Why make card games?

written 2026-06-23

So you want to make a game. Why make it a card game? Why not make it everything else in your vision, but without cards?

I think card game-ness is cargo-culted, because people don't know what turn-based strategy games look like without cards. There aren't many, besides abstract games like Chess, which are very different from what most designers want to make.

So, let's think about what a card game could look like without cards. I'll pick Cobalt Core, a roguelike deckbuilder about spaceship combat, because that's the game that got me thinking about this, and I already have a review that describes how it works. But this analysis applies to other card games too, at least others that involve cycling through your deck multiple times, like Slay The Spire, Hellcard, and Dominion.

Cobalt Core review

What could Cobalt Core look like without cards? You'd still have the mechanics of energy, hull, shield, evade, attack, and the midrow, but cards could be replaced with abilities that don't need to be drawn from a deck before being played. If you could play any of your abilities whenever, then stronger abilities would completely outmode weaker ones, so it wouldn't make sense for the starter deck to have cards like the generic Basic Shot which are strictly worse than character-specific ones like Block Shot. We could solve this by buffing the generic starter abilities, or nerfing the character-specific ones, but for this experiment let's say we want to change the game as little as possible, so we don't want to have to rebalance all the generic abilities, and we don't want you to be able to play any ability whenever.

Instead, we could have a cooldown system, where each ability can be played once every couple of turns, or every couple of other ability uses. This would copy the dynamic of not being able to just spam your best ability, and so Basic Shot could have a use because you might want to attack when all your other attacks are on cooldown.

We could also make the cooldown length depend on the number of abilities you own (the size of your "deck"), which would copy the deckbuilder dynamic of "the more cards you have, the less often you can play each one".

What about translating card draw? Well, what does it effectively do? It gives you more control of which of your cards you play. So we could translate a card that draws cards to an ability that lowers the cooldown on all your other abilities.

But in Cobalt Core, cards don't just go get discarded when you play them. Most cards get discarded at the end of your turn even if you didn't play them. (Some have an attribute that exempts them from this.)

What does that effectively do? It means you might draw your best cards together and not have enough energy to play them all. It also means you might draw parts of a combo separately, and not be able to execute the combo. It nerfs combos, long-term effects, situational cards, and makes tactical plans less reliable.

We can apply the same rule to our cooldowns, but then different abilities have to start on different amounts of cooldown, so they don't all go on cooldown at the same time and leave you with nothing to play. As for how much each ability starts on, there's a lot of ways to answer this. We could let the player choose it (effectively choosing the order of your deck instead of random shuffling), which would add a lot of interesting decisions to preparing for battle. But this wouldn't preserve all the balance effects: you'd be able to guarantee your combo pieces are available together and your long-term effects are available early, so those might be too strong.

Another option is that each ability has a fixed amount it starts on. This would mean that depending on which abilities you have, you might have some turns with many options, and some with few or none. This would add interesting decisions to deckbuilding, since you'd want to ensure you have enough abilities off cooldown to spend your energy every turn. And you might want to ensure you have some defensive and some offensive options for every turn. It would also add parameters for designers to tweak to balance abilities, for example combo pieces could be given the same starting cooldown to make the combo easy to use, or different cooldowns if the combo is too strong.

Another option is that their starting cooldown is based on when they were added to your repertoire, or their alphabetical position in it - that would also add interesting decisions to deckbuilding, since adding or removing an ability could change which others will be available together.

At this point we've copied most of the strategic dynamics of Cobalt Core, maybe added some, and maybe removed some. So what's the point? How would this version of the game differ from Cobalt Core as it exists?

Mainly, it would lack randomness. This would make the game more fair and enjoyable. But does it have downsides too? I think the only real downside is that it's harder to explain. The game as described above probably sounds a bit strange and confusing. But it's not actually more complex than card games, it's just close equivalents of the same mechanics. Why does it seem more complex?

Skeuomorphism

Because cards are skeumorphic. They represent a familiar physical object, providing intuition for your game mechanics. You don't have to teach players what a card, hand, or deck is, they already know it from physical card games, so if you make a card game, you get all the strategic dynamics that come with that without increasing your game's learning barrier. In my above version, you get the same strategic dynamics, but the mechanics are harder to explain.

But skeuomorphism only helps if your thing works like the physical object it represents. But many digital card games have abilities that don't work like cards, and they come up with convoluted ways to implement them as cards. In Cobalt Core, there are cards that represent innate abilities of your ship (like switching which of your cannons is active) that you always have access to. Some are implemented by giving them the "retain" attribute, which means they don't get discarded at end of turn, and the "recycle" attribute which means they go to the top of your draw pile when played instead of to discard. Others are managed by a status effect that checks if you have them at the start of your turn and adds them to your hand if you don't. These are awkward ways of implementing something you just want to be an ability not subject to card draw randomness. The card metaphor hurts here, makes the mechanics more confusing, not less.

Conclusion

So there's my point: not that being a card game has no benefits, but that card game-ness is cargo-culted, many designers make card games not because they considered it and decided it was appropriate for their vision, but because they've seen a lot of card games and so few non-card turn-based strategy games that they can't think of other ways to design a turn-based strategy game.

Actually, Cobalt Core has some abilities that are not implemented as cards. Evade and droneshift are resources you get from cards, but the buttons to spend them aren't cards. So why is the Ares cannon switch implemented as a card?

I also want to say that if I were to design a game with Cobalt Core's premise from scratch, I wouldn't necessarily copy all those dynamics. I wanted to show how it was possible, but that doesn't mean it's the only or the best way to design the game. If I were making it from scratch, having to rebalance things wouldn't be a problem, so I could, for example, make it so abilities don't go on cooldown when you don't use them.

I think the best way to end this article is to list turn-based strategy games without cards that I'd suggest game designers to study.

Prismata
Invisible, Inc.
Into The Breach

Invisible, Inc. and Into The Breach are about positioning multiple units, much like Chess. Prismata is very unique and was actually inspired by the thought experiment "what if card games didn't have randomness?"