Skills and the continuum of player versus character agency


Gary Gygax famously relates the inception of skills into D&D as a result of having a character cross a river and someone asking "Can my character swim?" Since then skills began to creep into D&D and other RPG's at a furious pace. Eventually, most games added some skills or even made skill-based systems -- dispensing with classes altogether. Most gamers accept the premise of skills blithely, seeing them as another thing that their character can do and fail to recognize that some skills take away player agency from the table.

Player agency is the ability of the player to direct and control the outcome of the game by interacting with the narrative constructs of the game. It is at odds with character agency -- the mechanical advantage that a character has within a scenario. I'm not in favour of either extreme. That is my bias. In early role-playing games, those games with few or no skills on the character sheet, player agency ruled the day. A clever player could run a stupid character in a way that they solved all of the puzzles, negotiated with tyrants, and wooed the NPCs. The introduction of skills like perception, investigate, persuade, and deception tilted the scales in the other direction. Players could circumvent the game by picking up the dice and rolling. Neither is truly role-playing as they are either ignoring the character or the context. It is only through GM fiat in either case that pushes the players back to the middle-ground between pure gaming and pure narrativism. I abhor such heavy-handedness on the part of the GM, while others see it as part of the GM's sacred duty.

So what is a skill? In D&D these terms get broadly used to include knowledge, tactics, and actions. I think that giving a skill too broad of a definition is a problem. Survival, for instance, is predicated on the rule of three: three minutes without air, three hours of exposure, three days without water, three weeks without food. Knowing the priorities of survival is essential to the skill of surviving. These are the tactics -- the hierarchy of goals. Knowledge is knowing what plants have water, what food will make you sick, etc. The actions that you have are -- pitching a tent, tying a rope, extracting the good parts of the plant, etc. If your skill broadly encompasses all of these things what is left for the player to do but roll the dice?

Let's examine Persuasion in D&D:

"Persuasion. When you attempt to influence someone or a group of people with tact, social graces, or good nature, the GM might ask you to make a Charisma (Persuasion) check. Typically, you use persuasion when acting in good faith, to foster friendships, make cordial requests, or exhibit proper etiquette. Examples of persuading others include convincing a chamberlain to let your party see the king, negotiating peace between warring tribes, or inspiring a crowd of townsfolk."

By mixing the skill definition with the tactic -- how to inspire a crowd, you have taken away part of the game. There is no need for the party to determine what the motives of the townsfolk are, to figure out what would inspire them. This was included in the skill description. Role-playing gets closed down.

Here is where I'd like some feedback...

What if someone skilled at persuasion meant that you deliver your argument convincingly? One could argue that is just like being able to carry a lot of weight per trip. Choosing what to carry is the realm of the player; just as choosing what to say should be the realm of the player. If the GM gave the player with a skilled orator character more sentences to deliver the message; they are in effect allowing them to "carry more weight." For example, a crude speaker tries to convince a crowd of something. The crowd will turn away or tune out unless the first sentence can captivate them. The skilled orator gets three sentences before the crowd turns away. In both cases what the player says is crucial. If the speech is protracted, the great orator can move them along better. At each step of the "speech" the player must hook the audience with something that keeps them going. The orator gets three sentences to find a hook; the lout gets one.

This methodology extrapolates to all things. A character who is perceptive gets three-sentence descriptions; one who isn't gets one.

At first, it seems a bit crunchy to break up a skill-test like this. The skilled character doesn't get more dice or pluses -- they can't roll-play their way through the scenario. However, they do get an advantage and have control of the scenario. What are your thoughts?

You raise some interesting ideas.

I can certainly agree with your remarks on the broadness of definition given to skills leading to problems. With Knowledge skills we have a way of specialising that involves characters reading books that have a narrow focus and gaining a skill check circumstance modifier in appropriate situations.

Digression alert: I set the DCs higher on Knowledge skills as they are a bit ridiculously low in the RAW. Especially for 'monster knowledge'. The RAW seem to assume shared universal knowledge exists everywhere about every creature in the world. I suppose this goes hand in hand with my irritation over the Universal Gold Standard Currency that seems to extend even into places designated as 'where your soul goes when you die'; and the Common language that seems to exist everywhere in the multiverse as well. "I summon a demon, and pay him for his services in gold pieces." Uh, no.

The Persuasion or Diplomacy skill is, I think, the one that most often gives rise to roll-play vs role-play arguments because persuading NPCs is so pivotal to the way that the plot develops in adventures that are not mere unavoidable hack-fests; or indeed the way that the wider campaign develops if the PCs start to interact with the political elite of their world.

In our D&D 3.5 campaign, I often find myself playing Persuader-type characters, and I almost never dump-stat Charisma. One of my characters has such high Diplomacy skill, that I re-wrote the Diplomacy system (a frequent occurrence, I gather) to make it harder for them to have all and sundry eating out of their hand. They can still usually sway 'typical' members of the populace to their point of view even with a rushed check at -10, but they can no longer walk into Asmodeus' palace and persuade him to hand over his scepter.

I've given time to considering why I am drawn to playing this type of character and am led to the conclusion that it's because I am so bad at spontaneous verbal dialogue in real life (though pretty good, though I say it myself, at public speaking as long as I have had time to prepare). Thinking about how your system would work for me as a player, my character would get more sentences in which to make their case and that works fine if the GM is liking the way I am putting my argument across, but if he or she isn't liking my style of delivery I'll actually be digging myself deeper as I blurt out my fumbling sentences. Then again I can see how it prevents unskilled characters from the opportunity of making uncharacteristically sophisticated arguments, so I see the merits as well.

In our game we use the (possibly abhorrent!) time-honoured "give a modifier to the skill check based on how well the player put their argument across". This still suffers from issues of inflicting a player's poor Charisma on the character they are playing to escape from the suckitude of real life, but at least their character's high modifier can compensate.

There are some who say that players who are not good at things, should not play characters who are not good at those things because they will play the role poorly. This is pretty harsh, but in the end I suppose it is a style choice for the referee to make and part of their contract with their players; if they choose to play by it.

I quite recently encountered this variant on the d20 Diplomacy system and it seems to have some good ideas in it. There are some similar ideas to our own house system though the rules I wrote are more about changing attitudes than persuading individuals to take a specific course of action. It doesn't really address your concerns as this system is still pure mechanical crunch and just a different way of resolving use of a skill modifier; but I found it worth a look.

I'll also note the divide between persuasion of NPCs, and persuasion of other PCs played by other players. Nowhere is the divide between player agency and character agency throw into such sharp relief. No matter how good or bad a Persuader your character is, your gaming buddies are going to treat your character's efforts to persuade their characters in exactly the same way! Usually, anyway. I have, on rare occasions, seen a player say "I'm not sure what my character would do here; let's use the mechanical system to resolve what they do." This can absolve the player of responsibility for their character's actions and some players find this a useful fallback.

"In our game we use the (possibly abhorrent!) time-honoured "give a modifier to the skill check based on how well the player put their argument across"."
I actually like that method. It addresses the problem using the mechanism inherent in the system. It may not divide the skill check in the way that I like, but we have used this method a lot and have found it enjoyable.

I'm an advocate of separating the delivery from the content in skill checks (or, as I phrased it above the tactics from the skill). For example, "I strike from his blind side" is a tactic. "I cut in an arc, allowing my bottom hand to drive the power and my top hand to keep the edge alignment." is describing the skill. If a GM judges how well the player delivers the speech they are judging skill, which I disagree with. They should only consider what the player says and substitute the "how the character would say it" when considering the outcome. This should also be applied to knowledge -- the player describes the intent; while the character drives the execution. This prevents the creation of gunpowder by players.

The GURPS skill system as of 2008 does it differently. ( ). Yes it does take a little more more on the GM's part before hand but it prevents player form having to rattle off tons of dialogue to do a task.