Solving the Spire — Better living through Card Games
Interview with Stephen Flavall by Luke Shaw (@FullOfFeathers on Twitter)
Stephen “Jorbs” Flavall has played over 3500 hours of Slay the Spire. It is his job now, to stream daily Spire runs and upload Youtube videos that range from highlights to in-depth, analytical seminars he dubs “Spireside chats” that focus on everything from card analysis, to conceptual discussions of the difference between Nemesis and Optimal players in Poker, and how that maps on to the idea of playing a perfect game of Slay the Spire. Gaming content on Twitch and Youtube has largely passed me by, but in my early days struggling with Slay the Spire, a frustrated google search led me to one of Stephen’s “over explained” runs, where he spoke in length about concepts such as “equity” during his runs, how and why he made decisions, and what the game was actually asking from him rather than what many people (myself included) may believe it is asking from them.
His style is calm, friendly, and relaxed, and his cool analytical style comes from his time playing online Poker professionally — he hired a poker coach who instructed him to analyse his own play as a method of learning: “He had me playing sessions of poker, recording myself for an hour and explaining everything I was doing to an imaginary audience and then at the end of my session I’d stop the recording, rewatch the video and take notes and criticise myself and say “hey, that thing you said doesn’t really make sense…” and “hey, right here… do you really know what you’re talking about?” This experience, coupled with a background in Chess, Magic: the Gathering, graduate level Statistics experience, and time analysing some of the earliest written stories — “the messiest but richest data in the entire human experience: looking at how people were thinking and what they cared about before we even had writing.” — has set Stephen up well for a career built around parsing what is essentially complex data, and working out underpinning theories that he can use to explain how to approach not only Slay the Spire, but everyday life in general.
Stephen came to Slay the Spire off the back of other games that rewarded the ability to examine, understand and reflect on difficult data sets. He had previously worked his way through the XCOM Long War mod, and was subsequently invited to work as one of the Legend balance testers for Long War 2. Working on videos and creating content made him feel like he was not only doing something he enjoyed, but also creating something that allowed people to “tune in, and feel like they were getting better at the game” which gives Stephen a unique selling point that sets him apart from many other streamers, which is why I decided to reach out with him to ask him why he’s made a game like Slay the Spire his daily job, and to dig into some of the nuances in both the game, and his approach to it.
Playing Your Cards Right
What are some concepts you think could help players in general to play Slay the Spire to a higher level of success?
I think the biggest, simplest lesson you can learn about Slay the Spire in order to start getting better very quickly is to recognise that you don’t build a deck which is an end product which gets tested like you do in Hearthstone or a Magic: the Gathering Draft. In Slay the Spire, pretty much every single floor you’re going into a fight with the cards you currently have. On one floor you need to kill three enemies with 12hp, on another you need to kill a 500hp enemy. In the first fight Area of Effect cards are really helpful, in the second they don’t do anything at all. What’s really the point of having your deck do the same thing when it’s being asked to do all sorts of things? That’s what keeps Slay the Spire fresh and interesting for me. It’s just a much different way to build a deck and have yourself test it over the course of a run.
Is this a case of players not recognising what they’re doing wrong? Is the game not doing a good job of telling you what your mistakes are?
I think at low difficulties you’re not being asked to do as many difficult things by the game, right? And even if you’re not doing things perfectly you can still get to the end a lot. So at low difficulty you get far less feedback telling that you’ve done something that wasn’t good as it could have been. It’s hard to even see that you’ve done something wrong when you’re just hanging out and playing the game and winning half of your runs — that’s really good actually, winning half your runs when you’re new to the game. It turns out if you do everything perfectly you should be able to win 99/100 of your runs at low difficulty, but the game doesn’t violently punish you enough for not not making the absolute perfect decision at the low difficulty so you don’t learn from your mistakes quickly. Once you reach A20 and you reach the Heart the game doesn’t hold any punches. At that point if you’re doing something wrong it’s going to be pretty obvious because you’re going to die very quickly.
This is abstracting away from the game which might be good or might be bad — I think Slay the Spire is complicated enough that when I think about how I’m learning it and how I am playing it, I am thinking about other things such as, how I am living my life? How I am interacting with other people? How do I learn how to do those things best?
If you imagine somebody who’s on a career path where they are going to be a teacher, sure, they have their career path that they’re on and they have their daily routine. They know what they value as a teacher and they have relationships with students and they understand the different lessons that they teach. They know what their day being a teacher looks like, and then they get to practice that for 40 years or so if they want, until they retire and that’s their life as a teacher.
In Slay the Spire there are so many complicated things going on and all that stuff is happening that decides whether your deck is successful or not, it’s almost the same sort of level as going through a day as a human being, doing something you’ve chosen to do. Except, instead of waking up every single morning and being the same person you wake up and you’re something completely new. And you have to work out exactly what you have to do this day, based on new cards that are given to you, and new enemies you have to face at the end of the Act, and new relics you get that you haven’t gotten before ever. If you’ve played thousands of runs of Slay the Spire you might have had similar cards and relics, had similar experiences to draw on, but you’ve never had this exact life before. You have to work that out over the course of 50 some floors, just on the go, and it is very difficult, and it’s very hard to take a world that you’ve never been in before really and work out what you’re doing wrong in it. You have to have a lot of innocence and naivety and be willing to try new things and struggle when they don’t work.
Has some of your objective decision making just become cultivated decision making? How often do you stop and make deliberate calculations, or do you only do that on your over explained runs?
Where I live there are mountains and passes called “Mount Deception, Deception Pass” and so on because of explorers who were exploring Washington State. They were like “Oh that’s the tallest peak in the area, we’re gonna climb it!” but Washington State is densely wooded, so they’re hacking their way through the undergrowth, and they get to the top and look around they see there was a taller mountain. So they get upset and they climb down and then go and climb that one.
In Slay the Spire when you are practicing your strategy, it can feel a lot like that. You see the top of the mountain you’re climbing, and you hack your way toward perfection. There could be a guiding principle you have that you believe is correct — “I should be fighting as many as Elites as I can get away with so I can get strong enough for the next act” — and you’re working on that.
The problem is that sometimes you realise there is a taller mountain somewhere else, and everything you’ve been practicing so far, every strategy you believe in… they were slightly wrong, so you have to start from scratch. A lot of the assumptions you were making were leading you consistently to the same mistake over and over again.
I have gotten to the point finally where I feel like it doesn’t happen to me very much unless I am playing The Defect. For The Ironclad and The Silent it took a very, very long time before I was able to say “Ok, my basic assumptions for this game are at least at the point where I am trying to climb the right mountain.”
What are some of these “deceptive peaks” you think are useful to identify with each character?
Slay the Spire is really nasty too because it gives you different difficulty levels. A15 was a different mountain to A20 Heart — the game was asking me to do something different. I think any time that you have some solution for the end of the game, and you’re building toward that solution for the end of the game, if that’s all you’re doing there’s just no way that’s all you need to be doing. You’re missing something big, which goes back to the idea of preparing your deck for fights on every single floor. That is something that new players struggle with, identifying their end goal and not their steps along the way.
For Ironclad in particular it took me a long time to work out how powerful the Exhaust synergies were. You don’t need Corruption and Dead Branch together. Corruption on its own is very powerful and has synergies with about 15 other things so if you take it it’s very easy to get another piece of the puzzle to get your deck to a winning state.
For The Silent it took me quite a long time to recognise how good Silent can be at turn one and two. It’s funny, I started the game and as an MtG player I liked playing combo decks so I tried to build only Silent decks that beat on turn one, and it took me all of a day to work out that wasn’t what I was supposed to be doing, even though it was fun for me. I got away from that because for A15 it was better to be building a value driven Silent deck which can play turn one decently and turn two decently and then it plays Footwork and Noxious Fumes and all of a sudden it beats every enemy in the game.
That was enough for A15. Once you get up to A20 on the Heart, I think Silent is often being asked to have that explosiveness that a Turn One kill Silent deck would have. But you can’t build that, you just need that level of explosiveness so you can get Noxious Fumes and Footworks down on turn two and cards like Calculated Gamble, Adrenaline or Acrobatics and relics like Bag of Preparation — things like that become very important for The Silent, and that’s the most recent thing I learned for A20 Silent was how crazy that could get.
In comparison, are you struggling with Defect?
Yeah, the way I played Defect before A20 was quite passive — scaling with my orbs and getting a lot of block. Damage was always an afterthought, every Act 3 boss in the game you can easily outblock the damage, so it made sense to set up, but if you try to do that against the Heart it will most likely kill you. So trying to work out how to do 800 damage whilst blocking becomes very interesting, especially because you’re letting Frost Orbs passively generate block, cycling Lightning Orbs doesn’t really fit in well with what I was doing before the heart, so I need to work out how to build a strategy that can block maybe with cards like Genetic Algorithm, Reinforced Body with Echo Form, or a lot of Coolheaded so I am evoking a lot of Frost Orbs, whilst I have to deal 800 damage with Storm and powers evoking a lot of Lightning Orbs or Darkness+ to make a lot of Darkness Orbs, or Thunderstorm, or Blizzard or Claw — there are lots of ways to do it but I was never being asked to do it before the Heart came out so it’s sort of a new space where I’m still learning things.
The State of Play
How do you feel about the overall balance of the game at the moment, at Ascension 20?
Slay the Spire balance wise is very difficult and I think the developers have done an incredible job with it. There is only one card pool for each character and only one relic pool and they’ve tried to make 21 different difficulties and like endless and daily runs and custom runs even, tried to make all of those be interesting despite being limited to the same relics and cards for every single time.
I play at Ascension 20 and I don’t expect the game balance to be perfect at that level because honestly the developers should balance the game so it’s fun for a new player I think. A lot of the balance for the game should be making sure it’s fun for a new player. But, people want to watch me playing the hardest difficult that they care about, so that’s what I wake up and do every day, sort of for my job? It’s what I want to play too, probably, but I don’t spend a lot of time thinking about it because I think it’s what a lot of people want to see me play and so that’s what I give them.
Balance wise, the Defect definitely struggles the most with A20, Ironclad and Silent I think are around about the same success rate. I have around a 30% Ironclad win rate, 35% Silent winrate and 15–20% Defect winrate I would say? In terms of how many different solutions the classes have for winning the game I think Ironclad only has two real ways to block for enough to win at the end of a run so that can be getting to a point where it’s a little stale because you’re so limited, and then Silent has four or five different options and Defect… ha, I feel Defect can win just about any way but largely that’s because Defect doesn’t have all that many sure ways to win, it ends up being a case where you find some very powerful interaction between cards or relics over the course of the run and have to probably get a little bit lucky as you go through the final part of the game as well. But Defect runs can be fun because of how many different ways you end up getting a winning deck.
You’ve worked in the past on solving games like Poker and chess, but you’ve said you won’t ever solve Slay the Spire because it’s just too big. Is that because RNG is too much of a factor?
Humans are in a very interesting place right now where they cannot solve the very complicated games like Chess or Slay the Spire yet, but they can build machines which can play better than any human can. So we can’t access that truth where we now understand this is the correct way to play the game. All we can do is create an artificial brain which can demonstrably do it better than we can, and it’s sort of a lame place to be as humans because the thing we really want to know isn’t what some machine that is imperfect but is better than us would do, we want to know what the right thing is to do.
When I’m sitting in a tough situation the Ironclad we want to have solved the game in a way that means we can look at our current situation and be able to work out that we should definitely play these cards in this order, and then have an explanation of exactly why that is. But, all we can get right now in a game as complex as chess or Slay the Spire, is the ability to play it a lot and get pretty good at it relative to other humans, or we can build some neural network or machine that uses a technique to understand the game which lets it process the game better than we do and then in turn we can ask the machine what the machine would do. The problem however is that machines aren’t very communicative all the time, so sometimes the explanation it gives us isn’t very satisfying, and also they’re also imperfect so sometimes they make wrong or suboptimal decisions based on how we’ve programmed them to learn or understand the game.
One thing I like about Slay the Spire over Poker is that there isn’t a machine that can play Slay the Spire better than me yet because no one has done the work yet. There are already machines that understand poker better than I do, definitely that’s true for Chess too, and Go recently with Alpha Go having come out in the last couple of years.
Have you thought about branching into teaching decision making, statistics, and game theory on its own, or do you stick with Slay the Spire because it allows you to demonstrate the concepts effectively?
My mission statement as a streamer is to take complicated things and make them approachable for my audience. Take something that is very difficult to engage with just because of how much stuff is going on, and boil it down, basically open the door and let them into the house where they can see all of the stuff that is going on and really enjoy it.
Your experience watching my content is largely what I am trying to do. You were saying “I’m not that great at the game” but you’re also saying people who win the game aren’t getting much interesting out of it — they’re talking about deck archetypes instead of the things that are going on under the surface, but I feel like your experience is cooler, you’re getting more out of it than they are. That is my hope as a streamer. I think games are an incredible medium to do this but something that I constantly try to do, is to try and take a lesson from a game and demonstrate how the sort of thing that we’re using the game to think about, could be something we thought about in real life. Whether that’s deciding which food to order or deciding how you’re going to enrich the life of the person you’re in a relationship with, or something like that.
So for example, in a typical stream I’ll be talking about making flawed assumptions based on what’s in front of me, and I’ll say “hey, this isn’t unique to Slay the Spire, this is something humans do all the time, there are all sorts of things we have flawed assumptions about, and if we re-examine our assumptions — which is a really difficult thing to do — maybe we would end up better at life in general?”
My friend who taught me poker now does productivity coaching. He basically tries to teach people how to solve life. He’s no longer about solving poker, he’s asking how are you going to wake up every morning and go through the steps of the day and make yourself as happy as possible, as productive as possible, and help other people as much as possible. And, that is sort of the end game for me as a teacher.
This is why the games are a great medium for teaching, because everyone comes to the game with the same rules: you generally don’t have to worry about some people having more privilege than others. For the most part when you come to Slay the Spire, you’re on the same level ground as everybody else and you have the ability to interact with things in the same way that they do, which means I can explain logic and have it appeal to a huge audience of different people and that’s very powerful compared to trying to help someone make small improvements in their life, but depending on what their life is like the stuff I’m saying may be completely irrelevant to everything they’re doing. I think it’s easier to demonstrate how to do that in a game, and then say depending what your life is something like this might be applicable to your life as well.
That’s an interesting concept, absolutely, especially when you talk about the muddy waters of privilege — of luck versus circumstances…
It’s not always true, because some games give you a narrative where it assumes you’d have a certain life experience where you’d understand where that narrative comes from and some games very much serve a particular culture, and I think Slay the Spire serves a culture of people who have played roguelikes and card games growing up, and so some people come to Slay the Spire and know more than others. Games do a better job of this than many things. It’s easier to connect to people having over a shared experience in a game than it is over a lot of other things.
For many people that is the end goal of games, to give people a shared experience.
So, using Slay the Spire as a kind of a leveller. Is there one main concept from Slay the Spire that you would love everyone who watches your content to come away with and take that away into other games or different life scenarios?
The two I would say which almost go hand in hand, and these sound like platitudes almost…
Number One: It’s cool to have a destination, but every step you take along your way to that destination is important and productive and something to be thinking about too. A goal is good because it teaches you where to go, but you need to use that for making decisions at each intersection. You can’t just aim for that goal and turn off the rest of your brain.
Number Two: It’s important to have assumptions about things. We all need to make assumptions about things so we can wake up and live our lives, or else we’d break. But it’s also important to understand your assumptions about things are wrong. Nobody has ever had assumptions about the world that have been right and they never will because the world is too complex and we’re asked to think about too many things. And so, it’s ok to have flawed assumptions and get through your life using those, but maybe when you have a moment, when you’re feeling extra productive, or you have a long weekend or something like that, take those chances to examine your assumptions and maybe work out how one of those isn’t serving you as well as it could be and maybe replace it with something that could work for you better.
Finally, is there a particular card, relic or enemy that you really irrationally hate?
As a streamer I talk to everyone, whereas a viewer sits in a space where they talk to a few of their friends, or post on a particular subreddit, so they may identify a meme that they identify with me as the only thing, but I get the claw meme, the boot meme, the catalyst meme, the demon form meme. I’ve never hated anything in Slay the Spire, but there are definitely things which are a lot worse than many people think they are, and explaining why that is can be very difficult because they come to me with an understanding that I am very good at the game and that I can teach them something, but also an understanding of what they think is good to begin with. Sometimes if I am trying to explain “hey no, that’s not very good” it’s hard to do that in a way that isn’t threatening or combative or feeling that it is stripping away what they understand about the game, so I have to try and work out how to do that in a way that is constructive or helps them grow and gives them something to stand on instead of just taking away the thing they are currently standing on. Which is why I make 15 minute videos for answering these questions instead of trying to do it in 10 seconds every time in chat. Yeah but obviously I hate the shovel. (That’s me memeing, I love the shovel)
Currently, Stephen has over 400 victories against the Heart on Ascension 20. He has finally managed to work out Defect, to a point where he’s winning around 30% of his runs with it, close to the 35% with the other characters. His streams are regular, and full of warmth and lots of interesting stray observations on the game, and decision making in general. If you are brand new to Slay the Spire, a struggling climber, or a total veteran, he has a plethora of content covering almost every aspect of the game, which is totally befitting of its endlessly spiraling and complex depths. If you follow some of his advice, you may come to realise that being dealt bad cards isn’t necessarily the worst thing that can happen if you start taking apart your base assumptions, and analysing what’s really on the table.
He’s also started working on his own modded characters for the game to try to bring more attention to the decisions he thinks are interesting in it. You can check them out on his Discord.
