Sandbox gaming

[chaochou]

Okay, general ponderings, and a lot of questions for people, about running NPCs.

When you are roleplaying an NPC, does that NPC tend to make optimal or rational decisions?

I ask, because one of the facets of the world is that lots of people - even very competent ones - are capable of highly sub-optimal, irrational and stupid decisions. Do you trust yourself to do the same with NPCs, or see it as a necessary or interesting facet in a sandbox game?

I'll give an example: We have a devil trying to outwit a king. As GM you could decide that the devil succeeds. You could decide the devil fails. You could do this: Roll d20 -
King: (1-3) Blind idiotic rage (4-7) Listens to wise counsellor (8-19) Makes best decision (20) Makes best decision with unforeseen additional benefits.
Devil: (1-20) Cold, ruthless, uncompromisingly rational

In time the devil will most likely gain the upper hand when the king gets angry and does something stupid. That's how devils are. They have limitless time and an otherworldly singlemindedness. The devil could also try to influence the wise counsellor. How wise is this wise counsellor anyway?

So some questions: When you 'simulate' an NPC, do you have a method to that simulation? Do you simulate fallibility or irrationality, and if so, how? If your notes say someone is 'calculating' are they always so? Are 'stupid' NPCs allowed moments of clarity, wisdom or far-sightedness?

 

[Umbran]

<yawn>

Um, chaochou, you should be aware of something - on EN World, whether or not you started the thread, you don't "own" it. While you may politely ask for some folks to remain on a specific topic, but that's about as far as it goes. He isn't breaking any rules. If you don't like what he says, you can ignore him, but getting dismissive, sarcastic, or personal is not a good idea on your part.

 

[Mallus]

When you are roleplaying an NPC, does that NPC tend to make optimal or rational decisions?

Not usually. Partly because I don't claim to be able to figure out what the "optimal" or "rational" decision would be, given the extraordinarily, and often extraordinarily batty situations non-player characters find themselves in, in your average D&D milieu. Or, from another angle, since the DM determines what constitutes "optimal" and "rational", those terms lose their meaning when the DM applies them to themselves. It's like being sole judge of a beauty contest you're also competing in. So I shoot for NPC's that make "theatrical" decisions; some with more pretense to the rational than others, depending on the character. I want their decisions to lead to entertaining, and genres(s)-appropriate situations.

I don't ask "what would a ruthless and intelligent person do?", I ask "what would Iago do?"

I ask, because one of the facets of the world is that lots of people - even very competent ones - are capable of highly sub-optimal, irrational and stupid decisions. Do you trust yourself to do the same with NPCs, or see it as a necessary or interesting facet in a sandbox game?

One little nod to verisimilitude I do like to make are smart, competent NPC's nevertheless make bad decisions. The trick is to make the reason for their NPC's irrational behavior evident, for instance, they're blinded by racism, species-ism, ultra-nationalism, etc. Their poor decision-making becomes a form of characterization.

Also, by accepting that NPC's, like real folk, sometimes act irrationality, I free myself of the need to figure out what the smartest course of action would be.

I'll give an example: We have a devil trying to outwit a king. As GM you could decide that the devil succeeds. You could decide the devil fails. You could do this: Roll d20 -

I don't like making encounter charts --so I don't-- but I frequently do this: partially randomize NPC's reactions to prevent, ahem, authorial over-determination, or, more plainly, so I can be (a little) surprised by an NPC's reactions.

 

[The Shaman]

I find the idea that these have been separated interesting. Not that I'm disagreeing with it, but I'm curious as to how much of a distinction you see between them.

They're not completely separate - the first applies to most cases of the second, but the first also applies to cases which are not the second as well.

For example, the adventurers decide to seek out information on a mysterious man in black from a random encounter. They decide to check out a tavern near where they encountered the mystery man; my prep assists me in improvising the encounter so that it's consistent with genre-emulation and the setting.

For example, you could improvise that there's a woman fleeing a decrepid house clutching a moneybag, hotly pursued by a noble trying to pull on his shirt. Or you could roll 'noble' on a random encounter table.

The difference is what I think of as having a 'light touch on the controls.' Am I directing an encounter at the adventurers, or am I creating an environment in which such an encounter may take place?

I only direct encounters at the adventurers in response to specific actions they take which are likely to provoke a response - frex, I'll assign a probability that the man-in-black hears word of the adventurers looking for him, and he'll react as appropriate to his nature and circumstances. The players act, through their characters, and I react, through the setting.

By randomizing those slice-of-life encounters, and including the possibility for coincidences, I'm setting the world in motion and to the extent practicable letting it play out on its own, subject to the actions of the adventurers. There are obviously boundaries and limitations to this approach, but it's as close as I can get to an autonomous simulation of the game-world.

Does it matter at that moment which meta-game method you used? I'm not seeing the distinction in the in-game result, despite two different methods. Is there some cumulative effect of 'random' encounters which over time must produce a different game to one which is solely improvised?

It heightens an element of surprise on both sides of the screen, it makes the role of the referee more reactive than proactive, and it preserves the feel of the game as a game, all in my experience, of course.

 

[Mark CMG]

Okay, general ponderings, and a lot of questions for people, about running NPCs.

When you are roleplaying an NPC, does that NPC tend to make optimal or rational decisions?

I ask, because one of the facets of the world is that lots of people - even very competent ones - are capable of highly sub-optimal, irrational and stupid decisions. Do you trust yourself to do the same with NPCs, or see it as a necessary or interesting facet in a sandbox game?

(. . .)

So some questions: When you 'simulate' an NPC, do you have a method to that simulation? Do you simulate fallibility or irrationality, and if so, how? If your notes say someone is 'calculating' are they always so? Are 'stupid' NPCs allowed moments of clarity, wisdom or far-sightedness?

In my Whispering Woodwind adventure product, I use the CMG PROSE System (attached). Maybe this will prove useful to others.

 

[nedjer]

Um, chaochou, you should be aware of something - on EN World, whether or not you started the thread, you don't "own" it. While you may politely ask for some folks to remain on a specific topic, but that's about as far as it goes. He isn't breaking any rules. If you don't like what he says, you can ignore him, but getting dismissive, sarcastic, or personal is not a good idea on your part.

Perhaps Morrus could offer shares in threads or executive subs so those who want to buy the right to be dismissive, sarcastic and all that other good bad fun could pay for the right to beat up on others. Kind of monetise hate

 

[Ariosto]

Now in a linear scenario, inspired by the graphical storytelling source, you really have a question of how and when Batman will defeat the Joker.

Speaking from experience, I will say that this kind of trip is so different from where the familiar RPG vehicle goes that you may as well leave that baggage behind.

Unexamined preconceptions, and legacy systems with no warrant but "how it's always been done", will mess you up.

Why are we rolling dice? Why do we have a Game Master? We ought to have answers to all such questions more sound than, "We're just imitating what others have done, without any real understanding."

Often, others did what they did as means to ends other than what we are attempting. They may be at cross purposes with our designs.

In this example, there can be a real game-play element in that question of how and when Batman wins. It can be possible for the Joker player thereby to win the scenario despite losing the conflict in the imagined world, just as the player of WW2 Poland or France might be able to win a scenario by points with a skilled defense.

(In a wargame, it typically would even be theoretically possible to defeat poorly played invaders.)

Going even further afield, actual strategic or even tactical potential could be almost negligible next to some sort of use of game trappings as means for telling a story. This is one direction that computerized "interactive fiction" has taken since Zork, continuing from such classics of the heyday as Trinity and the very political A Mind Forever Voyaging.

Great fiction can make for a poor game, and a great game can make for poor fiction.

 

[Ariosto]

Giving my own answers to questions chauchou posed to pemerton and The Shaman:

(1) Exploration: The pioneering RPGs were/are games of limited information. That is the prime need that the GM fills, more than procedural adjudication (which can be minimized in the "rigid kriegspiel" model to which, e.g., WotC-D&D comes closer than old TSR-D&D).

They were/are basically games of playing a role, in the sense of imagining oneself in a particular situation, with a particular perspective and set of resources, and responding on that basis. There are things that a knight, a cowboy, a tank commander, a spy, a spaceman, etc., cannot perceive or otherwise does not know because of his position in the imagined world -- just as each of us has such limitations in the real world.

Bran Mak Morn lives and dies in his world, not sitting at Bob Howard's typewriter in Texas. Part of the appeal of D&D, RuneQuest, etc., was/is that one can "change the story" from within.

(2) Random Events: These serve several closely related purposes.

First, they are -- along with a host of other occasions for dice-rolls -- part of the basic structure of a very probabilistic game with great strategic scope. This is for instance a significant factor in considering the random factor in character generation. We expect to explore the game as a sum of histories rather than as a single isolated slice.

It's like playing a card game through multiple hands. The spread of probabilities gives context to what one is dealt, and playing the odds is a long-term process.

Second, they serve in a sense to keep the GM "honest" when it is their probabilities -- rather than whims -- that dictate the frequency of this or that occurrence. We human beings tend to fall into unconscious patterns that can be skewed from what we think we are up to.

Third, they are another contribution to the ability of the game to take turns that surprise even the GM. Such surprises greatly help (in my view) to keep the moderator's role fun rather than drudgery.

Fourth, the use of tables and dice-rolls can be a form of "brain storming" for inspiration. As The Shaman observed, they typically give but seeds that the GM must interpret and flesh out. Those seeds can be a valuable aid to creativity when one is trying to come up with one interesting variation after another (as for instance in creating a subsector for Traveller).

 

[rounser]

There is also the option to go back the way you came. A "railroad" isn't a matter of place, it's a matter of the forced narrative from the DM/GM/Referee/Facilitator.

I disagree. PCs can abandon an adventure hook or plot in progress, just as they can abandon a dungeon. Dungeons make good settings for D&D because they limit options, and "force narrative" to a choice between room A and room B.

IMO walls do railroad PCs passively, and implicitly. Just because it's not an active channelling through hamhanded DMing doesn't make it a way to implicitly limit player choice. And that's not a bad thing IMO. I think, once again, people's definitions of railroading and sandboxes are too narrow, and their associations with which is good and which bad a bit too simplistic.

Shades of grey, not black and white...both in terms of definitions of these terms and what constitutes good or bad campaign design.

 

[The Shaman]

Dungeons make good settings for D&D because they limit options, and "force narrative" to a choice between room A and room B.

IMO walls do railroad PCs passively, and implicitly. Just because it's not an active channelling through hamhanded DMing doesn't make it a way to implicitly limit player choice. I think, once again, people's definitions of railroading and sandboxes are too narrow, and their associations with which is good and which bad a bit too simplistic.

I think, once again, that this is attempting to broaden a definition beyond the boundaries of comprehensibility.

A landscape feature, even a somewhat linear landscape feature, does not railroad players and their characters. A referee who coerces the adventurers by ignoring or negating their decisions to follow a specific path or course of action railroads players and their characters.

Expanding the definition of a railroad to include landscape features leads to . . . well, this.

DM: The innkeeper takes your coins and directs you upstairs to your rooms for the night. You see the stairway at the far end of the common room.
Player 1: Whoa, watch out for the railroad!
Player 2: Yeah, I'll cast fly to access our rooms.
Player 3: I'm going to climb the wall to the windows after the wizard flies up there.
Player 1:[/i] Okay, while the wizard and the rogue are accessing the windows, we'll build a pyramid out of tables in the common room and hack a hole in the ceiling with the dwarf's axe and rig a rope to climb up . . .

A linear landscape feature is a linear landscape feature. They may be gauntlets if the referee is so inclined, but they do not railroad the players or their characters in and of themselves; the referee does if he forces the adventurers through the gauntlet.

Sending guards after the adventurers when they commit a crime isn't railroading. An island which must be accessed by boat (absent other means like magic or a flying mount) isn't railroading. A road crossing a mountain range isn't railroading. Presenting the adventurers with a bad choice isn't railroading.

Saying the adventurers are captured by the guards no matter what the players try to do, forcing the adventurers onto the boat or over the road, and taking away any choices is.