Spaces moving back and forth
Some thoughts on card games and roguelikes
The following is the (3k words long!) draft of a chapter dedicated to Slay the Spire (seriously, do I need to link to it?) and deck-building games. I am still working through these arguments as part of my project on the roguelike as a poetic form, so a) I am still trying to figure out what I am doing, and b) changes will happen to these ideas. All images in this post are generated with Midjourney using the Mitsua One model, in an attempt to reflect the dreamlike nature of playing Slay the Spire. Here be dragons.
I've never been a fan of Magic: The Gathering. There, I said it. My game scholar credibility is lost. But let me give you some context. When Magic was released, I was very knee-deep into tabletop role-playing games, running years-long campaigns of Call of Cthulhu, and trying all possible game systems we had access to in translation, from Cyberpunk and Kult to Aquelarre and Fanhunter1. At that time, a game like Magic: The Gathering felt like constraining my imagination. I wanted to create worlds, to scare and entertain my players, and to have fun in an improvised way. The few times I played Magic back then did not convince me that the beauty of the assembling systems and evocative powers of the decks was more enticing than what words could do to my players. So, I never got into Magic. Card games were not my thing.
I am still surprised that I have logged over 1000 hours in Slay the Spire. I shouldn’t like a deck-building game, and yet here I am, trapped in the spire. Every day I play the game. In fact, the only thing I play is the daily challenge, a new run with specific constraints that I eagerly wait to learn about as a ritual to start or end my day2. My condition of being captive of the spire probably made me reflect on roguelikes more seriously. There’s a lot written about the mathematics of these games and the appeal of deck-builders as systems, and it is neither my interest nor my talent to investigate these games from that perspective. I will look at what the form of the roguelike does to the deck builder3 looking at Slay the Spire.
Since it’s been a while, let me quickly summarize my premises: a roguelike is a poetic form used to create novel forms of videogame expression, altering existing genres. This alteration results from applying permadeath, procedural generation techniques, and secret-based gameplay to video games' structures, tropes, and conventions. What Slay the Spire does, then, is inject the deck-builder with these elements, creating a different type of experience.
But how different? When I started looking into roguelikes, I realized that maybe all deck-building card games are roguelikes. After all, most deck builders I know have permadeath, meaning that once the game sequence is over or a player loses their cards, the game is over (for that player). They also have a form of procedural generation since randomness techniques are applied to constructing game elements (shuffling the card deck is a sure way of adding randomness to the game). While I don’t have any concrete examples, I am confident that some card games have secrets and powerful combinations that must be experimented on and discovered.
Understanding the commonalities between analog deck-builders and their roguelike interpretations is very productive in understanding the latter as a poetic form. A roguelike is a videogame form; therefore, we need to remember its nature as the software we play with. The starting point for understanding roguelike deck builders must be procedural content generation. Not just the addition of computable randomness, like shuffling the deck, but the creation of the main challenges of the game through the application of one or more generators.
Roguelike deck-builders make players construct their decks to face challenges they can know once they have played theg ame several times, but cannot predict at the time of play. That is, players need to create a strategy to play the game without knowing exactly how it will present them with challenges or whether their plan will succeed. Procedural content generation in a deck-builder roguelike is used to make the progression in the game a secret, making each of the choices taken by players when selecting their cards to play with more critical.
In Slay the Spire, players are given a starter set of cards. They start facing opponents with a varied set of attack and defense strategies. As they progress through the game, ascending the levels, they face different enemies and collect other cards. The enemies and the availability of cards depend on a series of randomizers.
When playing Slay the Spire, knowing what type of deck to build for each character is fundamental. My favorite, the Defect, has several different strategies that allow for other ways of dealing with damage, all connected to the capacity of that character to harness extra protection, extra turns, or extra damage. When I play Slay the Spire, I start by deciding which type of deck to build, and then I try to adapt it to what the content generator has decided will face me that day (remember, I play only the daily challenge!). Sometimes I succeed, sometimes I don’t. The strategy contained in my selection of cards may break down. Or maybe I don’t get to build the right deck because the cards are not part of the selection available this time around.
That is a critical difference between analog deck builders and Slay the Spire: PCG defines what type of deck I can build and how effective or inefficient it will be. In Slay the Spire, the type of cards available depends on the results of the randomizer. The system, combined with the careful design of the game, almost always allows for creating an appropriate deck. However, there is a risk of settling into a strategy that will break down because insufficient cards are available, or because some enemies appear too early in the run. In the case of the daily, it may happen that some the cards most appropriate to a character are not even part of the possible deck.
Part of the roguelike experience in Slay the Spire involves the uncertainty of deck building. The dance between the player's agency in constructing a deck and the PCG systems determining which cards are available for play creates a fun level of uncertainty. The Slay the Spire subreddit is full of stories of players who build a particular strategy only to discover that they do not have enough of the right cards to implement it. Therefore, they die in ridiculous and epic ways4.
The roguelike form adds a sense of uncertainty to the deck builder card game genre in the outcome of any strategy. Because the climb in any Slay the Spire run is randomized, a plan is only a starting point for a player. They soon need to learn all the cards, how they can work synergistically with the best cards for a particular strategy, and how they can collapse if the player meets the wrong enemies on the wrong level.
Let me analyze this: I like to think about deck builders as clashes of possibility spaces. A card deck is an open and broad space of possibilities5. A strategy is narrowing that space into a series of actions that allow players to traverse it with a high chance of winning by selecting cards. A built deck is a materialization of a possibility space. In games like Magic, we witness a clash of possibility spaces. Each player is trying to drag the other to their possible space since, in that space, they will have less agency and less possibility for action, and they will then lose. Winning in a card game like Magic and Dominion implies having one space of possibility to succeed over the other6.
How does this back-and-forth work in a roguelike version of a deck builder? In the case of Slay the Spire, players build their possibility space as they play the game. Once the run is started, the game's possibility space is already set but unknown to the player. It is a secret. The player does not have any knowledge about that possibility space; worse, players know that the possibility space of the game is not intentional but random. The only thing players know is the cards they have and the cards they can have. That’s why learning to play the game, developing knowledge about enemies’ tactics and potential decks, is unwrapping a sense of safety in how the game will go. The more we know about the game and the metagame (a topic for another post), the better I will be able to craft different navigatable possibility spaces. This pleasure of Slay the Spire will die when all that is left is marginal gains on the decks we can build, because most possibility spaces are known. The vastness of the combinatorial possibilities of the deck and the enemies makes that death as remote as the heat death of the universe, for most players. And even then, there will always be the pleasure of setting in motion the systems we’ve created and “contained” in our deck, the rube-goldbergian enjoyment of setting things in motion through the chosen cards.
Slay the Spire does not deploy a strategy to counter the player. As the gameworld lore suggests, it is a spire unknown to the player, ignores their presence, and needs to be conquered. Playing Slay the Spire maps the space and the possibility of a generator's system creation. The game is conquered by mapping the procedural generations’ possibility space and creating a flexible and adaptable strategy.
Slay the Spire uses roguelikes' poetic characteristics to mutate the deck builder genre. It uses procedural content generation to create an adversary to the player that is interesting and challenging but ultimately not guided by any human strategy. Randomness creates uncertainty in the outcome, for not all strategies will be possible in a given ascension/run. Furthermore, the fact that the system has balance but not intentionality feels uncanny, as it is the mind of a machine with no internal representation of a strategy but that acts and reacts to our actions as we play7.
Slay the Spire, wraps this generated level in secrets. We don’t know what the next foe will be. We won’t know what the available cards in the run will be. And we will not know, unless we play extensively, how cards work with each other and which cards are good. The deck and all its possible strategies are a secret that we unwrap as we play. Each run has a secret possibility space that we can unwrap every time we start an ascension.
Permadeath makes that secret and that strategy selection meaningful. We will only have one chance to explore this particular possibility space, trying to overcome it with the strategy we develop. Every card and every choice in our path is critical because it will bind us to a particular picture of that space we are forming as we play. If the image we are creating is more or less correct, and we manage to translate it to our deck, then our possibility space will overlap with the one generated by the system, and we will win. If we don’t, we will lose those spaces forever8.
This is the crucial difference between deck-building and roguelike card games like Slay the Spire. I’d argue that, conceptually, we can think about analog deck builders similar to games like American football or soccer (it is, at this point, this newsletter must mention football at least once per post). They are games of conquering space and taking as much territory from your opponent to score points. The spaces in deck builders are the possibility of action afforded by the cards collected in a deck. For a deck to succeed, it needs to minimize the opponent's possibilities and maximize the opportunities for the deck9.
In Slay the Spire and other similar games, the roguelike form alters how these spaces collide. It is not anymore a matter of two decks competing to eliminate the other but one deck being built as the game is played to map and conquer the space created by the randomizer. The Rube Goldberg device metaphor from earlier can help: playing Slay the Spire is the practice of creating systemic engines that can traverse a randomized possiblity space in the most efficient and pleasurable way. This practice is limited and enhanced by the randomized nature of every run: when constructing the engine, we don’t know what the enemies will be, what the shops will offer, and what the treasures may yield. We’re building an engine for an unknown space. As we progress, we try to adjust as best as we can our deck to that ideal engine, given the possibilities dictated by the generator. Sometimes we will succeed, but often we will yield to the inhuman beauty of the challenges created by the Spire.
The roguelike form transforms deck builders into games that are about simultaneously constructing and conquering possibility spaces afforded by the design of the card deck. That synchronicity and the raised stakes due to permadeath make the roguelike form a creative approach to change deck builders. Returning toSlay the Spireis a familiar territory in which I know that every day, I will have a new map to explore, a new territory to conquer, and a new space to create, one card choice at a time10.
No, I never really got into D&D, because I'm not too fond of swords and I have no fantasy. And yes, if you are not Spanish you won’t have heard of these two last titles, and you are missing out
Besides the brain hook that all roguelikes have for me, I cannot separate my interest in Slay the Spire from its context. When the Danish government announced the first lockdowns to prevent the spreading of COVID, I had just bought a Nintendo Switch. As those long days of uncertainty and home passed, I needed to play something, so I tried out that deck builder that looked interesting but had not convinced me a year earlier. I decided to play the daily challenge. And all of a sudden, the daily challenge in Slay the Spire became a milestone in my COVID lockdown days. It was a ritual that gave order to my day, an hour of evasion, and a connection to all those other players who were also trying to slay the spire somewhere in the world, probably under some form of lockdown and as worried as I was. For me, Slay the Spire was not a game but a ritual of certainty in a time of uncertainty.
The traditional history of deck-building games lists Dominion (2008) as the first proper example of this genre, while Hand of Fate (2015) is the first roguelike take on deck-building. We are talking here about a relatively young genre.
This is the time to write about roguelikes and stories and narratives before I get into Hades: Besides their aesthetic form, roguelikes are fascinating cultural objects because of the stories they create. Not the stories they contain, not the narratives suggested by the designers through the gameworld lore and the characters, but the stories we tell through them and what they say about us. Because of the procedural content generation element that makes each run unique and the importance of permadeath as making each run an exceptional event, roguelikes are storytelling machines. Only the stories are for us, players, to tell. They do not depend on the game, but they are accounts, maybe even epic in the old tradition of poetics, of our encounter with these forms of fun software. Without the stories we can tell about our runs, roguelikes are just video games. Those stories, whether we speak them out loud or keep them to ourselves, are essential to understanding why we play these games, how we learn to play them, and their lasting aesthetic power.
As my friend Martin Pichlmair wrote to me when reading a draft of this text, a card deck is more interesting than that: it’s both an open and closed space of possibilities. It is closed, containing only a limited number of cards. But shuffling, the OG procedural content generator, turns the deck into a relatively open and broad possibility space. I think that the interest in card games that some of the great game designers out there, from Martin to Zach Gage, is derived not only from the apparent ease of prototyping but also from this tension between designer control (deck size) and required openness (shuffling).
The concept of possibility space is an old one in game design studies, but one fundamentally undertheorized -unless I have yet to notice something entirely possible. Salen and Zimmerman’s classic Rules of Play mentions the concept and uses it extensively, and my favorite exploration of it comes from Zubek’s Elements of Game Design. I like to think about the possibility of space as the most crucial concept in game design theory. I usually conceptualize it as an umbrella term for at least three concepts: the designed space of possibility, the perceived space of possibility, and the actual space of possibility. There’s a lot of Donald Norman in my thinking. Still, the idea is that designers create a possibility space and try their best to communicate it to a player, who perceives a space of possibility that may or may not be isomorphic to that of the designers. Finally, there is an actual space of possibility that is different, sometimes broader and sometimes narrower, than both the designed one (as it often happens in games of emergence or roguelikes) and the perceived one (because players are frequently blind to some paths until they learn them). Maybe at some moment, I should write these ideas more clearly, as they thread my interest in play and my distaste for overtly complex games and puzzles.
Time for the semi-regular observation about cybernetics: what a beautiful black box Slay the Spire is! Pickering’s idea on how to think about cybernetic systems has this game as the perfect example: while we ascend in our run, we don’t know what the game looks like, so we need to react to how it is being presented. The level in Slay the Spire is not made on purpose; we don’t know how it was created or what it may mean. It ignores us, and we can only make sense of it by playing, shaping its contours, and building a strategy that will hopefully constrain what the game can do into something we can respond to and overcome. In other words, it does not matter how the run in Slay the Spire works from a technical point of view; all that matters is how we experience it as we ascend and how it affects our construction of the deck.
Alright, alright: technically, if we know the seed of a particular generation, we can replicate it. But seriously, what is the fun in that!? (I know what the fun is, but mine was a rhetorical question.) Embrace permadeath, and allow yourself to have fun!
And yes, many other ways of seeing these card games exist. Mine is, on purpose, not a statistical/mathematical take on them, even if these possibility spaces are mathematical. I think it’s productive to see Slay the Spire as a series of possibility spaces that evolve as the player collects cards and discovers how the game has generated a particular run. If you like another approach, please feel free to write your 2000+ word essay. Thanks for coming to my TED talk.
Yes, yes, what about Balatro? First, say no to drugs. Or maybe say yes to Balatro and no to all the other drugs. The question, I guess, is how this 3k-word rant/book chapter maps with Balatro. Well, Balatro does to poker what Slay the Spire does to Dominion: take a genre and game, add permadeath for meaning, procedural content generation to increase variability but also to immediately strange the player from know tactics and habits, and add secrets that relate the way the systems interlock with the meaning of elements in the game world. In Balatro, the jokers are secrets and part of the deck-building strategy, and the fact that the hands are randomized and that we don’t know what the result of our hand will be makes it closer to the pleasurable uncertainty of constructing a deck that can successfully overcome the game. The main difference is that while Slay the Spire has several enemies with different tactics, in Balatro, we often play against our stupidity. It’s sort of like the Soulslike of deck-building roguelikes. You usually die because you are too stupid or evil at playing the game. But the principle of action, creating a deck as the continuous expansion of a space of possibility, is similar to what Slay the Spire does