Allowing flavour to guide direction

Traditionally known as “top-down” design, the idea of starting with the fantasy behind the idea and translating that into mechanics is as old as games themselves. For Balance, I want very strong themes intertwined with the mechanics on a deep level, in order to create a very flavourful experience, much like Blades in the Dark does. Even though I wanted to approach everything from a flavour standpoint, that wasn’t realistically feasible, and the previous two articles about Warrior’s and Hunter’s design showcased that they were fulfilling the needs of the game rather than the flavour. When I got to making the first INT-based class, for reasons outlined in detail in the Warrior article, I knew I had to have a wizard – the traditional fireball slinging spell caster that’s so ingrained in the genre. Knowing full well I’d put that into the game, I thought that I wouldn’t worry about it for the time being and only figure out where they fit into the whole system, not how. Before I had the final version of the class structure the game currently operates on, I set out to establish the core fantasy of the system – playing Agents of Twilight, operating under the will of the Eternal Council. It’s a bit of Chimborazo lore taken to the extreme, where I can focus on and explore it in an alternate universe, free from the trappings of the original world – for instance, it in the members of the Order of Twilight were only Monks. Not knowing the details of how I was going to do the classes, I still knew there were going to be classes, and I didn’t want to stay confined to variations of Monks. Enter Witches.

Despite being an alternate universe take on the concept, I decided to stay as true as possible to the flavour of the archetypes I wanted to put into the game. If nothing else, it will be a nice bridge between Balance and Chimborazo. With this in mind – what are witches? Or, specifically, what are the themes that I want to convey with them? Well, their original “powers” are very loosely defined, much more in line with what a soft magic system has to offer, but they can be put into roughly three categories – arcane magic, blood magic, and making potions. Which actually works out quite well, because in Balance each class has access to three Paths on their skill tree. And this is where the problems begin.

While technically three, in truth classes so far really had two Paths – a light and a dark one. The twilight path was reserved for stat boosts where other paths’ skill would be, the flavour behind it being that going down this Path, the character spends their time drilling on what they already do well and improving on their core identity, a.k.a. what they are already good at. This wasn’t going to work very well for the witches, and I started asking some brave questions about the nature of the system I had built. “Would it really be such a big deal if one class didn’t follow the template?” It seemed like a big deal at the time, but I’ve always believed that a good designer has to know which parts of the system are sacred and which rules can be broken to create a better experience. So I started experimenting with having all three of the Witch Paths be full of skills. After a few preliminary play tests I was really happy with how that turned out – due to their nature of “soft magic”, individual skills don’t have a lot of raw power, but the larger toolbox allowed the Witch player be impactful in more ways. It fit the theme of “support class” that I wanted for them, so I really liked it.

Once this experiment was deemed more or less successful, I began wondering how I felt about that one class being mechanically different from the rest. I figured I wouldn’t mind it so much if there were three classes that sort of “broke the mould”, because triplets are one of the major sub-themes of the game – even the classes are separated into three groups (with a different main stat) plus Monk. The Monk has always been a special case, partly because he’s the original member of the Order. So instead of exploring which other two classes to “fit” under this three-path system (quite unnecessarily so, because I didn’t deem it necessary originally), I wanted to see where it made sense. And it turns out that was the Monk. They were always supposed to represent “perfect balance”, in both their themes and mechanics, and they didn’t have a main stat – or rather, they treated all stats as main. It made a lot of sense to apply the Witch approach to the Monk class, and it worked out great in the end, I was very happy with having translated the flavour so well, even if it meant having two that slightly veered off from the template.

Around that time I realised a major flaw in this system – these two classes no longer got access to built-in stat increases if that’s what the player wanted to do. I saw two solutions to that problem – either giving them a “free” stat increase every level, or denying them natural stat increases altogether. I deliberated on that a lot, but ultimate I went with a split approach, where I gave the Monk a free stat (of their choice!) upon level up, while completely robbing the witch from being able to increase her stats through the progression system. It just made more sense – a Monk is always training, constantly honing their body, while a Witch spends so much time preparing and carried away with immaterial stuff that she kind of foregoes that part entirely. In a way, this dichotomy turned the Monk and the Witch into opposites of each other, which I think is a cool theme on its own.
(Quick note on power level – none of this means that one class is inherently more powerful than another. With enough time to prepare beforehand and due to the “soft magic” nature of her abilities, the witch can actually be the most powerful class when given that opportunity. Conversely, though their base is stronger, the Monk’s abilities are a lot more linear than those of other classes.)

All of this bending of the system was in order to allow myself to better represent the Witch’s identity and flavour. However, mechanical shifts alone are not enough to convey very deep meaning – for all of this to have a point to it, I need to craft strong themes that resonate throughout her skill kit, so players can feel like they are actually playing a witch. In writing, an approach that’s very often used to make the reader subconsciously connect different things together is repetition. Specifically, the repetition of common elements. I was looking for some way to tie all of the skills on a given path together, when I found inspiration while making a status effect – bewitched. As a word, “bewitch” exists in a group of now obsolete transitive verbs, alongside the likes of befall, beget, betwynde, etc. It was an interesting idea, using quite outdated and sometimes completely obsolete words, but one that I think quite befits the nature of witches. It wasn’t long before I pulled a list of all words with the be- prefix and went through it, picking out a couple dozen of the more thematically fitting ones – I was then tied to what the actual skills with these names could do based on the word’s meaning, and in many ways my limited “dictionary” defined what the Witch will do on this path.
(More on strong identities in the next article)

Similarly, while I was playing MTG later on, I cast Gifts Ungiven, and by the time it finished resolving I already knew I had an inspiration for the class that is supposed to be able to divine information about the future and the past – and the more uncertain, the better. It’s what lead to the creation of skills like Sights Unseen, Words Unspoken, etc. The only remaining problem to solve now was to distribute the things witches do into the appropriate paths – which was tricky. From a popular culture point of view, witches are often seen as on the darker side of things, even if not outright evil. In the lore I’m pulling from, however, they are very much not that, even if they do play with some powers that can hardly be described as “good”. In every sense of the way, neither the light nor the dark path offer a clear answer to “What is a Witch on this path?” This meant that I had some flexibility in which pillar of gameplay I can explore with those paths, but it also meant that I had to artificially retrofit the themes to match the type of gameplay I settled on this way.

Fortunately, due to existence of the Wizard, I didn’t have too much freedom. With STR as their secondary stat, they had to be the INT-based class that does Combat, and I very much didn’t want witches to be good at combat, so one out of four was immediately unavailable. Due to them being perceived as dark, I didn’t want them to be very good at Social stuff either, so it seemed like the options for their Paths were to explore Tactics and Stealth gameplay. It fit their primary and secondary stats of INT and AGI respectively, so at least I was happy with that decision from a systematic point of view. And while quite comfortable with Tactics, as the class that needs time to set up and uses a bigger toolbox of skills, them having Stealth gameplay was something I wasn’t prepared for from the start, not in the slightest. I had to entirely re-define what Stealth meant when a Witch did it, and ultimate I set down on an unusual concept – enabling Stealth through information. That also fit their theme as a support class quite nicely.

Lastly, picking which type of gameplay to go into which Path was ultimately guided by flavour, too. I knew that I wanted the Witch to be the only actual “healer” in the game, and I saw the opportunity to have this cool thing where the only healer is on a dark Path, so I seized it and put blood magic there. I had no choice but to put arcane in the Twilight Path, because it made the most sense lore-wise by a long shot, so that left potion making for her light Path.

In the end, everything about the Witch’s design was guided by flavour. So much so that I changed entire parts of the game in order to accommodate it, which in turn allowed me to do more cool stuff with other classes and explore space I otherwise wouldn’t have. I think it was for the better.

For additional context, refer to the Balance class design intro

<< Previous article: Hunter | Next article: Knight >>

Leave a Reply

Your email address will not be published. Required fields are marked *