Bluffing Games

Bluffing Games Are All About “Why?”

This article is about bluffing games: what they are, why they work, and how to make them better, with a few examples. 

What is a “Bluffing” game? 

I use the term “bluffing game” to describe games with two critical components: deception and deduction. Strictly speaking, they should probably be called “bluffing and deduction games,” but that’s too wordy, and there is already a related (sub?) category called “social deduction games.” So here we are.

In a bluffing game, each player has a secret, and if that secret were known, their opponents would have an advantage. This secret could be an identity, a goal, or just a secret hand of cards that might be good or bad. Players must take actions that move them toward a goal, so players can get clues about each other’s secrets. Games with secrets but no actions don’t really have opportunities for deduction, so they are basically guessing games.

Poker may be the best-known example of a bluffing game. It’s full of secrets and lies. If you can guess what your opponent holds, you can make better decisions. In turn, you must conceal your own secrets, to deprive your opponent of the same advantage.

Other well-known bluffing games include liar’s dice, the dictionary game (Balderdash), Mafia/Werewolf, Stratego, Cockroach Poker, and Coup. I’ve designed a few bluffing games myself, including three that I’ll discuss below: Regent, Cursed Hand, and Pirate’s Bluff. I’ll also describe my house rule for Cockroach Poker that speaks to the core issue of “why?”

Bluffing games are not about lying for its own sake, or making promises you don’t intend to keep. Those games have deception but not deduction; I need to understand, within the context of the game, why you did what you did, and if that reason is just “because I’m mean,” then that’s an entirely different kind of game.

On the other hand, a game that’s all deduction and no deception is more like a multi-player puzzle. The key dynamic in a true bluffing game is not just that there is something to figure out (like solving a crossword), but specifically that each player’s secret is another player’s puzzle.

And finally, a game that has secrets but no way to tell them (no actions that reveal strategy) is a guessing game, not a bluffing game. One round of rock-paper-scissors isn’t a bluffing game, but a sequence of ten rounds can be.

Constructing a Bluffing Game

Good bluffing games are simultaneously short on rules and long on choices. Within a simple framework, players should have freedom to make meaningfully different decisions, even those that are tactically wrong, because this freedom allows them to communicate information (and misinformation) about their secrets.

Simple rules are always good, of course. But bluffing games want to be at the simplest end of the spectrum. Players need clear, tight boundaries for their decisions and their deductions. If a game has too many rules or possibilities, then a player’s actions will be harder to contextualize, and the central challenge of deducing their secret gets quickly muddled. 

Bluffing games don’t really “click” until players can grasp the structure of the game, so the first few rounds will always feel random. The more rules, the longer this random period will be. More complexity leads to harder decisions, which means less useful information in every action. If a decision is complicated or too close, then I can learn very little about your secrets, no matter what you decide.

I’ve seen prototype bluffing games crack under the weight of too many rules. Players can be too busy figuring out their own situation to really worry about deducing anything about each other. So, when building your bluffing game, first ask “is this system simple and tight enough for players to make educated guesses about each other’s motivations?”

Ideally, even if the rules are simple, the permutations of those rules, and even the different reasons why a player might do the same action, should be interesting. This means that within that simple rules set, you need to keep the players’ options, and the variability of the game state, as open as possible. You specifically want players to be able to make poor decisions on purpose, in defense of their secrets, without breaking the rules or destroying their chances of winning. You shouldn’t make bad decisions illegal, but you also shouldn’t render them strategically infeasible. (This is different from the best practice in designing most types of strategy game.)

“Too few options” may sound like a straw man, but it’s real. I recently tested a new prototype bluffing game in which players had secrets, but the rules only allowed them to do very specific things. If I’m trying to guess someone’s secret, and everything they do is their only legal option, I can reverse engineer their secrets by what they were allowed to do.

Every action should be an opportunity to communicate, and should have the effect of making an opponent wonder “why?” Every forced action (a decision point where the player doesn’t really decide anything) is a missed opportunity. Games that allow fewer free choices (for example, Hearts) have a strong deduction element, but not necessarily a strong deception element. Hearts is great, but because of the limits on choices I would not necessarily describe it as a “bluffing” game, because there is rarely a strategic advantage to playing the wrong card.

When building your bluffing game, ask “when I make a decision, what is my range of freedom, and what information does that convey? Can I make the “wrong choice” as an act of concealment, and does that have the potential to benefit me even more than making the right choice?”

Consider the typical bluff in Hollywood poker, where a player makes a huge bet with a terrible hand, This passes the “bad decision” filter: Can a player benefit from this often enough to compensate for the risk? Is that better than the value of simply playing correctly and folding? In poker, this answer is often “yes.”

In reality, this pure bluff rarely happens, but it’s still a reasonable model. More realistic bluffs are made with cards yet to come, when the player is “semi-bluffing,” and has a chance to catch a better hand if the enemy calls. This is still an action contrary to the value of one’s hand, because it can trick an opponent into making a mistake.

The real deduction in poker is more complex than a single hand: poker players pay attention to their opponents’ betting patterns, play histories, nonverbal tells, and even physical appearance and style of dress. Everything you do near a poker table communicates something about how you play.

Example Game: Regent

Regent is a two-player Pairs variant. It takes the simple core mechanic of “don’t get a pair” and turns it into a quick bluffing game for two players. 

The Pairs Deck contains the numbers 1 through 10, with each number N being represented that many times. This means there is one 1 in the deck, two 2s, three 3s, and so on, up to ten 10s. 

In Regent, each player’s secret is “do I have a pair?” You are trying to catch your opponent with a pair (two cards of the same rank) before being caught with one yourself. You are also simply trying to avoid getting a pair in your hand. Whether you are paired or not, there is value in convincing your opponent that the opposite is true. You can do this through your decisions, but also with table talk and nonverbal clues.

Each player starts with a hand of three cards: two down and one up. Score is kept with coins or chips.

Regent gives players a choice of five actions on their turn. Three of these actions continue the hand (usually), while the other two always end it. Generally speaking, all of these options are available all the time.

A player may:

• Draw: Take the top card of the deck into your hand, face down.

• Gift: Play a card from your hand face-up into your opponent’s hand. You are not allowed to give them a faceup Pair.

• Attack: Play the top card of the deck into your opponent’s hand, face up. If this gives them a pair, the hand is over and you lose. The penalty is the rank of the pair, for example 5 coins for a pair of 5s.

• Call: See your opponent’s down cards. If their hand contains a pair, you win. If not, you lose. The penalty is either the rank of the pair (if you are right) or their highest card (if you are wrong).

• Fold: Withdraw and pay a penalty equal to the lowest upcard (between both hands). 

The winner of each hand pays a penalty to the loser, and the game continues for any number of hands. The deck is not reshuffled after every hand, but only when it has been exhausted. Five cards are burned off each fresh deck so that the deck is never fully countable.

Short Rules: This is a simple list of options, though still more than the basic Pairs game. The possible actions can lead to situations where both players have many cards. Sometimes mostly face down, sometimes mostly face up.

Many Options: Most options on the list are available most of the time, unless the cards don’t allow it. For example, you can’t give a gift if you have no facedown cards, or if all of your facedown cards match your opponent’s upcards. The limitations are also informative: because you can’t gift a card that creates a pair, not playing a gift can send the message that you are unable to play one. With the added wrinkle of not shuffling the deck, there is a lot of information to track for players who are willing.

Game Objectives: A hand has one of two basic states: It either contains a pair, or it does not. If you knew your opponent’s secret, you could always make the right move, calling them immediately when they have a pair. There’s a bit more information as well; if you knew which face down cards your opponent holds, you could give them the right card to create a pair. And sometimes you can deduce what your opponent is likely to hold based on what they do, as well as the basic odds of the deck (high cards are more common than low cards).

Actions convey information. When you draw a card, you are gaining more secrets of your own, and possibly ruining a hand that was safe. So if you draw without fear, you are implying that your down cards are low enough not to be much of a risk. Or you might be implying that you have no other choices, because your down cards are not playable. Often you are sending a mix of signals with the same action. 

If you play a tiny card on your opponent, it makes them wonder why. Did you really have to break up a pair of 2s? Do you think you’re giving me a pair of 2’s? Or are you putting that low card out there because you want the price of folding to go down? And if you want to fold cheaply, does that mean I should call you?

In the process of choosing the available actions for Regent, we looked for that smallest set of choices (and accompanying rules) that would give the highest value in terms of guessing, deducing, and keeping secrets. 

Example Game: Cursed Hand

Cursed Hand also uses a Pairs deck. It is designed for a larger group, up to six players. The core mechanic is a single hand of cards which passes from player to player, always growing by one. As in Regent, the hand is “cursed” if it contains a pair. 

To start the round, shuffle and deal a hand of seven cards to each player (six cards if there are six players). Each player chooses one card to start their “Graveyard,” which is a set of faceup cards on the table in front of you. These are revealed simultaneously and the highest unique card will take the first turn.

The first player creates a “Ghost Hand” by drawing a card and then choosing any two cards from their hand. They pass the ghost hand to the player on their left, face down, and declare out loud the size of the hand: “two cards.”

The receiving player may either call, if they think the hand is cursed, or accept the hand, if they think it is not. If they receive the ghost hand, they must look at it, add one card to it from their hand, and pass it to the next player, who in turn may either call the hand or accept it, and so on. Each time the cards are passed, the passing player must declare out loud the number of cards.

When the hand is called, it is shown, and the contest is between the caller and the player who passed it along. If the hand is cursed, the passing player adds one of each of the paired cards into their Graveyard. If the hand is not cursed, the calling player takes the highest card, also into their Graveyard.

Note: The Ace is wild and always matches the highest card in the Ghost Hand. It is never taken as a penalty (you take the natural card instead).

The game has just one loser: the first player who accumulates a pair in their Graveyard. For a longer game, you can play until one player has lost three times. 

Short Rules: Again, this game is quite simple, and it uses a fairly common Pairs mechanic of “not getting a pair.” This is interesting in the Pairs deck because each rank has a different number of cards, and thus a different chance of being paired.

Many Options: Among the limited choices in the game, players have free choice of what card to add from their hand. This is important because, usually, players can decide not only whether to create a pair in the ghost hand, but also (especially in cases of the deadliest and most common cards), exactly which card to pair. This is important because of the losing condition, getting a pair in your Graveyard. I’m more likely to pass a pair of cards that would kill you, but not myself.

From the contents of each player’s graveyard, it’s possible to guess how they might play the hand. You would never pass me a pair of 10s, because you already have a 10, and that penalty would kill you. On the other hand, you might do exactly that, thinking I would never expect it.

One mechanic I’m especially fond of is the requirement that the passer must say aloud how many cards are in the ghost hand. For one thing, it keeps everyone engaged and updated on the size of the hand. But more important, this tiny speech is an opportunity for the next players to deduce, from verbal or nonverbal clues, whether the hand is cursed. If the players were allowed to act silently, there would be less opportunity for these kinds of tells.

Example Game: Pirate’s Bluff

Pirate’s Bluff is a new card game similar to liar’s dice. In both games, the object is to guess how many of each possible rank are in play in everyone’s hand, or to deduce when someone has guessed too high. The challenge is that it happens in real-time, so acting fast if often more important than thinking too much!

The deck contains eight copies each of the ranks 1 through 8, as well as eight wild cards ranked 0. About half of the deck (or less) is dealt out, so at least half the cards remain out of play. Wild cards can become any rank, but only if they are in a hand with natural cards of the same rank.

To begin, players ante one chip into the pot. They start with a hand of seven cards, and then play one of them face up. The lowest of these cards will make the first bid. Ties are broken by unique letters on each card. 

A “bid” is statement of a number of cards of a specific rank, for example “five 2s.” This bid means that among all the hands, including the face up cards, there are at least five cards of rank 2. (Wild cards will act as 2s if this is called, but only if they share a hand with natural 2s.)

After the opening bid, anyone can jump in with a higher bid, or call the last bid. A “higher bid” is defined either as (a) the same quantity of cards, but of a higher rank, such as five 4s, or (b) a larger number of any rank, such as six 1s. 

If a player says “call,” this ends the round, and players show their hand to determine whether the last bid was right or wrong. The winner takes the pot, as well as a penalty from the loser equal to the number of cards in the called bid. For example, if the bid was “seven 2s” but there are only six 2s, then the caller was right, the bidder was wrong, and the penalty is seven coins. 

Short Rules: Pirate’s Bluff is both a bluffing game and a speed game, both of which have to be incredibly simple to function. Players have just three choices: raise the bid, call the bid, or stay silent. Because of the ante, sitting quietly can be expensive in the long run, so players do have some incentive to get into the action.

Many Options: Within the short rules, players still have reasonable freedom about when to act and what to say. Specifically, players are allowed (and even encouraged) to make bids that they think are wrong, with the hope of drawing others into raising those bids even higher. A player with no 4s in her hand might make a bid of seven 4s, and then snap-call any player who bids eight 4s. As in liar’s dice, a deceptive bid can be powerful as long as no one sniffs it out!

Adding a “Why”: Cockroach Poker

Cockroach Poker (Drei Magier Spiele, 2004) is a great bluffing game, but it stumbles a bit off the starting line. I want to describe a house rule that helps cut to the chase, related to the core question of “why?”

The Cockroach Poker deck is similar to the Pirate’s Bluff deck: it has 8 cards in each of 8 kinds of insect / creepy thing (but without the ninth rank of wild cards).

To start a hand, deal out the entire deck. The action bounces around the table. At the start of a round, the active player chooses a card from their hand and passes it face down to any other player, declaring aloud the identity of the card. For example, “this is a stink bug.” (This can be a lie or the truth.) The target player now has three options: Yes, No, and Pass. 

1. Yes: Target declares that the passer is telling the truth. The card is revealed.

2. No: Target declares that the passer is lying. The card is revealed.

   1. In either case, the card is placed face up in front of the loser of the showdown, and that player starts the next round. (If the target guesses right about the passer’s declaration, true or false, the target wins. If not, the passer wins.)

3. Pass: Target player makes no judgement. They take the card, look at it, and then pass it to another player, making a similar declaration.  They are not required to make the same declaration as the original passer. The card can’t go to any player who has already seen it. If the card has visited everybody, the last player doesn’t have the option to pass. 

The game has one loser: either the player whose hand is empty first, or the player who accumulates four of the same kind of card face up. 

Overall, Cockroach poker works well, and satisfies the short rules and open choices conditions. A starting player may pass any card to any opponent, declaring any type of card, whether lying or not. The next player isn’t locked to declaring the same type, and so on. 

Where the game needs improvement, in my opinion, is in the opening rounds. At the start of the hand, there is no practical difference between the card types, except the isolated information each player sees in their own hand. Later in the game, if I say I’m passing you a cockroach, and you already have three of them, this is meaningful. But at the start, nothing matters. So I have trouble grasping the “why.”

House Rule: I start each hand of Cockroach Poker by dealing everyone one card faceup. This gets us past the first awkward steps, and makes each card meaningfully different from the start. Now when a player says he’s passing a Stink Bug, I can look at the cards on the table, especially in front of both of us, and form a more interesting picture of what that declaration means.

If this were my own game, I’d also figure out a way to use this opening deal to randomly choose a starting player, with something like a serial number on each card. But I think that’s asking more than the published Cockroach Poker deck can do. (On the other hand, the Pirate’s Bluff deck would be perfect…)

In Conclusion

If you want to design a bluffing game, or make your off-the-shelf bluffing game better, think about these ideas. Keep the rules simple, but allow lots of freedom within that simple framework. So much freedom that a player can “play wrong” and still win. Oh, and be sure you give them something to do.

The heart of any bluffing game is secrets: keeping your own, and learning those of your opponents. If you knew your opponent’s secrets you would always know how to win.

Whenever a player takes an action, you should be able to ask “Why did they do that? What does that tell me about their secrets?” If clues like this help you solve your opponent’s puzzles, then you’re playing a bluffing game.

References:

• Regent: from the Muses Pairs Deck, also found in the Pairs Companion Book.

• Cursed Hand: First published in Game Trade Magazine, Oct. 2018, also at cheapass.com.

• Pirate’s Bluff: Available from us right here!

• Cockroach Poker: Designed by Jacques Zeimet, Published by Drei Magier Spiele