How to Design a Card Game
This article is a primer for card game design ahead of the 2024 Duck Sauce Game Jam where we'll be creating board/card games in addition to the usual videogames. It won't cover everything and it won't go as in depth as other guides out there, but hopefully it gives you enough to get started.
I'm going to be focusing on card battlers, think Magic the Gathering, YuGiOh etc, although a lot of what I say applies to other genres like deck builders. Also, I'll mostly be talking about the nuts and bolts of how these games work and not so much on things like theme and visual design.
Making Fun Cards
We'll start by going over what makes a set of cards fun to play with, as I think that'll help guide how the game is made. Sid Meier famously said "A game is a series of interesting choices", and that very much applies to card games. From picking which card to play from our hand, to picking which cards to put in our decks, or even which cards to draft prior to building our decks, a lot of what we do in card games is just picking the best card from a selection.
Let's start with a boring card game and explore ways we can make it interesting. Without worrying about win conditions or mechanics or anything like that, let's imagine a simple units-on-the-board card game. Right now the "set" of cards we have are simply the numbers 1-100. Higher number wins combat against lower numbers.
Right now the puzzle is pretty easy to solve, you just choose the highest number, and people will just put the highest numbers in their deck. So we need to complicate things a bit. Let's say we add a resource system so that you get a certain number of resources per turn, say three, and now the cards have a cost to play. 1-20 costs 1 resource, 21-40 costs 2 etc. Suddenly it's not extremely obvious which cards are best. We can do better though.
Now let's say we give some cards effects. "12" could have an effect that forces enemy units to attack it instead of other units, which may make it a better card than "20" in some cases.
We can also create archetypes, being groups of cards that work well together. Say we give "3" the effect that units that are a multiple of 3 cost 1 less to play. Suddenly, with that card, "90" is going to be way more useful than "91" in a deck with "3", and maybe we even throw in that "12" from earlier to protect the "3".
Some of the cards from our game we just made together.
The goal with all this is to make the answer to the question of "What is the best card from this selection" vary as much as possible. It should change with the context of the other cards so that the puzzle is constantly fresh and evolving. It also leads to a variety of styles of decks, which is a hallmark of a well balanced game.
Here are a few examples of how popular card games do this with their units.
First off, most card battlers actually have two numbers, some sort of attack and some sort of defence. The importance of these two numbers is up to the player and often defines the card's role in the deck. Some games go as far as to have three numbers that are relevant to combat, but I wouldn't recommend that for a jam game.
Second, cards aren't numbered like we did, so instead of using the confusing maths I introduced, they use attributes. Attributes essentially group cards in some way. Often there is a main attribute like colour in MtG or element in YuGiOh, but sometimes there are also smaller ones that don't have much mechanical relevance unless a card refers to it, functioning more as "tags". A classic effect using this would be a captain having "All other soldiers you control gain +1", the tag attribute being soldier.
Third, cards should have some sort of cost attached to them. A resource system like I brought up earlier is often a core part of a game's mechanics and I'll explain that more later, but other conditions can exist that limit the ability to play a card, or makes them less effective if they are played outside of their archetype.
Of course, there's no reason to stick to just unit cards, most games don't. Having "spell" cards that are just an effect can open up a lot of design space and are a good idea to include. Equipment cards that make your units stronger are good, as are environment cards and even "hero" cards, which act as the character you're playing. Reaction cards that are played in response to certain triggers can introduce a lot of complication into the nuts and bolts of the rules work, but for a casual game, it'll be fine to include them.
I'd recommend picking a few types of cards and sticking to them, as there's not much point working out the rules for an entire type of card if they don't constitute a large part of the game. Established card games can have loads but that's because people already know the game and want new mechanics.
Remember, the goal is to make "which card is better" as unobvious as possible. If you make a card that seems worse in every way to another card you've made, it's a good idea to make it stronger or interface it with other cards to make it the better choice in some contexts.
Making Fun Rules
Win Condition
Let's switch our attention over to the game itself now. The first thing the player needs to know is how they win. In card battlers, the answer is usually one of two answers.
The first is "reduce your opponent's life points to zero". For this, you'll need to decide on a starting life total that players have, which you can adjust later on to give your game a good length. There is also a variation on this that works the other way, instead requiring you to gain a certain amount of your own points to win. This is rarer, but it makes the game work much nicer in 3+ player formats if that's something you want to build your game for.
The second common one is to deplete some sort of "shield stack". This comes from an extremely influential card game called Duel Masters where each player would place 5 face down cards in their shield stack, and whenever they were attacked directly, the top card of that stack was "checked" and placed in their hand. When a player is attacked when they have no shield cards left, they lose. This has a nice crunchiness to it that requires a lot less maths and does away with the need for a life tracker. It also gives the person who got hit an extra card in their hand which acts as an anti-snowball mechanic.
A lot of games use some variation of this, with the newer Digimon game being one of them. In Digimon, some cards have security effects, which trigger when that card is checked on the stack. If the checked card is a unit, then it essentially fights the attacking unit, which discourages attacks with weaker units as there's a higher chance the attacker will lose their unit. Either way, the checked card is destroyed, but this shows how much design space is in this shield stack idea.
A security stack from Digimon, with a card which has a security effect.
These are the two most common main win conditions, but most games will have alternative ones. Often "decking out", meaning trying to draw from an empty deck, will cause you to lose, which avoids stalemates and draws. Some cards might also just say that you win the game if you meet a certain condition.
Resource Systems
The resource system is one of the most important parts of card games and it's one of the trickiest to get right. At its heart, it's about saying "this card is strong so it should be harder to play", but often it ends up taking on other responsibilities.
For example, in MtG you need mana of a certain colour to play most cards. You get mana by playing "land" cards, capped at one per turn, which passively give a mana of that land's colour every turn. This ensures that you can't just put all the best cards in your deck as drawing the lands for all of them would rarely happen. This encourages players to build decks of one or two colours.
The land system also adds a dynamic to the game where players are able to play stronger cards the longer the game goes on, which is a nice progression and adds excitement. One of the core ideas of some MtG decks is ramping, which is essentially anything that lets you bypass the "one land per turn" pacing of the game, showing how important this idea is.
With resource systems taking so many responsibilities, I think it's worth talking about the options.
First off, you could just do something like MtG's land system, where lands are shuffled into your deck and there's some randomness to whether you will draw them or not, and you can play a maximum of one per turn. I'll be honest, a lot of people hate this system, as it leads to a lot of games where you can't do anything, either because you didn't draw enough lands, or because you drew too many with no cards to use them with. That being said, it's easy to design around and players will be familiar with it.
A variation of this comes again from Duel Masters, which instead lets you place any card as a resource card, which will then essentially act as a land. This retains the pacing element of MtG's system, and the cards still have colour so it encourages building decks with colour in mind. Again this is easy to implement while being more consistent than lands, although some players don't like having to choose cards they like as resource cards as they don't get to play them.
Duel Masters card placed upside down as resources, with the number showing their value and colour.
Flesh and Blood is a somewhat new TCG where every card has a "pitch" value, which is the number of resources they give when put into the "pitch zone". The difference here is that pitching is a one off thing, with the pitched cards going to the bottom of the deck at the end of the turn. FaB is notable for lacking the upwards pacing of most TCGs, with most games actually slowing down instead, and this is part of the reason. This style of pitching also creates an interesting tactic where you play to place a good series of cards at the bottom of your deck so that you can play then once you get through your deck, so players do get to play their cool cards.
Ashes Reborn is a personal favourite of mine that uses dice for its resource system which, given that you roll 10 of them a round, actually ends up being less random than lands, and the dice are coloured so they maintain that aspect.
If you're making a digital card game, then there are a few more options that would normally be too much upkeep for the player, but are handled automatically by the software. I think most people watching this will be making physical games though so I won't really go into these.
The Digimon game I mentioned earlier has a very interesting resource system. It has a "memory gauge", which has 0 in the center and goes up to 10 in each player's direction. When the counter is on your side, it's your turn, and the cost of playing cards is moving the counter a certain number of spaces to the opponent. This means that if you play a big expensive card, the cost of that is giving your opponent a lot of memory to work with, which they might use on a big card of their own or a bunch of smaller cards.
Digimon's memory gauge. Gaia Force costs 8, so playing it causes the counter to move 8 places towards your opponent's side.
Another interesting thing you can do is to move the cost of cards to the back end somehow. D Spirits is an indie TCG that allows its units to be played for free, but they cost their player life points when/if they're destroyed. This is a somewhat disguised cost that constantly begs the question "can I get away with this?"
Lastly, you could simply not include a traditional resource system at all. YuGiOh is the famous example of this, with many people pointing to its lack of resource system as its biggest flaw. But really, all this means is that you have to design custom costs to everything. YuGiOh cards often have very complex conditions and summoning costs for this reason, and this requires a huge amount of design to upkeep, although it does lead to a lot of interesting design space with a lot of decks focusing of gaining advantage from what was intended to be a card's drawback. Still, this is very hard to design and balance well.
Combat
Combat can be as simple or as complex as you want it to be.
One of the big decisions you can make is whether or not units can block by default. MtG allows this and it causes a much slower game that's more about developing your board. Other games that don't allow blocking, or reserves blocking to a few select units, end up having a much faster paced almost race-like feeling.
I'd say the latter is simpler to design, and probably more relaxed to play. Ashes Reborn has default blocking, quite complex rules to its combat, as well as that third combat number I mentioned earlier, and the result is a very chess-like combat experience which can be a bit of a brain burner. It's one of the reasons I typically don't like playing more than one or two of the games at a time, whereas I'd easily lose count of how many Digimon games I'd play before getting tired of it.
Another important decision is whether or not units have "summoning sickness", meaning "can they attack on the turn they're played". Usually you want this to give players time to react to cards, and also because being able to attack on turn one is usually a pretty big advantage.
Turn Loop
The turn loop is a good thing to figure out early on. MtG's turn loop is famously complicated, with a ton of unintuitive things that confuse even professional players, sometimes mid finals, but there are things you can do to avoid this complexity.
Earlier I alluded to reaction cards being a cause of complexity, and that's because you have to add little stages in all the parts where a reaction can be played, then you need to decide what happens if someone plays a reaction card in response to another reaction card. MtG solves this last problem with the "stack", where reactions can be played one on top of the other until no more are played, at which point they are resolved in the opposite order they were played.
Things like the stack are such a pain point for new players that some games, such as Hearthstone, just forgo this style of reaction entirely and their turn loops are much simpler for it.
With this in mind, a typical turn loop without all the reaction stages will look like
draw phase > main phase 1 > battle phase > main phase 2 > end phase .
The draw phase usually has you refresh your units and resources and of course draw cards. The reason there are 2 main phases is so players can play units before or after battle which is a nice way to give them more freedom in a non confusing way. The battle phase is where combat occurs and the end phase is just there to be referenced by cards that do something at the end of a turn.
You could also have as many swaps between battle and main phases as you want, although typically each unit would be exhausted/tapped after attacking so it only attacks once per turn.
So far this has been a pretty traditional turn loop, but there are other innovative systems that give a different feel. Usually these different systems are aimed at avoiding long player turns, as this can leave players bored while waiting for their turn. To do this, turns are either made shorter or designed in such a way where both players are making decisions on either player's turn.
Ashes Reborn goes the first route and simplifies the turn. Every turn you get a main action and a side action. A main action is something like attacking, playing main cards or summoning most units. A side action is optional and allows you to play some "faster" cards and effects. However, these much smaller turns are then grouped into rounds. Each round continues until both players pass their main action, and a first player token is passed to keep track of who gets first turn each round. Drawing cards and refreshing resources is done at the start of the round, so you can plan out several turns ahead as you know what you'll have for them.
Ashes' turn loop and first player token.
Legends of Runeterra has a similar round system, except instead of a "first player" token, it has a "battle" token. Players can only attack on rounds where they have that token which gives a very dynamic alternating offensive/defensive feel, all while keeping turns short enough that both players are constantly engaged in what's happening.
Personally, I think the turns-within-rounds style has a lot of flexibility and is quite an easy way to make your game feel different to what most people have played.
Balancing
The last topic I want to get into is balancing. Don't expect your game to come out perfectly balanced, as this is something that takes entire teams of designers and loads of playtesters to get right, and even then there are mistakes. Instead, this is just going to be some advice on how to design your game to make balancing a bit easier.
Avoid the number 1.
If your card has a cost or attack of 1, then it's very difficult to tweak that without either doubling it or setting it to 0. Of course you could do decimals, but don't. This is why games like YuGiOh will have units with thousands of attack, it offers granularity if you need it.
If you have a resource system, be strict with either resources or card draw.
If it's cheap to generate both, then you have to worry about infinite combos where cards are generating more than they're costing. This is fine in solo games like Slay The Spire, but in two player games balancing this can be difficult. The easiest way to do this is to make extra card draws very rare so that cards are the bottleneck, or make it difficult to recur resources within a turn so that bottlenecks. Typically drawing cards is fun, so most games choose to be strict with their resources.
Every card should belong in some decks, and no card should belong in every deck.
If you have a powerful card that every deck wants, then it's going to make the variety of viable decks much smaller. Exploring different ways to rein in powerful cards is a really fun thing to do though, and sometimes the card itself doesn't need touched. Simply adding a card or archetype that counters it can discourage players from using it in every deck, and some players love building around countering strategies rather than building decks in a vacuum.
Thanks for reading! If you have any questions, feel free to reach out on Discord or in person at the jam.
