could walk away with a million dollars. Producers have to prevent that from happening by controlling for risk. And Arthur Smith says the process begins with lots and lots of trial runs.

You bring in real contestants. It's not the producers playing. You cast real contestants and they play the game. You literally play them in a conference room and you simulate the game. Smith's team will do dozens or even hundreds of these trial runs. And they're constantly making adjustments

to the trivia questions and show design. They'll do this until the prize outcome is close to their desired average. You play the game multiple times and you go, you know, this is too easy, this is too hard, no one's going to win any money on this show unless we make the questions easier. Or, oh my God, we've gone way too far the other end, everyone's going to make too much money and it's not going to be interesting. Aaron Solomon says another important part of the testing process is choosing how to

design the progression of the prize money, or what they call money ladders. On who wants to be a millionaire, the total prize money goes up in increments over the course of 15 questions, from $100 to $1 million. At any juncture, the contestant can choose to walk away with what they've already won, or they can proceed to the next question and risk losing thousands of dollars. The conventional wisdom is that the typical contestant will, once they hit the hundred thousand dollar level, they're going to probably walk away. We want to make sure that the structure of the game is so enticing that once you hit that hundred thousand dollar level,

the leap from there to the next level is substantial enough that it just might tempt somebody into going for it. You look at a lot of money ladders for million dollar game shows, and you will see that the jump from a hundred thousand dollars often goes to two hundred fifty thousand dollars. The establishment of those thresholds on the money ladder can make all the difference in the world about whether people actually go for it and make it an exciting game or if they just predictably quit. The producer's decision-making requires a combination of probabilistic math and human psychology. You're predicting what is the likelihood that a contestant will want to go on.

And that starts to become above my pay grade. All of this testing only tells producers so much. No matter how people do in testing and what results you get in run-throughs, nothing compares to the actuality of standing under the lights in your moment. Whatever hypotheses we had about how the game would play out, suddenly, in real life, human behavior

plays very differently. Smith says that can sometimes lead to nightmare situations. I've heard stories where they tape the show, and, you know, episode one, two, and three, they gave away like the entire prize budget for the season and they're completely freaking out. Because testing is imperfect, producers like Smith and Solomon have to turn to another group of professionals who know the ins and outs of both game shows and statistical analysis.

Once it gets to a certain point of sophistication, that's typically where we would hire a game show statistician. That's coming up. Behind the scenes of your favorite game show, there's a mathematician. By day, I am a math curriculum writer for elementary school math materials. And by night, I am the actuary or mathematical advisor to television shows such as Deal or No Deal Island currently on NBC. Again, that's Bowen Karens. He's the math teacher we heard from earlier who was a contestant on who wants to be a

millionaire. After his appearance, he decided to combine his two passions, math and game shows, into a side hustle. Today, he's one of a small group of contractors who are brought in by networks during the testing process to run statistical simulations on prize outcomes. During a game show, a contestant has to make a series of decisions. Karens uses a software program called Matlab

to work out all of the possible outcomes these decisions might lead to. Someone has a choice between a dollar and a hundred thousand dollars and the show says, I'll tell you what, I'll give you thirty five thousand dollars for sure to just go away. What do they do? That's part of what I have to answer. What my software is able to do is look through the decision tree. Okay, well, I have at least a 60% chance of surviving. decision tree, okay, will I have at least a 60% chance of surviving if I take this deal? I should do that. Do I have less than a 50% chance of surviving?

Then I should not take the deal. And it's just a long set of choices, any of which decide whether someone will or won't take a deal. So that will simulate one run of one episode. But now, it's not enough. I actually run 100,000 iterations of that one episode. Karens does this for every single episode in a season. All the reports go into Excel where I will describe like, all right, the most likely thing to happen is this.

The worst case scenario of how much it would cost is this. The entire structure of the game can be changed based on what we see in the math. If the results of the analysis are unfavorable, producers might adjust the questions to be harder or even reconsider the show's format. But in the event of a big payout, they have a few protections in place. Many shows will pay their grand prize as an annuity, say $25,000 a year for 40 years instead of a lump sum of a million dollars all at once. This safeguards their annual budget from a potential disaster. They say it in the little part of the broadcast that nobody ever looks at, at the very end, where like 500 things are flashing

by you at once. And networks have another backstop. A lot of these reports come down saying, here's your average, but also you've got a 25% chance you're going to give away more than X. And that's where we start to involve insurance services. Insurance companies have underwritten game show winnings for decades. That might seem like a simple case of risk management. But as who wants to be a millionaire learned back in 1999, things can get complicated.

Again, here's TV producer Aaron Solomon. Everybody was watching it, you know, practically every night. And then after a certain amount of time, there was a sense of, okay, well, is someone not going to win the million at some point? So the first contestant who actually did win a million dollars, his name was John Carpenter. The conventional wisdom was that the questions that he answered were pretty noticeably easier than ones from previous episodes. And the perception was that the producers must have deliberately made the questions just a bit easier to incentivize somebody to not only go for it,

but to increase the likelihood that they would actually win the million dollars. For the viewers at home, that helped keep the show exciting. But to the insurance company, it might have looked like the producers were messing around with the odds in order to keep their ratings up. In 2000, a group of insurance brokers associated with Lloyd's of London sued Millionaire claiming that the producers had made the show too easy. If the producers, by their own discretion, decided to suddenly make certain questions much easier than they agreed upon, well, that changes everything in terms of their actuarial figures.

Insurance companies are a lot more particular these days. They're wanting to hold production companies more accountable and be more specific about the show they're actually insuring. When insurance companies do agree to underwrite a game show, they don't just want to see the show's estimate of the average winnings per episode. I have to share with them the entirety of my assumptions and my heuristics about why I think the players will behave the way they do. And then they have to pick over it and verify that they agree with it, both the quality of the computer programming as well as the quality of the assumptions. And then if they are happy with what they've got, they will take it on.

The scary part for everyone involved is that for all the simulations, in-person trials, budget hijinks, and insurance underwriting, you still can't predict what will happen once the cameras start rolling. 25 years ago, when Bowen Karens appeared as a contestant on Who Wants to Be a Millionaire, he had a strategy to distract himself from the psychological pressures of the set. If you watch my episode on TV, you'll sometimes see me looking up. And the reason I'm looking up is because that's the only direction you can look where there isn't somebody looking at you.

And if you look up, you don't see any of that. You get to chill out and be back to a normal person for a second. But in the end, he didn't walk away with a million dollars. His performance, as it turned out, was just about statistically average. I won $32,000 in 20 minutes and I'm still looking for that kind of a rate on new jobs. It feels like it didn't happen and it feels like it did happen. I've got the check on my wall. I've got the video evidence and it's like something

you can barely wrap your head around. For the economics of everyday things, I'm Zachary Crockett. This episode was produced by Michael Waters and Sarah Lilly and mixed by Jeremy Johnston. We had help from Daniel Moritz Rapson. And thanks to listeners Devin Powers, Bernard Midgley, and Mario Milosevic, all of whom suggested this topic. If you have an idea. All right, until next week. If the game show was a movie, how do you want your movie to go?

You want a little disappointment, you want a few laughs, and then you want a big celebration. Big celebration! The Freakonomics Radio Network. The hidden side of everything.