It's not the situations that are being set up that's causing this, it's caused by the underlying rules clearly and obviously granting a greater chance of success when only the "face" does the interacting.
<snip>
there's strong encouragement built into the system for most if not all characters to participate in combats; not least of which is that to just leave it all to the tank (analagous to leaving an important social interaction to the face to deal with) means you're going to bury a lot of tanks and then spend a lot of time recruiting (and rolling up) new ones.
But the example that you give is not about system. It's about situation.
If every combat was framed as a duel between one NPC and the PC "tank", then there would zero encouragement for the thief (or mage, etc) to ever get into combat. But the default in D&D a illustrated in modules; as set out by rulebook advice on encounter design) is not to frame every combat as a duel.
It's the framing that means only the "face" talks. That framing is, in turn, related to bigger issues of encounter and scenario design - for instance, if most interactions with NPCs are of the
get some information out of them or
justify why the PCs shouldn't be arrested/attacked for being here variety, then of course only the most persuasive PC will do the talking.
But it's not very hard to frame social interactions where those are not the only things at stake. It's as simple as the NPC saying to the fighter, "So, what's your view of this matter?" If the fighter stands there and looks dumb, well, that's a loss (but see further below). If the fighter replies, well, now the fighter is part of the social interaction.
If, in fact,
it doesn't matter that the fighter just stands dumb when asked questions by NPCs - that is, if social standing and reputation and good relations with NPCs and the like
don't matter in the game, then it's going to be hard to frame social encounters that engage all the PCs - just as, if it didn't matter whether or not PCs took damage in combat, then the mages and thieves wouldn't both with fighting and would just leave it to the tanks to mop up. Or to put it another way: social encounters where
it doesn't matter if you're a social failure are like combat encounters where
you heal all damage as son as it's taken, so it doesn't matter if you get hit.
No one would dream of setting up that as the default approach to combat; but I think in many D&D games it probably is the default approach to social interaction.
Okay, lets get a little ridiculous, lets say the player characters are in a shop trying to buy weapons, and one of them asks, "How much is that weapon?" The DM informs the player that he must roll a D20 to decide the outcome of this encounter. So the player rolls a D20 and the result is a natural 1, so the DM informs the player that the store proprietor gets mad, grabs the sword that is on the table and he attacks the player characters with it, and tells the players to role for initiative to determine the order of combat. The player who rolled the dices asks, "What happened, what did I do?" The DM tells the player, "You rolled a 1 on the d20 and as you know a natural 1 is an automatic failure in whatever you are trying to accomplish, sorry, just bad luck I guess."
What you describe is no more or less ridiculous than the following:
The PC walks into a room with an angry orc in it. The orc draws a sword; the player says "Well, I draw too and try and cut the orc down!" The player rolls a 1. The GM, for the orc, rolls a 20, then the damage dice, and the damage result exceeds the hit point total on the PC sheet. The GM announces "You're dead!" The player asks "Why, what happened, what did the orc do?" And the GM answers "Well, I rolled a 20 and you know that's an auto-hit and crit, and the damage roll was more than your hp. Sorry, just bad luck, I guess."
In other words, the core of D&D combat can be resolved without knowing anything more about the fiction than that character A has a weapon in hand, is in the immediate vicinity of character B, and attacks. If that's not a problem for combat, why is it a problem for haggling? Or, if you have techniques for dealing with this issue in combat, than why wouldn't you use the same techniques for a haggling scenario?
At it's core the game has always been, roll a die or dice and try to get over or under a target number.
<snip>
To support combat and social situations you add modifiers. Regardless of flavor, it's a +1 through +5 usually and most mods cap out at +2 to give you a 10% swing on a D20.
To support combat, exploration and logistics you add distance. This is usually in a number of feet, or number of squares (as an abstraction of feet) per turn.
To support social situations you add an initial disposition table. This is usually a d10 or d20 adding a modifier (usually charisma)
To support all the pillars you add a basic economy so that things may be purchased, bribed etc.
Everything else is DM fiat. Which is as it should be.
The combat resolution mechanics have never involved just the rolling of a to hit die.
There have always (in the post-Chainmail era) been damage rules, related to rules for hit point ablation and the consequences of that. From time to time there have been facing rules, positioning rules, movement and engagement rules, etc.
And these haven't always just been GM fiat. Gygax's AD&D has intricate rules for facing, how many figures can attack a single figure, when the shield bonus to AC does or doesn't count, when DEX bonus to AC is retained or lost (based on position and other elements of status), etc.
4e has very substantial non-GM fiat elements in its social resolution mechanics (ie skill challenges) (the most important being X successes before 3 failures; but other ones too, like the requirement that the player declare an action for his/her PC that makes sense given the fictional situation; and the distinction between primary checks (that, if successful, will advance the PC's goal in the challenge) and secondary checks (that, if successful, don't directly advance the goal but open up some other resource or opportunity)).
An initial disposition table (ie a reaction roll system) is not essential (4e doesn't have one - the GM sets the "initial disposition" as part of the framing of the situation). And it won't help if there is no system that permits the players, via their PCs, to actually generate changes in the relevant fiction. And I take it as given that the 3E version of that system, in its Diplomacy rules, shows why a simple "Roll X to improve the disposition N steps" mechanic, that has no scope for considerations of framing and context, is not feasible in a game which allows for essentially open-ended bonuses. (Traveller social interaction does use a "roll X to win" system, but bonuses in Traveller are tightly capped. And even then I'm not sure the Traveller system is impervious to breakage.)
Having mechanical means for suicidal challenges means that the PCs are subject to the same kinds of social "attacks."
It doesn't have to. In Classic Traveller players make morale checks for their PCs on the same basis that the GM makes them for NPCs; and in Burning Wheel the GM can call on a player to make a (comparable) Steel check.
But in AD&D and Moldvay Basic PCs are immune from the morale rules. And in Traveller, while players can make Admin and Bribery checks to resolve interactions with officials, there is no system that allows the GM to make Admin or Bribery checks on behalf of NPCs to resolve interactions with the PCs.
In 4e, a social situation may be resolved via a skill challenge, but the only way for an NPC to "win" the challenge is for the players to lose (ie 3 failures before N successes). What consequences flow from the NPC "win) (= player and PC failure) is up to the GM to narrate based on the logic of the ingame situation, but I think it is a matter for individual tables to decide whether or not a failure can include a PC having his/her mind changed.
I wouldn't do that at my table. The closest I have come is much more minor manipulations of PC behaviour as consequences for failed checks on the way through the challenge. Eg one time the PCs were negotiating with some witches, one of whom was a Pact Hag (and thus has, on her statblock, a whole host of verbal manipulation abilities backed by magic). A player failed a check, and I descibed the result as being that, as the hag talked to the PC, she led him to move through the room from point A to point B - point B being where there was a trapdoor, which the hag then activated by pulling on a cord so as to drop the PC into a pit.
For me, in 4e, that's about the limit of "mind control"-type consequences I will impose on player for a failure in a social skill challenge (in combat terms, it's a modest amount of forced movement). When I have posted that example in the past, though, many responders have thought that it broke the limits of acceptable consequence narration. But it's quite feasible to resolve social interactions via the skill challenge mechanics without using even consequences of that sort.
If anyone has experience with games that do a good job of mechanically implementing exploration and interaction, I'd love to hear about them. I hate glossing over them at my table, but I've yet to come across a better alternative.
Exploration's not really my thing, but OD&D used the Outdoor Survival game which some liked (eg someone just upthread mentioned it favourably); Moldvay Basic's dungeon exploration rules are highly regarded and are the inspiration for the contemporary game Torchbearer.
I think Classic Traveller's jump space mechanics work nicely, because (like the Moldvay exploration rules) they use a tight resolution sequence: roll for an encounter as you leave the system; roll checks for drive failure and misjump; spend a week in jump space (which can inclue a roll for hijacking if there are NPCs on board; and other situations the GM chooses to run, but none are obligatory); roll for an encounter upon arrival in the destination system. (It's quite similar to Moldvay's
move the PCs their movement rate, map, resolve any declared checks to look for traps or secret doors, mark off the appropriate number of turns, roll a wandering monster check every X turns.)
Conversely, Classic Traveller has poor rules for planetary surface exploration. The weakest moment of play in my current Traveller game occurred when we got stuck in these rules - which are basically "GM fiats all distances and directions, the players roll for vehicle malfunction every ingame day, the GM checks for an encounter/event twice a day, every day uses X amount of rations/oxygen". Everything turns on the passage of ingame days, but these are just parcelled out by GM fiat. (On the return trip I had the PCs, in their ATVs, under bombardment from an orbiting starship, and used the "quickie" resolution system that is intended for when a small craft comes under attack from a starship. That system worked well, and let us avoid the tedium and fiat of the proper rules.)
For a completely different sort of exploration resolution system there is Cortex+ Heroic. The default published version of this is Marvel Heroic RP, but the Cortex+ Hacker's Guide has a number of variant rules, including exploration rules for fantasy gaming. The basic unit of play in Cortex+ Heroic is the
scene. In MHRP these are either action scenes or transition scenes; the Hacker's Guide introduces exploration scenes. In
my own Cortex+ Fantasy game I don't use exploration scenes, but just treat them as a species of transition scene; and by a combination of house ruling plus induction from various rules that are part of MHRP, I use a rule that in a transition scene each player is allowed to declare a single action (rolled against the Doom Pool). If that succeeds, then the player can establish an asset - so a player might make a check to establish a Found the Path Through the Forest asset, which would then provide a bonus die on salient actions in the next Action Scene.
Exploration-type assets can also be created in Action Scenes; when the PCs were trying to rescue some villagers who were locked up inside a giant chieftain's steading, one of the player's established a Hole in the Pallisade asset which then gave a bonus die for the action to rescue the villagers (which, from memory, in mechanical terms was an action aimed at degrading the Imprisoned Villagers Scene Distinction).
In 4e skill challenges can be used to resolve exploration - I've done a few of those, as well as slightly more conventional approaches.
For social resolution I'm finding Traveller fine at present (it has a few subystems - a generic reaction roll, which I let the players make as a type of "influence" check; the systems for resolving bureaucratic encounters; the patron encounter system for estabilshing contacts/recruiters). Burning Wheel is very different - it has the Circles mechanics for establishing friends/contacts; and the Duel of Wits (which can bind on players/PCs as much as the GM/NPCs) for resolving disagreements/persuasion attempts.
In Cortex+ Heroic social conflict is resolved in a mechanically identical fashion to any other conflict, but instead of Physical Stress and complications like Stuck on the Top of the Washington Monument, you deal Emotional or Mental Stress and inflict complications like Smitten. (Those are examples from actual play: in the end, Ice Man resolved the conflict between the PCs and B.A.D. by riding in on ice slide and carrying Diamondback off into the sunset (in mechanical terms, stepping up her complication to a disabling level).)
In 4e skill challenges work well for social conflict resolution.
Here's an actual play example.
