I’ve been pretty interested in what Jay Barnson has been writing about over the last few days. In his latest “RPG Design” series, he has been discussing the pros and cons of using a Skill-based system over that of a Class-based one — although, he’s mostly been looking at Class-based so far. He picked up my interest for two primary reasons: 1) I’ve had a similar internal battle over trying to connect them to Minty Breath and 2) I think there might be an easier way to think about this problem: think of them as symbols.

[Before you read anything else, I want you to be reminded of one thing: I have yet to finish making a game. I’ve started a few in the past, but have yet to ever get past playing around with prototypes for things. Jay, as far as I can tell, has done that. So, while I might mildly disagree with something he wrote, you might want to think about it before siding with me. He has the experience, I just happen to be able to ramble more with words.]

In considering how to approach the problem of whether or not to have skills — or classes — in Minty Breath, I struggled with trying to relate them to language constructs. For me, I think better with words, with sentence structure and within grammar rules. You may have noticed, within my own writing, that I am very particular about how I use certain syntax rules and how few mistakes I make — although, yes, as was pointed out to me last week, they do happen. I like constraints when I create. I like having having borders to work within and against when I work. They help me lower the scope of a problem. So, I went about trying to link all design ideas into syntax descriptions.

For Minty Breath, I decided that I would use skills. This came, in part, from how Line was already talking about her design — I borrowed, with permission, many of her original ideas in my initial design. However, skills presented a problem. How do they fit into the idea of making tactical decisions within a conversational model? If you were communicating with another object — entity or person — how would a skill manifest itself? Would you just possess some temporary increase (passive) or would they require some type of enabling (active)? I have ideas on both, but am still working on it. However, I did come to a conclusion: skills should be verbs.

In my model, skills are something you can do, an Action. Therefore, they are verbs. In a syntactically correct usage, an object would use an action in a context to another object. In other words, “Sam sits on [the] chair.” Sam is the subject. He performs the action of sitting and the direct object is the chair. In translating the in-game events into a visible language, I could construct (some fraction) of the fabula as it happened and also present the player with a written record of events. There would be the story as the player acted and then a narrated version that could then, hopefully, be fed back into the engine to reconstruct the same actions that just happened — that objective, to read and translate the ‘game language’, is a long, long way off though.

My point in saying all that is to make the statement that using skills gives greater choice to the player. They can pick the expressions that they want their character to use in addressing the world. Jay mentions that point too:

“In a skill-based game system, your character progresses by individually increasing or acquiring each different skill and ability individually, without a predetermined class framework to guide or limit the choices. Oftentimes, there isn’t even a concept of “level” – the improvements are bought directly with experience points… The Fallout series are solid skill-based RPGs, and Skyrim, the latest of the Elder Scrolls games, has moved almost entirely to the skill-based camp.” (emphasis added)

Bethesda developed games, as Ari and I were discussing in the comments the other day, are about giving the player the most options, the most freedom. They want the experience to be designed in such a way that the experience is tailored toward how the player chooses to express their individuality through their characters. Using skills, allowing the player to customize which abilities — in this case, what the character is able to do — they have lets the character fit their needs to the occasion.

Skills, then, are like clothing or any other outward-facing cultural choice: we choose because first it speaks to us and second because it speaks to others. In games like Oblivion and, as I hear, Skyrim, player use skills, pick skills, in such a way that they are constructing their own personal set of symbols for the world to see and for how they communicate with others (mostly enemies). It is not unlike how most people follow certain paths or perform certain acts; we build little rituals in our lives that reflect moments that are important to us. The skill set of each player is different, it is how they play the game. Their personal collection of skills, of verbs, is how they directly interact with other in-game objects.

Classes, on the other hand, are easier for developers to create. Basically, they are archetypes. They represent, in short-hand, some set of ideas that surround some person or label. If you, as a developer, want to jump-start the character creation process, you give players classes to choose from for the game. They can then learn how you, as the developer, think about certain ideas and thus adjust themselves to the world they are about to interact with and perhaps, hopefully, spend many hours enjoying.

Jay points out the same thing:

“Simplicity and Ease of Introduction: Class-based games break different styles of gameplay into simplified roles that are easier for newcomers (to that particular game system, or to RPGs in general). Particularly during initial character creation, it can be really hard to predict what skills or styles of play will be viable, and a smaller palette of roles can ease the guesswork a bit.” (original emphasis)

In his second post, Jay spends more time on the problems with Class-based systems and speaks to something that I want to highlight as a problem, but not in the way he thinks it originates.

“Unbalanced Content: Some content may be far too difficult or far too easy depending upon which classes are absent or available. An example would be the original Bard’s Tale, as far as I remember it correctly: If you had a bard in the party and the horn from the default starting party, surviving to level three or four was relatively possible. Without a bard and the magical horn, the early stages of the game were absolutely brutal. Another example is the cleric class in D&D (up through 3rd edition, not including Pathfinder) and undead encounters. The difference between an encounter being trivially simple and brutally devastating might be the presence of a cleric. One can argue (and I often do) that this simply means the class abilities are imbalanced with such an all-or-nothing difference in the experience, but the point of the specialization really is to make the character truly shine in particular circumstances, not to just provide a marginal advantage. It’s a fine line…” (original emphasis)

While it might be true that the content in said examples are actually unbalanced — I’ve never used them, can’t speak to that — I would like to shift the blame away from what is not to what is. Something I personally struggle with, in my own world creation, is how much exposition to give and when to give it to the reader. In this case, I would like to think that the games were not as unbalanced as he says but that the designers had different ideas in mind when they created the classes that they wanted people to play with in their games. They did a poor job of explaining how their worlds worked and had a number of expectations about how people would play, which roles they would us and how those expressions between players would work.

That’s the problem with symbols. Their meanings are all relative. Most survive only through constant maintenance and usage to reinforce what their associated meanings are. What I think might have happened in at least one of the cases that Jay mentioned is that when the designer used the archetype of cleric, they assumed that no party would not be without one. They assumed that someone would be playing in the role of healer — yes, I know, not all clerics are healers — and did not think outside of that archetype. It made it easier to design situations and powers with the thought that their symbols, the archetypes they chose to represent the limited choices of expression allowed for the players, would be used in certain ways and within certain contexts.

It comes back to, at least for me, in how you are explaining your game mechanics to players. If you can explain to them why you are limiting their expressions in a reasonable way, they will believe you.

To give an example, Trine has only three classes but allows the player to switch between them. Each class has a very limited number of expressions. Yet, the game is able to construct very elaborate puzzles for the player because they do a very good job of explaining, within the first few levels, how each class works and why they are better at certain actions. Once the player learns the game’s symbols, how the developers see each class, the player can arrange their own expressions using such symbols to create the perfect (or approximate) input needed for each puzzle.

Be it skills or classes, each are symbols for how the designer is explaining to the player the range of expressions they are allowed in the world, which verbs they can use. Skills, as I prefer, allow a greater range of customization for the player but leave the designer with planning for complex chains of verb+noun combinations with various objects in the world. Context becomes the key to how things are used. Classes, as I mentioned, work great when expressions can be limited to increase the complexity of the mechanics of the game. By decreasing the possible verbs, the number of interactions can be better managed by both player and designer. The designer can create a complex chain of events and know that the player will (hopefully) be able to translate it and then respond with the symbols they have learned.