Black Box Mechanics
Black Box Mechanics
And the Ethics of Gambling in Games
I got a question about ethics and gambling game design on Twitter last week. It was a hard one, not quite appropriate to a 280-character answer, so I’m writing this article to attempt a more complete response.
I will warn you now that my conclusion is “it depends.” This post is more a discussion of my perspective on this question, and what I think we’re talking about, but there’s no definitive answer at the bottom. Just perhaps a little help in finding the answer for yourself.
The Question
EP Walker (@voxen) asked:
@cheapassjames Is there such thing as an ethical payout? Does the quality of experience increase with the stakes, or is it arbitrary?
This was part of a thread about ethics in game design. It is a brief conversation that runs the gamut from the ethics of games without agency, to the use of regret as a design metric. Since this question was directed to me, I figured I’d try to expand here on my initial reply, which was:
Are gambling games ethical at all? That's the root question here and of course the answer is complicated. In theory? Yes. In practice? Not always.
Let’s see if I can do that answer a little more justice below.
Background
I invent games. I’ve been doing this professionally for 27 years. I focus on lightweight and traditional-style tabletop games. I’ve also worked in digital games, mostly casual online games, mobile games, and console games. I’ve designed several gambling games on multiple platforms, including practical, digital, and casino table games. I’ve designed monetization strategies for free-to-play games, including games for children, and I’ve written many classic pub-style gambling games including the Pub Games in Fable II. Recently I sold a casino table game called Ricochet Poker. That game debuted in the month before the Covid lockdown, so that was rotten timing to say the least.
I’m also a gamer. I grew up playing and inventing many kinds of games. I learned draw poker from my grandmother, and license plate bingo from my grandfather (on the other side). He was an enthusiastic casino gambler and regular card player, who introduced me to the Martingale betting strategy, all of his favorite games, and even Las Vegas itself. As an adult, gambling is still a large part of the games I play, both at home and in the casino.
As a game inventor, I think gambling games are a fascinating design challenge. To succeed, these games must be extremely simple, yet still enduring and fun. Packing hours of replayability into a tiny set of rules is not easy, even when you’re designing for casual play. Selling games to the casino industry is vastly harder, but it’s more a challenge in deal-making, not design.
Unlike most of my design essays, this article isn’t about how to make these kinds of games. It’s more about whether we should.
“Gambling Mechanics” and the Black Box
I often see folks in computer games decrying the use of “gambling mechanics” in games, but I think we need to be a little more specific. Gambling games use many different mechanics, most of which are found in other games too. Dice rolling, card shuffling, bluffing, counting, keeping score. Gambling games have been around far longer than for-profit casinos, and “gambling mechanics” is a vague term that could describe rules from almost any game. So let’s start by talking about what we’re talking about.
Gambling, by definition, is any game that is played for money. This could describe literally anything, but at the top of the list are games where money is the score, and the mechanics are predominantly random. Players wager money on each round, and over the course of play, they try to win more than they lose. The list includes games like poker, roulette, prop betting, gin rummy, and many other games with an element of chance. People also gamble on skill games, like darts and billiards, but those are not really “gambling games” in the same sense.
To qualify as gambling in the eyes of the Law, a game usually has to pass the “three C's” test: Consideration, Chance, and Compensation. If players must put money in, and if there is a random element (sometimes this must be “primarily” a random element), and if players can get money back out, then it’s probably gambling. If the game fails any of these tests, it’s probably not. (“Compensation” is usually listed as “Prize,” but that wouldn't be three C’s.)
But when I hear “gambling” from digital game designers, I think they’re usually talking about one very specific type of mechanic: the “black box” or Jackpot game, which often doesn’t even satisfy the three C’s.
Basically, a black box is a game element in which a player places a wager and receives a random result. This result could include the same currency as the wager, as in a slot machine, but in most digital games it’s paid out as items of value within the game. Those items might or might not be convertible back into the original currency (and this conversion may either be sanctioned by the game, or it might take part only on an external marketplace).
Games and mechanics in the black box category include casino slot machines, RPG loot crates, and trading card game booster packs. For a fixed and repeatable investment, usually in real money (a wager), a player receives an unknown reward picked from a variable pay table, and that reward is paid by the game itself, not by other players.
The random reward schedule from this Skinner box mechanism makes it susceptible to pathological abuse (read: addiction). When the game is played for real money with no mitigating factors (such as online moderators, real world shop keepers, limited product supply, or observant casino staff), this addictive play can have consequences for the player. And that can happen whether the prizes are worth real money or not.
So does the potential for addictive behavior mean that all black boxes are unethical by nature? Nah. Let’s get into those weeds one step at a time.
Ethical Payouts
The first part of the question: Is there such a thing as an ethical payout?
On the surface, this seems like a question of mathematics. It sounds like it could be restated as “is there a value of return below which no games should be acceptable?” But I think it’s impossible to answer the question in those terms, because the only defensible answers seem to be “no way sucker, anything goes,” or “yes, and that limit is 100%, because gambling is bad.” It would be hard to make the case for some other arbitrary threshold, say 85%, being the precise breaking point.
So let’s define some terms here and then come around to what I think is the answer, which is less about mathematics and more about consent.
A “payout” from a black box game can be expressed as a percentage value, often called the RTP (return to player). This is calculable if the output is the same as the input. In other words, it’s a slot machine, money-in and money-out. With a booster pack or loot box, the rewards can only be valued in the input currency if they have firmly established values. Game developers usually try to make this evaluation complicated or impossible, precisely to avoid the implication of gambling.
Anyway, for clarity let’s talk about the slot machine. The RTP of a slot is its expected value, is its average payout over time. If a game has an RTP of 95%, that means for every $100 a player bets, she will win an average of $95. This is not a guarantee, just an average of all equiprobable results.
Of course, these games have wildly variable pay tables, so you will rarely if ever see exactly a 95% return. What’s more, you can never play long enough to guarantee that you will hit that average , and you’re much more likely to walk away with less than average, because only a small number of players will hit the biggest jackpots.
In fact, even if a slot machine had an RTP of 100% (or more), it still wouldn’t pay that amount back to all players. A large chunk of the prize money would still be paid out as a few big jackpots, and so most players would still walk away with a loss. This is part of why it’s difficult to choose a perfect RTP and say “below this value, a game is no longer good.”
Because of the random distribution of prizes, the nonlinear value of money, income taxes, and a thousand other factors, the only way to make a 100% payout machine completely equitable would be to replace the black box with a clear plastic tube, guaranteeing that every dollar wagered produces exactly the same dollar in return. Obviously this isn’t a game anymore, but I feel like the plastic tube is the non-gambler’s fantasy of the perfect gambling device.
Entertainment Value
But here’s the thing. The return of a game isn’t entirely measured by the money won or lost. Gamblers generally understand this: In the long run, they know they are probably going to lose money. They understand this principle even when they can’t explain the details. They enjoy the challenge of beating the house, but they know it’s not likely. Instead, they write off their losses as the cost of entertainment.
I know game designers, even in the casino business, who think this attitude is insane, and that gamblers are just stupid. As a gambler I take offense. I feel that the money that I risk (and usually lose) is spent on entertainment, in the same way as money spent on a movie ticket or a bike rental. I don’t expect to “come out ahead” on either of those transactions either, but I do expect to have fun. If I don’t have fun, I don’t rent the bike again.
In the world of digital games, the entertainment value of the loot crate or booster pack is clearly the point of the purchase. Within the context of the game, that’s exactly what you’re supposed to be paying for, and it doesn’t need to have any real-world value. It’s only questionable for reasons we will talk about later.
But first, let’s talk about keno, which is mathematically one of the worst games in the casino. Travel with me now, back to the winter of 1977. You’re sitting in a breakfast restaurant near the casino floor, waiting for waffles and your bottomless coffee refill. There’s a keno board at the back of the room, displaying a new game every five minutes. Those games are broadcast live through the magic of electrons from a keno lounge at the other end of the casino. And there’s a cute keno runner who comes to your table and sells you a ticket for the next game.
You pay a dollar, circle five numbers on the ticket, and send it off to the lounge, where 20 numbered balls will soon be pulled from a group of 80 to decide your fate. You picked lucky numbers like your daughter’s birthday, your license plate, your favorite quarterback. For your dollar, you stand to win as much as $400 if all five of your numbers come up.
You watch with excitement to see if your game will be posted before your food arrives. While your game is playing, you watch as those 20 numbers are drawn and posted one at a time, checking and rechecking your ticket to see if yours were drawn. Each time you make a match, you imagine what that $400 will buy you.
And then you lose, because you usually lose. You know you will probably lose most of the time, but a jackpot once in a while would be nice. It might even get you back even.
Mike Shackleford’s 2001 survey of the keno games in Las Vegas showed them with a range of RTPs between 65% and 80%. I’ve seen keno pay tables as low as 54%, including online games that are running right now. So if expected value is your only metric, keno is a disaster. And most mathematicians experience nothing but dread when they play.
But my grandfather loved this game. Was he wrong? It’s hard to convince someone they aren’t having fun. Did he know it was a losing proposition? I think he did. But I think he also would have told you that the excitement of possibly winning $400 was worth the dollar, and I don’t think he was wrong.
In keno, you’re paying for drama and convenience. Something to do while your food comes. Or at least this was true back in the days of cafe boards and keno runners. Today you can play video keno, with odds just as bad, but instantaneous results. I don’t know if that’s equally fun, but those machines do exist, and people do play them. So, are they having fun, or are they addicted, and is there a way to know the difference? I don’t have an answer to that, but I’m sure it has something to do with whether it ruins your life.
Gambling certainly didn’t ruin my grandfather’s life. He just loved to gamble.
In general, I think visitors to Las Vegas understand that they are paying for entertainment, and they’re okay with that. When I talk to garden-variety gamblers, I sense that they know the odds are against them, even if they couldn’t show me the math. Almost no one is genuinely surprised when they lose. They suffer their losses as part of the fun.
Is the keno player’s excitement over his potential $400 jackpot really worth 25 cents of expected loss? Does he know he’s paying 25 cents? Would he be just as excited if he knew that? Is he even capable of thinking about the game in those terms?
Questions like this are a mixed bag. They presume that 25 cents makes any difference at all to a specific person. They presume that money is the only important thing in the equation, and this presumption is anchored by the fact that money is the only stated prize.
Maybe the player doesn’t know the exact cost of his keno bet. But he knows that the most he can lose is the entire dollar, and he’s okay with that, no matter the expected return. He also knows generally that the games in Vegas are not fair, so the upper bound of his expected return is 100%. Is knowing that your expected return is bounded by an upper limit of 100%, and a lower limit of zero, sufficient for an adult to give consent to play? I think it is.
I’m using keno here because (a) it has terrible odds and (b) it costs next to nothing to play. The exposure on a single $25 hand of Jacks or Better video poker is also about 25 cents, but the potential loss is a lot higher, and hardly anyone plays that game perfectly. Still, for the people I watch playing in those high-limit rooms, $25 is still basically nothing.
My opinion is this: if the rules are clear and the game works exactly as advertised, the game is fair. Not “fair” in the sense that it has a 100% RTP, but in the sense that the player has ample information to make the decision to play.
In the case of keno, the mechanics are breathtakingly simple. Twenty balls will be drawn at random from a pool of 80. Each ball has an equal chance of being drawn. This is both the promise of the rules and the truth of the mechanism. The math to calculate the RTP isn’t trivial, but it’s also not that complicated, and if you can’t work it out yourself there are plenty of books about how to do that. Even in 1977, without the internet in your pocket, it was easy to find a book explaining how to calculate the odds in keno.
Given that this game meets the qualifications of (a) being in a casino, where players are adults and they understand that the odds favor the house, and (b) the rules are sufficient to completely evaluate the payout, this game has what I would consider an ethical payout because it delivers exactly what it promises.
Craps, roulette, Big Six, and other basic casino games fall squarely into this category. Their rules are honest (if the dice are balanced, the wheels are fair, etc.,) and their solutions are simple. Players have everything they need to determine what is at risk. For all the popular games, there is ample literature explaining the rules and the odds for those who are curious but lack the skill to figure them out.
Blackjack is not so simple, but it’s still not hard to form a working strategy and understand one’s level of exposure. The same goes for video poker and a few other decision-based games. Actual poker is an entirely different animal, but for the most part you’re taking money from other players, not the house, so in terms of the house’s edge it couldn’t be more clear: They rake a little out of every pot.
And then there’s the slots, where a player really has nothing to go on, unless they are in a jurisdiction where payouts are constrained by law. If RTPs are not posted, a player can only trust in their personal experience if they want any proof that the return is higher than zero. Still, I consider those games acceptable, because they aren’t making any promises. The mechanics are: you put your money in, the machine makes noise, and that’s the end. As advertised.
What’s an “unfair” game under these rules? Any game that cheats, lies, or is deceptive about its workings. I’m talking about literally breaking the rules as stated, like a blackjack dealer who deals from the bottom of the deck, but only when she suspects one of her players of counting cards. (That player was me.)
I once worked on a cruise ship that offered a daily “mileage pool” game, in which players bought tickets and placed bets on how far the ship would travel by noon the following day. Every day, the cruise director would call the bridge for the actual mileage and then, if necessary, he would substitute a credible alternative that paid out the fewest tickets. Yes, for real.
One could argue that all casino marketing is deceptive, or at least false without being untrue, and therefore to anyone who believes those ads, the games are not honest. But this is how much of marketing works, and it brings us to the issue of context.
Context Part 1
The second half of the question relates to the idea of playing for money. Does the quality of the experience increase with the stakes?
The cliche of “let’s make it interesting” tells us that playing for money is more fun, even though we also know that this phrase is typically followed immediately by someone getting swindled. But it’s certainly true that most gambling games are more exiting when played for real stakes. The decisions in poker, for example, are meaningless without some real risk for losing. And games like craps are pretty dull without a little money on the line.
Games take place in an abstract realm that designers call the “magic circle.” This is the framework within which the rules are valid, defining boundaries of the game’s imaginary space. Monopoly uses fake money and charges fake rent and causes fake bankruptcies and the game is enjoyable nonetheless, because everyone knows that it is imaginary and self-contained, at a level of abstraction within which all those things can be treated as real.
Playing for real money, even the smallest of stakes, shatters the magic circle and makes the game about something bigger, expanding the circle to include the “real world.” (This is a somewhat misleading phrase referring to a larger magic circle called “commerce” where the players agree that money has value. You’re playing it right now.)
So, does a game become more fun when the stakes are raised? Maybe. It’s not really a question about the game. It’s a question about the players.
There is a type of player for whom the magic circle is essential, and games can only be fun if they are contained within it. There is another type for whom the exact opposite is true. For some gamblers, a purely abstract game is a waste of time, because games are only fun if something “real” is at stake. And of course there is a broad range of players in between these extremes, including me.
Las Vegas is a magic circle. When you’re there, the rules are different, and that’s part of the appeal of the place. Actually, that might be all of it.
As a tourist on the Strip, I operate on the premise that everyone is there to rob me, from the casino executives up to the baristas. It’s a nonstop dance of deception, where high rollers earn free rooms and good tippers get sweet secret stuff, while casinos grow bigger and stronger on the backs of all those suckers who just don’t get it. The object of this game is to get the most entertainment for the lowest possible cost. And hey, you might even get lucky.
As I said above, I don’t believe that gamblers expect to win all the time. Rather, their basic approach to the casinoland metagame is to try to make their bankroll last as long as possible, and maximize their fun. If they win, they are happy, but they’re not really surprised when they lose.
This victory condition is usually summarized as “getting free drinks,” but it can mean different things to different players. I don’t drink, but I do play hours of video poker in exchange for comps like free rooms and food. These perks have value to me because I enjoy coming back and playing more video poker, and that value is (slightly) higher than the theoretical amount that I risk on that game. For me, gaming the system is fun, and I’ll go out of my way for a stupid free thing like a coffee mug, because that’s all part of the game.
So whenever I go to Las Vegas, this game is on. The context is known, the players are identified, and the goals are clear: Play smart, get comps, lose as little as you can. In other contexts, the same games would not necessarily be fun, because they are not surrounded by the metagame that is Las Vegas.
A World Without Swimming Pools
Is gambling ethical in any context? Should an industry be allowed to profit on the sale of, well, nothing? Should casino executives drive fancy cars and build even bigger casinos because “math is hard”? Ha ha, but yes.
If you built swimming pools, you might hear questions like “how can you work in an industry where ten people drown every day?” That’s how it sounds when people ask how I can justify creating games for casinos, when ten adults in a thousand meet the NCPG standard for problem gamblers.
Are swimming pools dangerous? Sure. Should there be laws? Actually, there are. A swimming pool is what they call an “attractive nuisance” (insert joke about blackjack dealers here). So if an underage or inebriated person wanders into your back yard and drowns in your pool, it’s your fault, unless you surround your pool with a fence at least four feet high with a self-closing, self-latching gate.
The theory here is that once the victim has overcome sufficient challenges to get into your pool, then it’s then their own fault when they drown. That’s obviously an oversimplification for effect; sorry if you have a history with pools. But the point is, should casino gaming be treated as an attractive nuisance? Obviously yes, and actually it is, and there are strict rules about where casinos can be built, who can play the games, and what kind of odds they can offer.
That last one is the most complex, and the rules about odds differ from place to place. Some jurisdictions don’t allow games with a return lower than, say, 70%. Some places actually don’t allow games that return better than 100%. Some places outlaw card games, some outlaw dice games, and the idiosyncrasies of jurisdictional rules are as politically motivated as they are nonsensical. In Las Vegas there’s no upper or lower boundary on payouts; odds can be as bad as the casino chooses, and it’s caveat emptor, all the way to the bottom of the pool.
So, does the existence of laws for pools imply that pools are evil? I don’t think so. Dangerous, yes, but evil, no. Could unscrupulous pool operators bypass the laws and lure people into the water who have no business there? Of course. Do they? Well… now we’re deciding if it’s unethical to design and manufacture any tool that can be abused, and that debate is probably as old as the flint axe. I’m not gonna settle it here.
One obvious problem with the pool metaphor is that swimming is not, in itself, profitable. Swimmers don’t gradually leak money out of their pockets while they swim, so the industry isn’t actually structured around maximizing revenue from people who don’t swim very well.
Instead, if you want to make money with a swimming pool, you should surround it with ancillary services like restaurants, equipment rental, hotel rooms, mini-golf, and so on. Las Vegas has this stuff too, plus automatic gun ranges, so it wins.
Is there ethical swimming? Of course there is. How can I do it safely? Learn to swim. Or splash around in the shallow end. Is a hotel operator being evil when she invites people into her pool, without really checking whether they know how to swim? Is an age limit or height limit really sufficient to distinguish the swimmers from the non-swimmers? Just because someone can physically open the gate, do they necessarily belong in the pool?
This is how I hear questions about gaming, because I also grew up near water and I love to swim. And even though people drown every day, I don’t think that’s an argument for closing down all the beaches and pools. (In this metaphor, beaches are home games.) It might, however, be an argument for teaching people how to swim.
Before I spent any time in a casino, I learned how the games worked. I read up on the rules and basic strategy for blackjack, craps, roulette. I love knowing how slots work, and I love playing slots despite their terrible odds. I’m a strange case, since most of the game developers I know despise slots and slot players. But as I explained above, I am paying for entertainment, and that’s fine with me.
I talk with a lot of casino players, I think that for most of them every casino game is essentially a black box. They have lots of interesting guesses about how the games work, but to be honest, their models aren’t very accurate. Nevertheless, I think they still meet the qualifications for informed consent. They know generally how these games are designed. They have shown their ID, opened the gate, and stepped into the pool. At this point, the responsibility to swim really does belong with them.
But even the best swimmers I know probably don’t enjoy being pushed into the water.
Context Part 2
The reason that black box mechanics are so dicey outside of the casino is that it’s not always clear that a player has opted into the metagame, or that they are qualified to give consent. Those qualifications are, basically, maturity and intelligence, specifically with regard to the mechanics of the game.
Any repeatable real-money transaction, such as a loot box, breaks the magic circle and turns every armchair into a swimming pool. This is actually true of all such transactions, not just black boxes, and the “caveat emptor” defense becomes a lot more specious when you can’t be sure who your emptors are.
This has less to do with the textbook definition of gambling, since the familiar tropes of gambling aren’t really necessary to make a repeatable transaction susceptible to addictive behavior. In fact, since gambling laws basically make real-money slots difficult or impossible outside a casino, the “gambling-style” mechanic in digital games is rarely technically gambling; it’s a black box purchase with random in-game rewards. And, random or not, any purchase that unlocks progress in a game could turn into a bottomless money hole.
This is, of course, why game publishers love these mechanics. To put it in the publisher’s terms, “We offer a game with a completely flexible price point, where every player can decide exactly how much the experience is worth to them.”
Yeah.
As either Upton Sinclair, H. L. Mencken, William Jennings Bryan, or Mark Twain probably once said, it’s hard to convince someone of anything when their salary depends on not believing it. Ethical or not, predatory or not, this is the world we live in.
In a perfect world, the “flexible price point” defense describes microtransaction-based games as a benefit to a marketplace with variable preferences and needs. In reality, it’s a bit more likely that these products, like real world slot machines, are capable of efficiently gobbling up a player’s bankroll and returning a loud burp.
And while one can expect an average tourist in Las Vegas to be responsible with their bankroll (yeah, not really, but it’s expected anyway), you can’t be so sure when you don’t know who your player is.
So, then, is it the responsibility of the publisher or their platform to qualify players before allowing them to participate in games with real-money transactions? Is it the responsibility of the game designer not to create these games for an unregulated marketplace? Or is it the responsibility of the players (collectively or independently) to “learn to swim” and accept that in this world, the magic circle is impossible to close?
I actually don’t have answers to these questions, but it’s fun to ask them. Most of us have the curse (or the luxury) that our bosses, or the law, or the “invisible hand of the market” will ultimately make these decisions for us.
Drawing The Line
Above my desk is a sign that reads “Write the game you want to play.” I think the contrapositive is valid here: Don’t write a game you wouldn’t want to play.
That means the boundaries of ethical mechanics are going to be different for everyone. If you have contempt for gambling games and/or the people who play them, then I urge you not to design those games. Leave that to the people who love it, please. We will make better games.
This goes for any market, really. I believe I can design children’s games, because I genuinely enjoy children’s games. I don’t create them out of a sense of obligation, and I don’t write “down” to an audience without enjoying the product myself. On the other hand, I doubt I’ll ever work on a first-person shooter because I hate playing those games, they make me motion sick, and I really don’t need another realistic simulation of repeatedly killing everyone in the room. But someone loves making those games, and that’s fine for them.
As a freelancer and self-publisher, I’ve had the luxury to think like this for most of my career, to work only on those projects that pass this test. I realize that some folks are working under different circumstances and may have qualms about being asked to create games (or products of any kind) that aren’t the product they would use. You have to make that decision yourself, because now we’re talking about the ethics of following orders to keep your job, and that question is also way beyond the scope of this paper.
So let’s look at our questions and see if we have any answers.
Is there such a thing as an ethical payout?
In a gambling game, I think so. It’s any payout that matches the promise of the rules. A game is acceptable (ethical) if it truthfully describes how it works, so that a player has the tools to decide whether to play. If the game is deceptive, no matter what the return, it is no longer fair.
Does the quality of the experience increase with the stakes, or is it arbitrary?
Gambling games are informed as much by the context in which they are played as their mechanics and the stakes involved, and different players like and dislike risks of all levels, including many players who strongly prefer a risk of zero. So, different answers for different players.
And the underlying question: How do I know if my game is ethical / fair?
Here’s the first test. Would you play it yourself?
There, I guess I could have just said that.
Afterthoughts
There are many other questions of ethics in game design which do not relate to gambling mechanics or real money transactions. Games that reinforce bad behaviors or teach bad facts can also have a negative impact outside their magic circle. However, for the most part, I think games that allow players to exhibit “bad” behavior within the circle can be harmless, informative, and cathartic. Or else I wouldn’t make so many games where you play the villain.
Encouraging bad behavior is fine in games, as long as players understand the basic rules of abstraction, namely that what goes on inside the circle needs to stay there. Las Vegas co-opted this notion and flipped it on its head with their famous slogan, “What happens here, stays here,” which basically means exactly the opposite of what we’re talking about. I mean, cool bro, so then can I have my money back? Or is the money the part that stays there?
I will also mention that while I will gladly defend the validity of gambling games in the abstract, I make no apologies for the sociopathic behavior of every decision maker in the casino industry. Outside of politics, prisons, and the Mos Eisley spaceport, there is no more wretched hive of scum and villainy. So… losing money to those people does sometimes lower the entertainment value of the games. But winning it back also feels that much sweeter.