Designing Traditional Card Decks
I’m embarking on a new game project called Fairmarket, which will be an artifact from the Dew Point universe. I want to use a traditional-style deck, meaning something that’s as ubiquitous and versatile as a poker deck, but unique to that world.
The modern 52-card deck can play thousands of different games, from poker to gin rummy to solitaire, some with hundreds of variations. What makes this deck so powerful, and how would a game designer create an equally useful deck from scratch?
This article explains my thoughts on what makes a good multi-game card deck. I’ll talk specifically about how these considerations affect the design of Fairmarket. I’ll also talk about other decks I’ve created in this space, and whether they hit the right marks.
To begin, let’s talk about the poker deck.
Section 1: Dissecting the Poker Deck
The modern poker deck has evolved into its current form over more than a thousand years, consolidating from decks of many different purposes, sizes, and compositions. I won’t retell that history here, but basically the deck we use today could be regarded as the best version of all the ones that preceded it. Or at least, it’s the one that survived. For most modern people, it is literally the only conceivable deck of cards.
I believe that nothing about this deck can be taken for granted, from the aesthetics of suit colors and corner indices, to the distinction between court and number cards. There are so many other ways to create a deck of cards, I think it’s worth examining in detail why this version prevails.
Card Size
The most common sizes for playing cards in the US are “poker” (2.5” x 3.5”) and “bridge” (2.25” x 3.5”). Both sizes are popular for different applications. Wider cards are more durable and easier to read. Narrower cards are easier to handle.
Standard European card sizes are essentially the same, being rounded down to the nearest millimeter from the imperial dimensions above. Many other sizes and variations exist. But this basic range seems best, simply because it is the easiest to shuffle and deal, fits well in the hand, fits well on a table, etc.
For Fairmarket, I’m choosing poker-sized cards, because I know that’s a standard size at DriveThru Cards, the factory where I will print them. (Some design decisions are easy.)
Deck Size
How many cards are in the deck? As with card size, this answer seems to involve versatility and ease of shuffling. I’ve worked with decks as small as 36 cards and as large as 110, but the range for easy shuffling seems to fall between 48 and 60 cards.
The exact number of cards will result from decisions we make later, but 48-60 seems like a reasonable comfort zone. A smaller or larger deck is not out of the question, but I think shuffling and dealing becomes difficult above about 72 cards (a tarot deck is 78), and I worry that the versatility of the deck is compromised if we go below 36 (with fewer cards, not enough different games are possible). I could easily be wrong on both counts.
For Fairmarket I once again have my printer’s strengths as a consideration. DriveThru Cards can make a deck with literally any number of cards, but they also make tuck boxes for poker decks of 54, 72, 90, and 120 cards, or 80 tarot-size cards. So, while these box sizes are not a huge consideration for game design, I’ll still keep them in mind.
Card Design
The poker deck has the following information on the face of each card: A rank and a suit. This is indicated by an index in the upper-left and lower-right corners. In some decks, primarily in Europe, the index appears in all four corners, so the cards can be fanned in either direction.
In the middle of the cards is artwork that reinforces the same information. On the number cards, there is one pip (one suit icon) for each value in the rank of the card. On the face cards, there is a unique character portrait, such as the Queen of Hearts.
Note that all the information defining the card appears on the corner index. Technically the rest of the graphics are redundant, including the index in the opposite corner. However, counter-indexing (putting the information in multiple corners) makes it easier for players to read a fan of cards.
Why do the pips still appear in the center? Before indices were invented, the pips were generally the only art on the card; the five of hearts just had five hearts on it. But even today, the center art is still extremely useful, because it’s easier to read at a distance.
Why are the high cards represented by pictures instead of large numbers? The first reason is likely that above ten pips, it becomes harder to count the dots than to recognize a portrait. The images for Jack, Queen, and King may be easier to distinguish than counting eleven, twelve, or thirteen pips. But more importantly, the different design of face cards creates a wholly different category of card. More on that later.
Why is there not more information on the cards, such as game rules? Because, in general, the more information you put on a card, the fewer games you can play with it. The 8-ball isn’t marked with “sink this ball last” because that’s not true for most pool games. The Ten of Diamonds isn’t marked “Big Cassino” because that’s not usually true or relevant. Also, traditional-style card games tend to exist in civilizations with lower literacy rates. I try to include only language-neutral components in a traditional-style games. So their simplicity makes them versatile, but also makes them more accessible.
All the card backs are identical, and designed to prevent the cards from being identified when face-down. This is a nearly universal trait of card games, because cards are more useful when they can be shuffled and dealt in secret. It’s not mandatory, and some modern games use a different back on every card, but that decision drastically limits the utility of the cards.
Since Fairmarket will probably have ranks and suits, I won’t stray too far from these general design principles. Examples from my first draft of the deck are in the graphic above.
Deck Composition
Each card in the poker deck is identified by a unique combination of rank and a suit.
Rank determines each card’s relationship to the others, with number cards ranked in numerical order, followed by the face cards. The ranks of Jack, Queen, and King must be memorized, but the order is fairly easy. The high Ace is somewhat counterintuitive at first, but it’s helped somewhat by the use of “Ace” rather than “one” as the name of the card, placing it somewhere among the letters rather than the numbers. (And Aces are not always high.)
The four suits do not have an obvious order. Instead, their order may be defined by game rules, or they can be treated as equivalent. This gives suits a different kind of utility than just another set of ranks. Combining 13 ranks and four suits gives us 52 unique cards in a comprehensible sequence.
Why is the poker deck 4 x 13, instead of 4 x 12, or 6 x 9, or 5 x 11, or some other more interesting arrangement of cards? Why are there suits at all? Why isn’t the deck just the numbers 1 through 52, which could arguably be made to do the same job?
I believe the main reason is what I’ll call “chunking:” different ways that the deck can be divided into blocks, so that certain sets of cards can be assigned different behaviors, depending on the needs of the game.
To play any card game intelligently, players must understand what’s in the deck. The regular arrangement of the poker deck, along with its clearly defined subsets (such as ranks, suits, face cards, and so on) makes the deck easy to understand. Imagine trying to make decisions in a game where spades has two more cards than the other suits, the Baron and the Ghost, and the ranks 5 through 8 were high in hearts only. These kinds of wrinkles might be exciting for a specific game, but they would be annoying as a feature of a core deck.
From the player’s perspective, this is a question of triage: How quickly can I determine whether a card is “right” or “wrong” for a certain job? The faster I can answer this, the faster I can play the game. This is a function of the graphic design as well as the recipe of the deck. And while individual determinations may seem negligibly fast, they can add up to a huge cognitive load when performed thousands of times in a game.
Every time you look at a card, you process what it is. You may not notice this with a poker deck, because you’re used to it. But remember how slow it feels the first time you play a game you’ve never seen. If a card is harder to read, or if information you need is buried, or if the deck is weirdly constructed, this will slow the game down. Imperceptibly in the short term, but fatally overall.
From the designer’s perspective, clear subdivisions help to create new games. How many ways can I slice this deck into chunks of the right size? How do I designate a particular fraction of the deck to behave in a specific way? If the deck is just the numbers 1 through 52, can I say “the multiples of four” are in their own category (for example, trump cards)? Is that really the same as just saying that trump is spades? Functionally, yes, but each time I check that a number is a multiple of four, I spend some cycles on cognitive load. This seems hard enough, but if I wanted trump to be hearts, I’d have to designate all those cards that are one rank less than a multiple of four. Possible, yes. Practical, no.
The more different-sized chunks I can define, the better. This means that if there are several ways to create the same size chunk, some of them might be redundant. For example, I briefly experimented with a 49-card deck representing all possible rolls of two 7-sided dice. Each card had a black die and a white die, each of which could have a value between 0 and 6. One shortcoming of this deck was that the black and white axes were identical. There was less variability in the 7 x 7 chunks than you would find in a non-square deck (e.g. 4 x 13, 5 x 11, 3 x 17, etc.).
What are the chunks in the poker deck? For a chunk of four cards, I can designate a single rank, such as “eights.” For a chunk of two cards, I can designate a color and a rank, such as “red eights,” and for a chunk of one, I can designate any single card, such as “eight of diamonds.” For a chunk of thirteen, I can choose a full suit. For exactly half the deck, I can choose either the red or black cards. I can also cut the deck at any multiple of thirteen by designating a break at a certain rank, such as “cards nine and above.”
That last example is why I think face cards are quietly doing an important job: They are their own obvious chunk of 12 cards. In theory, I could also choose a set of 12 cards by nominating the “red primes,” but obviously I would only do that to make my players suffer, or at least make them argue about whether Jack and King are prime.
Note that dividing by color can do more than just select a block of 26 cards. Splitting any subset by color lets you slice any other chunk in half, such as choosing the 6-card block of “red face cards.” This is part of why I think the two-color poker deck is still the most useful version, despite the advent of four-color decks. If each suit has its own color, the four suits are certainly more distinct, making it easier to identify flushes and harder to accidentally mis-play in trump games. But without “red and black” I can no longer chop other subsets in half.
This all sounds academic but believe me, chunking is a hugely valuable tool in game design, and if your chunks are easy to describe (as opposed to “those cards ranked one less than a multiple of four”) then your players will thank you for it, usually in the subtle form of not noticing. It’s comparable to the difference between burying your UI in layers of relevance, instead of putting all options on the home screen (or, if you will, Apple versus Microsoft packaging).
For example, blackjack says that all the face cards are worth ten points. This is invisibly clean, and you might not even have noticed the value of that chunk in this context. But imagine trying to implement the same rule if those cards were labeled 11, 12, and 13.
Section 2: Other Games
Let’s talk about some of the traditional-style card decks I’ve worked with, and how they compare on the aspects above. I doubt if I’ve made anything better than the poker deck, but at least I’ve found a few interesting new corners of the design space.
To be clear, I know that many of these decks have been made before. Adding more suits to a poker deck isn’t exactly rocket science. But I think the details are still interesting. Even the smallest differences can have dramatic effects.
Renfield (Cheapass Games, 1999)
Renfield is a combination trick-taking and gambling game about gravediggers eating bugs. The deck is a three-suited deck designed specifically for this game. It has three suits numbered 0 through 17, for a total of 54 cards.
Each card also has a unique name, as well as additional game information, including a point value and a dollar cost. These values are the same across each suit, but they don’t directly correlate to the ranks.
Renfield is a game with many variations, similar to poker. But the deck is somewhat limited in scope, because the cards include so much game-specific information. It would be hard to modify or ignore the extra marks if you wanted to make a meaningfully different game.
Also on the downside, there isn’t anything to distinguish sub-blocks within each suit (for example, numbers vs. face cards). It’s just a long run from 1 to 17, with the 0s being slightly special because they are not normally shuffled into the deck. Seventeen cards is a big block, and it needs some kind of partition.
We’ll circle back to this idea later, because I’m also using a three-suited deck for Fairmarket.
Fortune’s Tower (Microsoft / Lionhead Studios 2008)
Fortune’s Tower was part of a suite of gambling games that I invented for Fable II, by Lionhead Studios, and was also in the stand-alone product Fable II Pub Games, from Carbonated Games / Microsoft.
The Fortune’s Tower deck contains four “knight” cards, along with several identical blocks of cards ranked 1 through 7. The number of blocks is different from game to game, depending on the house rules: there can be as few as eight or as many as ten number blocks, though there are always only four knights. Since the knights are good for the player, the worst decks are therefore those with more number cards.
While Fortune’s Tower feels reasonably “traditional,” and is certainly at home in a medieval fantasy world, the deck was designed for this game alone, and might not have a lot of other applications. It could be considered a modification of a traditional 4x8 deck, in which there are four suits including numbers 1 through 7 and four knights, for a total of 32 cards. The game was designed and tested with poker cards, and it plays just fine in a deck with suits.
No other games have been written with this deck, so its versatility remains to be seen. I wasn’t as interested in high-versatility decks in those days. But without suits, this deck can only be broken down into ranks. I’d probably change it to a suit-based system if I needed to develop many more games with the Fortune’s Tower deck.
Quicksilver (Unpublished, 2012)
A precursor to Pairs (below), Quicksilver was a 62-card deck of tarot-style cards, intended as the core component of a collectible game with a medieval theme. The core mechanic was based on a two-player press-your-luck game, which Paul Peterson and I originally roughed out with a poker deck.
The Quicksilver deck has six suits of different sizes, containing 16, 14, 12, 10, 8, and 2 cards. Each card has a unique name and a serial number showing its exact position in the deck. The 16 “Commons” are numbered 1 to 16, then the 14 “Merchants” are 17 to 30, and so on up to the 8 “Clergy,” which are 53 to 60. The last two cards have no suit, and are both numbered 0.
While I liked the idea of blocks of different sizes, players took a while to understand the sizes of each block. Theoretically, if this were the only deck in your universe, you’d know the suits. But that’s not practical in this modern world. So while I could chunk this deck in interesting ways, players didn’t easily learn the sizes of the chunks, which made it harder to make good decisions or understand what was happening.
As with Renfield, the Quicksilver cards contain a lot of game-specific information, including a second set of suit symbols, a coin value, and a death icon. This makes the deck less generic, probably suitable only for this game. I’d likely strip off all this information, possibly replacing it with an indication of the suit size, if I redesigned this deck as a traditional deck. But it was designed for the hobby game space, which explains the details (more on that at the end).
Quicksilver never made it into production, but it did lead to a breakthrough with Pairs.
Pairs (Cheapass Games, 2014)
From Quicksilver, Paul Peterson and I took the idea of different-sized chunks to its logical conclusion in Pairs, revisiting the “triangular” deck design that had been used in games like The Great Dalmuti (Wizards of the Coast, 1995) and my own 12 Days (Gamesmith / Calliope, 2011).
The triangular deck gets its name from the distribution of ranks, not the shape of the cards. The Pairs deck has ten ranks, and each rank is marked with the number of cards in that rank. The deck therefore has 1 x 1, 2 x 2, 3 x 3, and so on up to 10 x 10, for a total of 55 cards. With this construction, understanding the size of each chunk is easy. If the card is an N, there are N of that card.
Because each rank is a different-sized block of cards, a designer can designate a chunk of any size just by picking the correct rank. Do you want a block of six cards? Choose the 6s. Do you want a block of 10 cards? Use the 10s. Or if you prefer, choose the 7s and 3s, or all cards of rank 4 or lower.
The triangular deck turns out to be exceptionally versatile, and it’s uniquely suited to a family of games you can’t play with a poker deck. But it can’t do everything that a poker deck can do, in part because most individual cards are not unique. Every 10 is the same, so if a tie must be broken between two of these cards, the game must use alternate tiebreakers, or just be satisfied with ties. So the deck loses value in sequencing by gaining simplicity in chunking.
Trying to address this problem, Paul and I also invented a more versatile deck at the same time, in which every card was unique. We called it a “pyramid” deck, which basically meant a triangular deck in three dimensions. This deck had three sets of markings on each card, which corresponded to three different axes of numbering, all running from 1 to 10. It was still 55 cards, but the three numbering systems were superimposed on the same triangle of cards
We tried making games with this deck, but most of them ended up focusing on only one or two of the three numbering systems, rarely all three. The deck had similar problems to the 7x7 deck mentioned above, in that all three chunking systems were the same size, so they often felt redundant. In addition, any marking we ignored became a distraction. In the end, this was just too much extra information crammed into one deck, and we decided that the single set of numbers was simply better.
And thus, the Pairs deck was born.
Willow (Cheapass Games, 2014)
Willow can be found in the Cheapass Games Poker Suite. Before that, it was part of a pitch for a gambling game in a fantasy universe. Willow uses a 36-card deck, which is three suits, ranked Ace through Queen.
The game has a structure similar to stud poker, using hands of only two cards. The value of a hand is based on the distance between the two cards, with pairs being the best, followed by cards that are the farthest apart.
Because the client didn’t buy the game, I’ve never done much with this deck, but I’m intrigued by some of the principles relating to a three-suited deck. I’m especially fond of the fact that there can’t be a tied pair, because there are only three cards of each rank. So some of what I learned from creating Willow might come in handy if I develop any poker-type games for the Fairmarket deck.
Rochi (Unpublished; open beta 2017)
Rochi is a gambling game from the fantasy world of Sonia Lyris’ The Seer. Like the Quicksilver deck, it contains tarot-like cards in blocks of different sizes. Like the Pairs and Fairmarket decks, it’s generic enough to play many different games.
The Rochi deck has 72 cards in six blocks of different sizes. Rochi’s blocks are number ranks: 2, 3, 4, 5, 6, and 7, each of which contains twice as many cards as its rank: four cards of rank 2, six cards of rank 3, and so on.. Each card also has a unique name, similar to a tarot card, and a designation of either “day” or “night.”
Each block is labeled with a number that is half its size, which seems strange at first, but supports the game’s core mechanic, called “cutting.” This term means dealing cards until half of a block has been revealed. For example, dealing 2 of the 2s, or 3 of the 3s, etc. So labeling the cards with the “cut” value rather than the “total” value makes the counting much easier.
In-world, this mechanic dates back through several dice and card games that precede the modern Rochi deck. One would assume that a person who grew up with this deck wouldn’t need the numbers at all, being intimately familiar with the sizes of each block, but of course we’re creating this deck for real people, so the numbers on our version of the deck must be large and clear.
I have developed only a couple of games with the Rochi deck, but I expect it has the same basic properties as the Pairs deck, and many similar games can be made with it. It also has something the Pairs deck lacks, which is a distinct identity for each card, so that games can rely on a full sequence of cards. Even with all else being equal, two cards can be ranked by their name.
The Island Deck (Crab Fragment Labs, 2020)
The Island Deck is a six-suited deck with nine cards in each suit, numbered Ace through 7, King, and Queen. It was created for a specific game called Showboat, but it’s generic enough that it can certainly play many other games.
As described above, I think including face cards within each suit is a huge improvement over just using ranks 8 and 9, because it gives a clear breaking point to designate a different type of card.
Showboat was initially developed as a custom-deck game called “Stuck with the Check,” about skipping out on a restaurant bill. The client didn’t pick it up (sound familiar?), and shortly afterwards, I realized that the game could be altered to play with a simple six-suited deck.
In the original game, the deck was composed mainly of numbered cards in six categories of food. There was also a block of non-food cards called “stories” that did practically the same job (and appeared in the same proportion) as the face cards in the Island Deck.
Since I prefer traditional decks to specialty decks, I developed the Island Deck so I could publish Showboat at Crab Fragment Labs. I haven’t had the chance to develop many games with it, because we’re all in quarantine now.
I worry that the 9 x 6 structure of the Island Deck might be less dynamic than decks with one or more prime axes, like 5 x 11, 4 x 13, and 3 x 17. But I do think this arrangement of cards will be better for some applications. This deck will make some interesting poker hands, but will probably be tough for trick-taking games. I don’t yet know what else it will be good at, but I’m itching to find out.
Ultimately the Island Deck will probably excel at games in the same space as Showboat, which is fundamentally about playing cards in sequence. I’ll be working on more games as soon as we can gather in large groups again.
Section 3: Inventing Fairmarket
With all these considerations and more than 20 years of game creation, I’m finally returning to the three-suited deck for my newest game, Fairmarket.
As with Showboat, the core mechanic for Fairmarket preceded the development of the deck. Fairmarket is based on a game called Merchant’s Row, which I created with some friends and a poker deck in the spring of 2019.
The structure of the game is something of a cross between a programming game, a trick-taking game, and a bidding game, and it uses a deck of cards that could be almost any linear sequence, including a deck of the numbers 1 through 52. You could play Merchant’s Row with the Island Deck, the Renfield deck, the poker deck, or really any deck of unique, serialized cards. But the cards need to be individuals, so you can’t play it with something like the Pairs deck.
In the original Merchant’s Row, suits were only used as a tiebreaker, so the Ace of Spades was strictly the highest card, and the Deuce of Clubs was strictly the lowest. High cards are better than low cards, and the game is definitely marked by “good” and “bad” hands.
We included a rule that makes low cards marginally useful, but of course this just passed the “worst card” torch to the middle cards, and bad hands are still bad hands. When re-opening this project, I wondered if I could make a version where “bad hands” were not always so bad.
Usually when I run into a design block, I look to the story for help. Fairmarket is a purely abstract game, but the Dew Point setting gave me some clues about how to approach the problem. People of that world see things in chunks of three, in the same way that we see the world in “black and white.” This means that there is a left, right, and center; an up, down, and middle; a sea, land, and sky; and so on. There are three main mythological archetypes, corresponding to birds, beasts, and trees. And three gender categories: strong masculine, strong feminine, and everyone else.
From this starting point, I was excited to take another look at the three-suited deck. Last month I played a few hands of Merchant’s Row using a Renfield deck, almost immediately adding the rule of trump to make different suits high on different rounds. That worked well enough, and now I’m designing a new three-suited deck specifically for Dew Point and Fairmarket.
In order to create manageable, versatile chunks, I am creating three blocks within each suit. The first block is just one rank of cards, like the zeroes in the Renfield deck. They are labeled “O” for Opas, essentially the “gods.” Like Aces, these cards can rank high or low. (The “O” also has the benefit of looking like a zero to modern eyes, though I’m not sure Dew Point has a concept of zero, and even if they did, they wouldn’t start counting with it.)
The second block is the numerals 1 through 12, and the third block is five face cards labeled B, C, D, E, and K, for world-specific (and non-gendered) variations of the words for Baron, Count, Duke, Earl, and King. This set of initials has the advantage of ranking in alphabetical order, so new players can understand the order of the five face cards.
The three suits are based on archetypes for masculine (arrowheads), feminine (owls), and neutral (trees). These are loosely based on the three major political camps, which are also entwined with the prevailing origin myth, all of which is so early in development that I can’t point to anything to read.
And that’s where I am right now. It’s actually very early in the life cycle of Fairmarket; I’ve only been on this track for about three weeks. But I wanted to talk about what I’ve learned about traditional decks, and how I’m using that knowledge to create this new deck. I’ll post the deck and the alpha rules shortly, as soon as I get a chance to play with the first proof decks.
So how is the Fairmarket deck measured by the standards above? This 54-card deck splits easily into thirds (by suit), into any block of 5 or 12 (faces and numerals), into any block of 3 (by rank), or a block of 15 (all face cards). Each card is unique and falls in a sequence. Interestingly, there’s no easy way to split the deck in half, short of choosing “cards 10 and higher,” but I like the idea of forcing players to think in threes when they’re playing in the Dew Point universe.
I worry that the spread of numbers between 1 and 12 is a really big slice, and I might want to think about ways to break it up even more. It might naturally break at 10, but I’m not sure. I don’t know if my theory about the poker deck is correct, that the 10, 11, and 12 will look too similar. So I’ve printed some decks with the graphics shown above, and I’ll be testing them as I refine the rules to the core game.
Without experimenting, I can’t really know.
In Conclusion
The specifics of the poker deck make it ideal for creating different games. There is little or no superfluous information, and the cards are marked in ways that players can easily break them into multiple subsets. Designers can slice up the deck into many smaller groups, to designate different chunks with different behaviors. The ranks have a clear sequence, and the suits do not, so these two divisions can accomplish different things.
To create a new deck that does the same job, it’s important to think about the various ways in which that deck can be subdivided. Are those chunks different and useful sizes? How will players determine what chunk each card belongs to? Are some cards identical, or can any two random cards be related to each other?
Breaking up a set of ranks into numbers and face cards can help further the goals of ranking and chunking cards, so the break point between numbers and non-numbered cards defines the sizes of those blocks.
If similar blocks are not the same size, how can players easily learn the difference? What other ways are there to subdivide a pack of cards into clear, useful chunks? These are the considerations that go into creating any nonstandard collection of cards.
At the end of this exercise, I have essentially described how to make a poker deck with three suits, by stretching the quantity of numerals and face cards in a pretty obvious way. But I still think this list of considerations will be useful as I fine-tune that deck. And I hope it helps you in thinking about designing traditional-style games of all kinds.
Afterthought: What’s Wrong with Specialization?
Why are some games played with a generic deck, while others require specific cards? Obviously these are just different types of game. But the distinction is really about commerce.
When I invent a new game with a poker deck, or any existing deck, I can’t make a lot of money selling it. That’s because the components already exist, and to play the game, someone only needs the rules. If I’m lucky, the game will become popular. But if I am trying to make money, tons of free play doesn’t matter. I need something I can sell, or another way to market this new game.
On the other hand, if I create a card game that requires a custom deck, I can manufacture and sell that deck. If the game is popular, I can make money. This principle is the driving force behind the hobby game market, and explains the types of games we find in stores: They are boxes of components that, basically, can play only one game.
Of course, there are plenty of products in the gray area (or “in the center” as they’d say in Dew Point). Hasbro markets games like “Canasta Caliente” which is just a double poker deck, marked with some extra information specific to canasta. UNO can be played as Crazy Eights with a standard double-deck. And plenty of hobby games are structured around a poker deck, with just enough changes to make them require a purchase. Personally, these kinds of games make me itch. But they also make money, sometimes a lot of it. (I'm looking at you, Lost Cities.)
For the most part, new games that use a traditional deck, or even multiple games with the same generic deck, are hard to sell, and are mostly found in noncommercial places like collections at the back of magazines, games woven into works of fiction, or on the websites of semi-retired game designers.
On the other hand, outside the marketplace of hobby games, a traditional-style deck clearly makes the most sense to a player. If I am a medieval courtier, or a character in a fantasy world, or a hiker right now, why would I buy (and carry) a deck that can play only one game, when there is another deck that can play a thousand different games? That one game must be pretty amazing.
Ultimately those specialized card games are a distinct form of entertainment, and shouldn’t be judged by the same criteria as traditional games. Magic: the Gathering is the epitome of a game with specialized cards (thousands of them), and that game does just fine. But you probably wouldn’t find a trading card game in a fantasy world, for all kinds of reasons.
In the end, though, I’m still fascinated by the challenges of creating a versatile and traditional-feeling card deck. I’m looking forward to discovering more about Fairmarket as I develop a few more games with the deck and find out what else it can do.
Update: March 7 2021
The discussion of the Fairmarket Deck above was from the very early stages of development. In the intervening weeks, I’ve worked some more on this deck, and come up with a new version.
The “block” of one rank (the three zeroes) turned out not to be very interesting. Honestly, any single rank is already a block of three, so there’s no point in describing the zeroes as anything special.
I wanted to break the 18-number sequence into three more powerful blocks, and I also didn’t want to reinvent too many wheels. I thought about other ways to create clear sequences, but letters and numbers are really in a class by themselves. So for the middle block of cards I considered using double-numbers: 11, 22, 33, etc, probably up to 66.
I refined that a little and settled on using “tens” for that block. So the numerals now start with Ace (one) through nine, and then there is a “ten” card with a different pip style, followed by 20, 30, 40, 50, and 60, and then three face cards: Page, Courtier, and Queen. As in the poker deck, the Ace can float to the top of the face cards.
These blocks can be divided in several ways: the Ace can move from the bottom to the top, and the 10 can be considered either the last of the sequential numbers, or the first of the 10s-block. So we have an 18-card block that ordinarily breaks into 9, 6, and 3 cards. But if you move the Ace to the top, that’s 8, 6, and 4, unless you attach the 10 to the low block, in which case it’s 9, 5, and 4. Lots of possibilities there.
With point values ranging up to 60, this deck is interesting and might open up some new design space. Some existing games use scoring that goes that high, although in most games the points are irrelevant, just the sequence of the cards. In a trick-taking game, the 10s block is just the “clearly higher” set of number cards.
Anyway, the game still isn’t nearly final, so I won’t claim that this is the greatest solution, but I’m curious to see where it goes next. I can imagine a lot of very good trick-taking games with this deck, if Renfield is any indicator.
The current incarnation of the Fairmarket Deck is nicknamed the “Knotts” deck. This deck uses the three suits of Flags, Cups, and Bells, and is a common deck along the South Coast of DeVere.