Solving the Spire - Better living through Card Games

“I will never solve the Spire, there’s too much of it.”

Stephen “Jorbs” Flavell has played over 2500 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, 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 USP 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 nuance 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 Spore in order to start getting better very quickly is to recognise that you don’t build a deck which in 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 same thing when it’s being asked to do all sorts of things?”

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? Even if you’re not doing things perfectly you can still get to the. At low difficulty you get far less feedback telling that you’ve done something that wasn’t good as it could have been. The game doesn’t violently punish you enough for not making the absolute perfect decision so you don’t learn from your mistakes. Once you reach Ascension 20 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 somewhat - 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 also thinking about other things such as, how I am living my life, how I am interacting with other people.
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. Except, instead of waking up every single morning and being the same person, you wake up and you’re something completely new. You have to work out your deck over the course of 50 floors. It’s very hard to take a world that you’ve never been in before, in terms of the same cards and relics, 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 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 a 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 realise there is a taller mountain somewhere else, and everything you’ve been practicing so far, every strategy you believed 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?

For The 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 many 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 she can be at turn one and two. I liked playing combo decks so I tried to build only Silent decks that won 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. On Ascension 20 you need have the explosiveness that a turn one kill deck would have, but you can’t build just that. You need your scaling, but you need cards like Calculated Gamble and Acrobatics, and relics like Bag of Preparation to make you explosive.
The way I played defect before Ascension 20 was quite passive - scaling with orbs and getting a lot of block - damage was always an afterthought. 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. 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.

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 it 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 that is “this is the correct way to play the game”. We can create an artificial brain which can demonstrably do it better than we can, but the thing we 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. We can build some neural network or machine that uses a technique to understand the game which lets it understand the game better than we do and then we can ask the machine what to. Machines aren’t very communicative however, so sometimes the explanation it gives us isn’t very satisfying, and also they're imperfect too so sometimes they make the wrong decision.

One thing I like about Slay the Spire over poker is that there isn’t a machine that can play it 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.


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.

In 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?”

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 conceit, 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 StS serves a culture of people who have played roguelikes and card games growing up, and so some people come to StS and know more than others. Games do 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.

So, using Slay the Spire as a kind of a leveller. Is there one main concept from StS 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 us that for making decisions at each intersection. You can’t just aim for that goal and turn 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.


---

Currently, Stephen has over 200 victories against the Heart on Ascension 20. 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 spiralling 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.