Game Mechanics And Player Agency

The concept of player agency is a central pillar of all role-playing games. It is a balancing factor against the omnipotent, omniscient Game Master. For the purposes of this article, we will be focusing on the smaller-scale application of player agency and the role of game mechanics that negate or modify such agency.

View attachment 95347​

From the very first iteration of Dungeons & Dragons in 1974, there have been mechanics in place in RPGs to force certain decisions upon players. A classic D&D example is the charm person spell, which allows the spell caster to bring someone under their control and command. (The 1983 D&D Basic Set even includes such a possible outcome in its very first tutorial adventure, in which your hapless Fighter may fall under the sway of Bargle and "decide" to let the outlaw magic-user go free even after murdering your friend Aleena!)

It didn't take long for other RPGs to start experimenting with even greater mechanical methods of limiting player agency. Call of Cthulhu (1981) introduced the Sanity mechanic as a way of tracking the player-characters' mental stress and degeneration in the face of mind-blasting horrors. But the Temporary Insanity rules also dictated that PCs exposed to particularly nasty shocks were no longer necessarily in control of their own actions. The current edition of the game even gives the Call of Cthulhu GM carte blanche to dictate the hapless investigator's fate, having the PC come to their senses hours later having been robbed, beaten, or even institutionalized!

King Arthur Pendragon debuted in 1985 featuring even more radical behavioral mechanics. The game's system of Traits and Passions perfectly mirrors the Arthurian tales, in which normally sensible and virtuous knights and ladies with everything to lose risk it all in the name of love, hatred, vengeance, or petty jealousy. So too are the player-knights of the game driven to foolhardy heroism or destructive madness, quite often against the players' wishes. Indeed, suffering a bout of madness in Pendragon is enough to put a player-knight out of the game sometimes for (quite literally) many game-years on end…and if the player-knight does return, they are apt to have undergone significant trauma reflected in altered statistics.

The legacies of Call of Cthulhu and King Arthur Pendragon have influenced numerous other game designs down to this day, and although the charm person spell is not nearly as all-powerful as it was when first introduced in 1974 ("If the spell is successful it will cause the charmed entity to come completely under the influence of the Magic-User until such time as the 'charm' is dispelled[.]"), it and many other mind-affecting spells and items continue to bedevil D&D adventurers of all types.

Infringing on player agency calls for great care in any circumstance. As alluded to at the top of this article, GMs already have so much power in the game, that to appear to take any away from the players is bound to rankle. This is likely why games developed mechanical means to allow GMs to do so in order to make for a more interesting story without appearing biased or arbitrary. Most players, after all, would refuse to voluntarily submit to the will of an evil wizard, to faint or flee screaming in the presence of cosmic horror, or to attack an ally or lover in a blind rage. Yet these moments are often the most memorable of a campaign, and they are facilitated by behavioral mechanics.

What do you think? What's your personal "red line" for behavioral mechanics? Do behavioral mechanics have any place in RPGs, and if so, to what extent? Most crucially: do they enhance narrative or detract from it?

contributed by David Larkins

 

[Ovinomancer]

I guess I missed it somewhere: how would you phrase answers to an action like "can I tell if he's lying?" (let alone "do I think he's lying?" which I suspect, would earn a "IDK, do you?") or whatever, to both avoid telling the player what the PC thinking, and convey anything like useful information? I mean, objective: determine if the NPC is being truthful method: I pay careful attention to his body language, looking for any clues that might indicate deceit, perhaps.

How'd you narrate the resolution if you thought it was uncertain, and decided to roll..?


As a bonus: how would you determine & narrate a failure giving wrong information vs no information?

Quoting you, Tony, instead of a later post on this discussion so that I can thank you for providing the impetus that gelled my thinking on this. So, thanks!

ETA: First, and because I failed to make this initially clear in the below, I'm discussing social skills here. You'll see where I realized this. Also, this discussion is D&D centric. (I think Story Now and Narrativist games don't have this issue because each check is made only by players and stakes for both success and failure are established, so a limitation on future actions is entirely permissible in those systems and doesn't violate agency because the player has to stake that outcome.)

Here's my thinking -- we use the phrasing 'you think he's lying' because it's a handy shorthand for imparting the kind of otherwise dense and hard to articulate information necessary to provide the input. So we skip the actual facts -- the looking left, the perspiration, a tightening of the irises, etc -- because it may not convey the proper information to the player (the player may misunderstand or the GM may not be able to convey in a way the player would understand, etc.) and instead use the shorthand to skip ahead to the point we're trying to make, specifically in this case that the wealth of information your character is receiving and the character's ability to perceive it indicates lying is going on. In short, the 'you think' is a shortcut to clarity because trying to describe the actual information the character is receiving is long and may be misunderstood in the telling.

The problem we're discussing occurs right after this. On one side, there's the camp (Camp 1) that believe that this information is just that - informational - and that the player has full authority to declare any actions for her PC that she wants. The "you think" statement is taken solely in an informative sense, a fact or bit of knowledge that they can use or ignore as they wish.

On the other side, there's the camp (Camp 2) that thinks this information is limiting on the player's ability to declare what their PC does. That the player is not free to violate this established fact and instead must or should adhere to it.

(There's also a third camp, which says that the information should strongly bias the player's declarations, but I'm not sure this really differs from Camp 1 or not, because the end point is that the player could still override that bias, they just should be do so judiciously or rarely.)

On both sides, this really is irrelevant as to who's rolling - the player or the NPC. However, for Camp 1, NPCs rolling checks against PCs tends to be viewed as irrelevant or unwanted. This is because the player can still do whatever they want, so the die roll is largely meaningless in regards to player decisions. Therefore, Camp 1 tends to adopt playstyles where NPCs don't initiate rolls against players but instead use their skills as challenge difficult benchmarks against player declared actions. Camp 2, however, seeing the information imparted by the rolls as binding, sees NPC initiated rolls as just another method for rolls to bind players and so doesn't draw a distinction between NPC initiated or player initiated rolls. But, the real core difference here isn't if NPCs checks can bind PCs, but how the results of a check are viewed -- is the result of a check informational or binding?

Clearly, myself and [MENTION=97077]iserith[/MENTION], [MENTION=3400]billd91[/MENTION], and other are in Camp 1 -- checks are informational. Tony's and others are Camp 2. One camp or the other aren't better, but this explains the core philosophical issue that divides this discussion (I believe, at least).

So, looking to other areas of the game beside social checks, does this continue to play out? Well, we'll have to divide checks into two categories: informational checks (which I'm discussing above) and those checks used to accomplish a task (like lockpicking). As for what constitutes the difference between a task resolution and an informational check, I going with whether or not you'd describe the result as something the PC knows or thinks is informational, if you instead describe a change to something outside of the PC that's task resolution.

Firstly, for task resolution checks, I think both Camps engage the game the same way -- a success means the task is accomplished. There are other considerations for failed checks that I'm not going to go into in this post (fail forward, degrees of failure, etc.). But, I think both camps see task resolution checks the same way. Camp 1 isn't going to treat a success on a lockpicking check or a check to force a door open as informational, for instance.

For informational checks, though, I can see a few problem areas, one huge one being illusions. Illusions muck with the information presented to the players, but many tables allow for a PC to act as if something described as present is an illusion, ie they allow the player to declare actions against what their PC observes. This touches on the informational check Camps in that a check may be made and result in false information being provided due to an illusion. I still believe there are tables that are strong Camp 2 here and that will require the PC act as if the illusion described is real even if the player has doubts, but I also think there are a lot of Camp 2 on social checks GMs out there that are more Camp 1 in regards to illusions. At least, I've played at such tables where you had to go along with the social check results but could, at any point, declare an attempt to disbelieve something as an illusion or act as if it were an illusion (this often ended poorly, iirc).

I'd be interested in hearing from many of the posters in this thread about this, from both sides of the discussion. I don't think it shows any incoherence to be strong Camp 2 on social skills and weak Camp 2 or even Camp 1 on illusions -- this isn't a trap question, it's a curiosity.

Aside: I've been quite, um, vigorous in my defense, err, offense(?) on abridging player agency. Honestly, I think this was because I didn't have a solid understanding of the real divide in thinking, and had an extra dollop of righteousness over my own conversion (ever meet an ex-smoker who's almost violently anti-smoking?). I find now that I can see a clear divide in the discussion and it's as simple as adopting the 'you think' as binding versus informational, I'm less, well, adamant about the issue. Well, less adamant that I have the right opinion, not less adamant about my opinion, at least. I suppose this is a beating-around-the-bush apology for some of my more vigorous statements from earlier in the thread.

 

[iserith]

Here's my thinking -- we use the phrasing 'you think he's lying' because it's a handy shorthand for imparting the kind of otherwise dense and hard to articulate information necessary to provide the input. So we skip the actual facts -- the looking left, the perspiration, a tightening of the irises, etc -- because it may not convey the proper information to the player (the player may misunderstand or the GM may not be able to convey in a way the player would understand, etc.) and instead use the shorthand to skip ahead to the point we're trying to make, specifically in this case that the wealth of information your character is receiving and the character's ability to perceive it indicates lying is going on. In short, the 'you think' is a shortcut to clarity because trying to describe the actual information the character is receiving is long and may be misunderstood in the telling.

The problem we're discussing occurs right after this. On one side, there's the camp (Camp 1) that believe that this information is just that - informational - and that the player has full authority to declare any actions for her PC that she wants. The "you think" statement is taken solely in an informative sense, a fact or bit of knowledge that they can use or ignore as they wish.

I think this helps explain the divide. I will just add for my part, I avoid saying "You think..." when narrating any outcome, just like I don't describe what a PC does physically. (In fact, I endeavor not to start ANY narration with "You.") That's the player's role in my view. I'm a stickler for proper action declaration and if a player is trying to have his or her character figure out if an NPC is lying, then I want the player to declare the reasonably specific clues he or she is looking for that would cause his or her character to think the NPC is lying. The result of the check, if there is one, will determine whether the character notices those clues. Essentially, the player is saying "My character will think X if he or she notices Y." Then I decide if the character notices Y or the dice do.

In my opinion, a lot of issues in the game come from inadequate offers from the player with regard to describing what they want to do. Fix that and a lot of other issues downstream are sorted out. The player has a right to have agency in the game, but like many rights, that comes with a responsibility: to be reasonably specific when it comes to action declarations.

And, of course, my position is based solely on D&D 5e here. If the game said it was fine for the DM to declare what the PCs think, then I wouldn't have this position.

 

[Tony Vargas]

Quoting you, Tony, instead of a later post on this discussion so that I can thank you for providing the impetus that gelled my thinking on this. So, thanks!

You're welcome, I guess, but you did not actually answer either question, at all.

Here's my thinking -- we use the phrasing 'you think he's lying' because it's a handy shorthand for imparting the kind of otherwise dense and hard to articulate information necessary to provide the input. So we skip the actual facts -- the looking left, the perspiration, a tightening of the irises, etc -- because it may not convey the proper information to the player (the player may misunderstand or the GM may not be able to convey in a way the player would understand, etc.) and instead use the shorthand to skip ahead to the point we're trying to make, specifically in this case that the wealth of information your character is receiving and the character's ability to perceive it indicates lying is going on. In short, the 'you think' is a shortcut to clarity because trying to describe the actual information the character is receiving is long and may be misunderstood in the telling.

It's necessarily the character drawing a conclusion based on it's ability to spot and interpret cues/clues indicating deception. Literally what he thinks.

The problem we're discussing occurs right after this. On one side, there's the camp (Camp 1) that believe that this information is just that - informational - and that the player has full authority to declare any actions for her PC that she wants. The "you think" statement is taken solely in an informative sense, a fact or bit of knowledge that they can use or ignore as they wish.

Actually, I'm pretty sure there's no other camp. Afterall you can choose quite a lot of actions when you think someone is lying - including to second guess yourself and decide you must be wrong and he's telling the truth (maybe because the lie is so consistent with the facts you do have, and no alternate theory makes any sense, perhaps because you just /really/ want to accept the lie because the alternative is hard to deal with in some way, be it logistical or emotional).
Alternately, you could just be prone never to trust anyone, which gets very inconvenient, IMX.

Tony's and others are Camp 2.

So, my actual camp divides the issue in an entirely different way: do you resolve tasks based on the abilities of the character, or those of the player?

If you're resolving certain tasks based on the abilities of the player, then the character needs no skill or attributes in those areas. If the player controls what the character thinks/feels, then there's no point having morale checks, charm spells, social skills, reaction rolls, INT, WIS, or CHA. (It's worth noting, BTW, that when 1e AD&D had morale checks, PCs were exempt from them, and it didn't have much in the way of 'social skills,' so the game has long come down - partially - on the 'player resolution' side of things). If the player controls /how/ a character searches for a secret door or picks a lock or which wire he cuts to disarm a trap, there's no need for skills in those areas. Etc...

I'm in the camp that says resolve tasks based on the abilities of the character. It allows players to play characters very different from themselves along a wider range of dimensions. Players still portray & make decisions for their character, of course, so in games with formal 'framing' (or GMs, like Iserith, who choose to impose it) I think, that heads off any issues around resolving social tasks using character abilities carrying a risk of undermining agency. In other systems where that sort of step is left vague, obviously it can cause issues.
Even in such systems, it's worth the risk to player agency, IMHO, because it gives players freedom to play characters substantively different from themselves in more ways than physical & supernatural abilities.

 

[Ovinomancer]

Huh, I guess I was optimistic about people giving the above a chance.
[MENTION=17343]Tony_V[/MENTION]argas

You first start by stating the check resolution unequivocally what the PC thinks.

You then say that the PC is free to think something different.

I can't reconcile the two statements.

Your last bit about whether skills are character based or player based is pretty far of of what I was saying. I wasn't even considering that divide (sorry, [MENTION=29398]Lanefan[/MENTION]), and my entire pay was based on character based skills -- no player skill substitutions involved.

So, good point that I forgot those that prefer players to do skill resolution in free play, but that doesn't really address what I was talking about with Camp 2. I'm fact, if there's not a check at all, the very basis for the Camps doesn't exist -- outs all premised on a character's check result providing information to the player.

 

[iserith]

I'm in the camp that says resolve tasks based on the abilities of the character. It allows players to play characters very different from themselves along a wider range of dimensions. Players still portray & make decisions for their character, of course, so in games with formal 'framing' (or GMs, like Iserith, who choose to impose it) I think, that heads off any issues around resolving social tasks using character abilities carrying a risk of undermining agency. In other systems where that sort of step is left vague, obviously it can cause issues.
Even in such systems, it's worth the risk to player agency, IMHO, because it gives players freedom to play characters substantively different from themselves in more ways than physical & supernatural abilities.

Like a true leader, I only scan posts till I see my name and then read the sentences immediately before and after it. Can you clarify what you mean here with regard to what you perceive are my methods?

 

[Tony Vargas]

Huh, I guess I was optimistic about people giving the above a chance.
You first start by stating the check resolution unequivocally what the PC thinks.
You then say that the PC is free to think something different.
I can't reconcile the two statements.

Make different decisions, anyway. Afterall, someone could be telling the truth, but have the wrong information, himself, for instance. He could be lying, but you don't want to let on that you believe that. Thinking someone is being truthful is just that, not thinking he (or you) is necessarily infallible.

At least, if the possibility of 'false positives' exist. If the resolution of a roll results in knowing he's lying or not or "you're not sure." Then, actually, you're always sure, you can never get a mistaken impression that a liar is being truthful or mistake mere nervousness for lying or whatever.

Your last bit about whether skills are character based or player based is pretty far of of what I was saying.

Yes, it is. That's why I felt compelled to spell it out, because that's the important distinction, in my view. Not whether the details of a system that does model character abilities rather than resorting to substituting player abilities, might end up abridging player agency in some instances.

That happens a lot in these discussions "risk I'm willing to take" turns into "something I demand all the time."

That's why we have a battlemaster in 5e, calls for a fighter that could actually do stuff, even if it meant complexity, became calls for a 'complex fighter' for complexity's sake.

Can you clarify what you mean here with regard to what you perceive are my methods?

I'm blanking on the words you usually use, but objective and method? Players need to tell you what they're trying to accomplish, and how. That's a form of 'framing,' I think, and avoids some pitfalls.

 

[iserith]

I'm blanking on one of the words you usually use, but objective(?) and method. Players need to tell you what they're trying to accomplish, and how. That's a form of 'framing,' I think, and avoids some pitfalls.

Ah, right. Goal and approach.

 

[Lanefan]

Huh, I guess I was optimistic about people giving the above a chance.

You first start by stating the check resolution unequivocally what the PC thinks.

You then say that the PC is free to think something different.

I can't reconcile the two statements.

Maybe the statements are using the word "think" where something else would be better? As in:

You first start by stating the check resolution unequivocally what the PC believes to be true.

You then say that the PC is free to act as if it is not true and do something different.

Howzat?

Your last bit about whether skills are character based or player based is pretty far of of what I was saying. I wasn't even considering that divide (sorry, [MENTION=29398]Lanefan[/MENTION]), and my entire pay was based on character based skills -- no player skill substitutions involved.

So, good point that I forgot those that prefer players to do skill resolution in free play, but that doesn't really address what I was talking about with Camp 2. I'm fact, if there's not a check at all, the very basis for the Camps doesn't exist -- outs all premised on a character's check result providing information to the player.

Your point about illusions caught my interest.

In my game, if a PC (or player) thinks there's something fishy about my description of a scene they are always free to state "I disbelieve" - a full-round action during which they attempt to see though an illusion. I roll in secret, and on success the illusion becomes apparent to that PC...who can then give a significant bonus to any other PCs attempting the same disbelief. This is pretty much straight out of the 1e rules and it's never been a problem.

Without a successful disbelief, however, the illusion is "real" to the PCs until and unless they interact with it in such a way as to break it (e.g. someone steps on to the illusory floor over the shaft and falls through).

I have a hard time thinking of social stuff in the same light. Illusions, for one thing, are pretty much binary - in the end you either believe them or you don't - while nearly anything social is going to be much more of a series of sliding scales and never-the-same-twice variables.

Lan-"Illusionist has always been one of my favourite classes to play"-efan

 

[Tony Vargas]

Ah, right. Goal and approach.

Yes, that. My faulty memory notwithstanding, it's good, succinct, guide, that seems more intuitive than the Game Theorist 'setting the stakes' or 'preserving agency.'

I've been meaning to ask: did you pull that from somewhere or was it inspired by anything in particular?

 

[iserith]

Yes, that. My faulty memory notwithstanding, it's good, succinct, guide.

I've been meaning to ask: did you pull that from somewhere or was it inspired by anything in particular?

It's likely distilled from a number of things in the D&D 5e Basic Rules, but it's key to "descriptive roleplaying" on page 66. "Drawing on your mental image of your character, you tell everyone what your character does and how he or she does it." The "what" is the goal. The "how" is the approach. "Active roleplaying" gives the same information, but in a different way and what you can't act out you have to describe.

The specific terms "goal and approach" probably came about as a shorthand in Twitter discussions between myself, [MENTION=6776133]Bawylie[/MENTION], and [MENTION=6801813]Valmarius[/MENTION]. I know it wasn't something I came up with on my own. And others have probably arrived at the same terms from some other direction.