[Lendrhald] helping GMs create Challenges

[David Berg]

In the game I am designing, I would like for the system to help GMs in constructing Challenges (in the Gamist sense).

Does anyone know of any games out there that do a good job of this?

Challenges in my game are intended to adhere to the following parameters:
1) all aspects of the situation must not violate the logic of the world (said logic involves medieval tech, pragmatic human cultures, invading orcs, and pockets of malicious magical evil hidden away from most folks' awareness)
2) situations must cover a variety of difficulty levels (so they don't feel "rigged" -- the players shouldn't expect every enemy to be something they can kill)
3) situations must allow a variety of choices by the players as to the precise nature of their Mission (it could be a useful and worthwhile mission to kidnap one orc; or kill 2 orcs; or track 10 orcs and rally locals to attack them; or scout the camp of 100 orcs and persuade the government to call in the legions.  The players get to choose.  If they want to walk away from 100 orcs without doing anything, that's fine.  If they want to fight them, they should all die.)

Besides simply recommending to GMs that they just do the above, I would like to provide as much assistance as possible in this potentially challenging task.  Making step #3 happen is where I suspect some system could be of most help.

Note:
I currently have one mechanic in place to ensure that Challenges are of varying difficulty levels.  The GM is instructed on an "average Lendrhald difficulty level", and then told to roll 2d6, with the outlier rolls making the Challenge easier than average or harder than average.

[Vulpinoid]

Does your core game feature leveled characters? Or, does it use a more open experience system?

I only ask this because level based experience systems give characters a nice incremental scale that can be reflected through challenge difficulties.

This is a difficulty 7 task, a level 7 character should have a 50-50 chance of passing it.

If you assume that tag teaming characters have an innate bonus because they have complimentary strengths and weaknesses...

Here we have two level 3 characters, that means a difficulty 6 challenge is about right. But since there's an extra character involved we'll push it up a notch to a difficulty 7 challenge.

If you characters don't come down to a simple level based progression, things get trickier.

Arcadia: the Wild hunt had a nice system of "Waylays" which were complications that the characters encountered as the moved around the game as they completed their quests. There were usually easy ways to face a waylay, or hard ways. You could face off against a troll in combat, but you'd probably get flattened, or you could try to trick him through your cunning...I don't remember specifics, but there was some kind of rule where if you fought him and won, you'd remove him from the game; while if you tricked him, you'd get past him as an obstacle but others might have to face him later (he'd become a recurring villain).

V 

[Judd]

Agon comes to mind when I think of gamist challenge building.

d20 too but I've never been particularly good at that aspect of that game, so perhaps I'm not the one to say.

Beast Hunters also might be well worth checking out.

Good luck, David!

[David Berg]

Judd,

I don't remember d20 actually showing the GM, "Here's how to build a good challenge."  Is there some chapter of some book that I'm blanking on?  I actually have some d20 books, if you could help point me to the right place that'd be awesome...

I'll check out Agon and Beast Hunters; thanks!


Michael,

I'm fine with my current 2d6 mechanic.  I guess I should clarify that the difficulty level is evaluated relative to the world (i.e., w.r.t. an "average" dangerous mission), not to the characters.  The difficulty range of almost all missions (unless you roll a 12, basically) will roughly correspond to the range of character improvement, such that starting characters can just "beat" the easiest mission, while played-for-years characters can just "beat" the (nearly) toughest mission.  I want Competence 3 parties to run into Difficulty 1-8 challenges at random.  The next step is ensuring that they still have something fun to do when they hit that Difficulty 7 challenge.  So, I hope to design a system that helps GMs write those options into that challenge.  Discussing that is the purpose of this thread.

The mechanic you describe from Arcadia sounds appropriate to a structured game, but maybe not to an open-ended one where the characters can go wherever and do whatever they want (which my game is).  Varying levels of reward depending on how you tackle a challenge is definitely a good direction to go in, though (as long as players are made aware of that).



One idea I'm pondering is having set pay rates.  Money is important in my game.  It gets you gear to fight with.  Most characters start out with little money and little gear.  The Empire definitely pays adventurers to take care of minor threats to civilized areas.  I was thinking that perhaps they pay a more for some achievements (utterly destroy evil force) than others (spy on evil force, give report of their capabilities & movements).  This should get me partway where I want to go... but not all the way, I don't think.

[Selene Tan]

IIRC, both Agon and Beast Hunters use a resource-limited GM. i.e., for every unit of play (mission, adventure, hunt, etc.) the GM has only so many resources to spend on opposition. That may not fit so well with the "go wherever and do whatever" style of game you want.

D&D has some text in the DMG about calibrating challenges. NPCs have Challenge Ratings (CRs) based on race and class. If the encounter is with onlyone NPC, then the NPCs CR is the Encounter Level. When you put together several NPCs, there are rules that determine the resulting EL (it's not strictly additive, and it's not well-defined for mixed-CR encounters). Assuming a party of 4 equal-level PCs, an encounter with EL equal to the average party level is an appropriate encounter. It should use up about 1/4 of the party's resources, and 13 such encounters should give them enough experience to advance one level.
There's also an interesting numerical analysis/method for determining appropriate encounters that someone wrote up at http://www.epiphanies.org/games/cr_ep.pdf

There was a challenge-creation system I was toying with for a game I was working on that used multiple encounter difficulty ratings. The system had 5 major abilities/strategies the players could use on challenges, and each challenge had a value for how strong it was against each of the 5 strategies. Depending on the desired challenge difficulty, the GM had X points to distribute between the strength ratings.
The challenges had what were basically hit points, with the strength ratings determining how much damage would be taken from a particular type of "attack". So players could futz around with a variety of strategies as suited them, e.g. use attacks the challenge was somewhat strong against because the players were really strong in that area.

[Callan S.]

Currently I'm looking at multi generational procedure generated* situations. A quick example of procedure generated content in this case is a unit which is a farm, another unit which is a rabbit population, and grass land units. Each has their set of rules which act as you might expect them, but when the rabbits eat the grass, breed, eat all of it (the grass, by it's rules, couldn't regenerate fast enough) and start destroying the farmers crops and by extension threatening his life, it generates situation.

By multigenerational I mean you don't stop there - lets say the farm is destroyed, producing a vagabond. The rules of the vagabond are followed and he moves to join the local bandits. The local bandits are strengthened enough to stop the trader units passing and an economic downward spiral occurs. Keep running generation upon generation.

With units who's rules aren't specifically designed to create a situation, but do interact with other units, and by running several generations of results, a completely unknown situation can be generated. One which no one would have thought up. Truth is stranger than fiction, if I may refer to procedure as truth

Looking at the situation and being moved to make something a goal to win (eg, clear the bandits, or even clear the bunnies) is another step and the most vital (without it, it's all just 'shit happens'). But I think it's a relatively fun and intuative step and should be easy enough when the time comes.


* This is the name I'm using for it - some might not agree it's the right name

[Vulpinoid]

There's also an interesting numerical analysis/method for determining appropriate encounters that someone wrote up at http://www.epiphanies.org/games/cr_ep.pdf

That's a good resource. I'd probably err between the two methods of XP distribution and set the wghole thing up in excel, or a flash based program that I could use to plug numbers in and get XP out. But of course, this relies on a similar character level mechanic as found in D&D.

Back to the core topic though, if we are looking at the "main encounter" for a session, I'd consider a system where a core encounter difficulty was modified by +2d6 as was proposed in the original post. Characters are paid a certain amount of resources for taking on the challenge based on this core difficulty +d6. This is where the employers of the group know half of the difficulties that might be faced by the group.

The twist I'd use is that once the characters had accepted the job, roll the other d6 (or keep it secret from the players), and they only discover the true difficulty of the job once the remaining obstacles are encountered. The secret die could be a 1, and therefore the job is easier than initially expected, or it could be a 6 and there could be some true nastiness in store for them. On the whole though, the character have an inkling of how difficult it will be and shouldn't get too far over their heads (as long as they take note of the events happening around them).

V

[David Berg]

Characters are paid a certain amount of resources for taking on the challenge based on this core difficulty +d6. This is where the employers of the group know half of the difficulties that might be faced by the group.

Hmm.  The Imperial government needs a job done that they only know a certain amount about, and they set a price based on that knowledge.  That all sounds right.  No die-rolling should be needed for this, just guidelines on how much the empire pays for stuff.

The twist I'd use is that once the characters had accepted the job, roll the other d6

It does indeed make some sense that there be some disparity between what the imperials think is going on and what is actually going on.  I think, however, that the nature of the Challenge situation needs to be determined first, and then the Imperial perception of it logically determined based on that nature.  Whether the empire is willing to up their pay if a mission was harder than expected is a good question.  I would guess the answer is, "Not without mid-mission negotiation."

There are also payment questions for any given Challenge situation:
1) does the Empire already know about the situation
2) how easy is it to contact an Imperial that will quote you a price, pay you, get you help, etc.

This is interesting, but I fear this thread getting sidetracked, so let's please steer away from all discussions of Level Of Difficulty for now.  I think Callan may be on to something that is more what I'd like to discuss.

[David Berg]

Currently I'm looking at multi generational procedure generated* situations.
. . .
With units who's rules aren't specifically designed to create a situation, but do interact with other units, and by running several generations of results, a completely unknown situation can be generated. One which no one would have thought up. Truth is stranger than fiction, if I may refer to procedure as truth

This is 100% appropriate to my "relies on in-gameworld logic" mandate.  And the resulting complexity and details will help tremendously in lending missions the proper color.  Bravo.  Some tough questions, though:

A quick example of procedure generated content in this case is a unit which is a farm, another unit which is a rabbit population, and grass land units. Each has their set of rules which act as you might expect them

So, are you suggesting that the game designer (me) provides the GM with an exhaustive list of units (to cover the different areas of the world) complete with rules that are thorough enough to describe their interactions with all other units?  That sounds desirable, yet utterly impossible.

With these units and rules, the GM then picks an arbitrary starting state, and then:

but when the rabbits eat the grass, breed, eat all of it (the grass, by it's rules, couldn't regenerate fast enough) and start destroying the farmers crops and by extension threatening his life . . . the farm is destroyed, producing a vagabond. The rules of the vagabond are followed and he moves to join the local bandits. The local bandits are strengthened enough to stop the trader units passing and an economic downward spiral occurs.

proceeds to "run" all this in his head for some arbitrary duration until he's decided it's ready for the player characters to stumble across it?  Unless all the rules of all the units in play are very clear and simple, this also sounds pretty impossible.

Were you thinking of writing some software to handle this or something? 

Or did you just have more elegant ideas of designer and GM tasks than I'm coming up with?

Looking at the situation and being moved to make something a goal to win (eg, clear the bandits, or even clear the bunnies) is another step and the most vital (without it, it's all just 'shit happens'). But I think it's a relatively fun and intuative step and should be easy enough when the time comes.

I agree that having a reliably realistic and complex scenario awaiting the player characters should make it easier for them to find ways to latch onto them...  I dunno if "easier" is still "easy", though.  With no GM guidance to the situation-formation process, it actually strikes me as a complete crapshoot as to whether the variety-of-goal-options I'm hoping for will actually arise or not.

[masqueradeball]

The reason most RPGs don't give strict guidelines for encounter design is because difficulty is an amazingly variable thing when considering a wide range of character abilities. In Dungeons and Dragons, for instance, CR's are completely inappropriate if your party is weak in a given character type (undead, for instance become exponentially more threatening if your party has an extra rogue instead of a cleric).
How does this pertain to your question? Well, I guess I'm saying that I would be able to give advice (maybe not good advice, but advice) on how to generte encounters if I had some idea of the scope of your characters. Otherwise, its hard to make any suggestions.
Also, what your describing sounds like some methods of adventure design I've seen presented out there. My advice would be to start out by creating a reward that the players can receive and a method that they find out about this reward, and then create a number of encounters, each keyed to a different set of character abilities. There could ba combat way to get the reward, a magic way, a social way, etc... and your game could provide rules for building each. Then each adventure could have one reward, a number of challenges that would lead to the reward and then the encounters would be keyed to a map or time line that the characters could explore/move through. The decisions that they then made as which encounters to take on would determine how challenging it was to gain the reward, based on their strength. To make more complex adventures, you would simply add more rewards or more steps towards achieving a single reward.

[David Berg]

I would be able to give advice (maybe not good advice, but advice) on how to generte encounters if I had some idea of the scope of your characters. Otherwise, its hard to make any suggestions.

The players' strategic options are limited to the options they would have if they were actually in the gameworld, possessing the knowledge, skills and equipment of their characters.  Said skills correspond to what might be expected of a human lviing in medieval times, ranging from farmer to veteran solider.  Some characters will be good at sneaking, others at shooting with arrows, others at striking with swords, others at reading mannerisms to detect lies.  All of these can be improved with the points gained through playing.  No superhuman abilities will be gained, but the occasional supernatural possession may be found (usually more Nifty than Badass).  Some players will be good at coming up with ruses, others at giving tactical orders, and others at speaking in convincing or charismatic fashion.

These examples are not all-inclusive, just representative.

My advice would be to start out by creating a reward that the players can receive

Might be sensible.  I'll go through this with an example:
The Empire will pay 5 shillings to anyone who can drive off a small band of Orcs attacking the town of Fjotdale, plus an extra shilling if you can prove you actually killed all of them.

and a method that they find out about this reward

Empire posts a sign where the West Road meets the road to Fjotdale.

and then create a number of encounters, each keyed to a different set of character abilities. There could be combat way to get the reward, a magic way, a social way, etc...

GM determines (through whatever means, let's not discuss that now) how tough the Orcs are, and figures that the PCs could not take them in a head-on fight.  There are several things that might work, though:
- picking them off one by one
- leading them into some sort of trap
- enlisting 3 or 4 more fighters to help
- other things that the players may come up with that the GM hasn't

One of the players is a good smooth-talker.  Another player's character is particularly stealthy.  A third player has the best fighter, and a fourth player has the best archer.

The GM relates these to his list:
- archer picks them off
- stealth guy tries to lure them into separating, where they can be picked off or trapped; or, stealth guy tries to lure them all into trap
- smooth-talker convinces imperials to lend weapons and locals to lend bodies
- fighter kills individual orcs that have been separated from their group

How does this turn into a list of potential encounters?  Should it?

and your game could provide rules for building each.

This is the step I'm hung up on.  How's the game text going to help the GM take all the options above and Make It So?

the encounters would be keyed to a map or time line that the characters could explore/move through.

I have some rules in place for communication between GM and players about who wants to play what kind of session.  The GM then uses this info to guide his prep.  However, the game cannot include any sort of "map or timeline" which makes the players view any encounters as having been put there for their benefit.  Everything in the world can be ignored if the players so choose, and ignoring any one situation shouldn't cause play to stop.  Likewise, if the players in the above example decide, "Well, let's go kill orcs," and just wade into their camp with swords drawn, the characters should all die, and the players will be more strategic next time.  This (and other failures) is something the GM should not prevent from happening.

[xenopulse]

Here's my core question for you:

You say the challenge needs to conform to the laws of the world, but do the game mechanics actually have to simulate those laws?

Beast Hunters' solution is to have the mechanics be open, with the specifics determined by the narrative. Those mechanics don't simulate in-game causality, that's left for each action and the GM's adjudication. That is, each action grants a certain amount of advantage, taking all the narrative details and circumstances into account, but that advantage is handled the same whether you're sneaking or fighting or laying traps. Similarly, the damage that's inflicted is generic and can represent a multitude of things, it does not model any particular injuries or morale in different ways (other than the mental/physical/social categories).

If you want a system that actually models the laws of the world, which also simultaneously allows a wide range of tactics and a balanced creation of challenges, you've got a very ambitious project ahead of you, and one we can't tackle without knowing the specifics of your system.

[J. Scott Timmerman]

I think accomplishing what xenopulse is suggesting is primarily a matter of making context-dependent options available and necessary regardless of specific situation.  Here, I'm not talking about the more broadly defined Situation, simply about having mechanics adaptable to different types of conflict. 

It's something I'm attempting to do with the basic logic behind my system as well.  Basically, my suggestion (and it may feel like hitting your head against a brick wall sometimes) would be to start with axiomatic notions of in-game-world logic, but only those notions that apply to the outcome of conflicts.  Think directly about how these apply to Player options.  For instance, does the weight of an unconscious peasant matter?  Or is it simply enough to say that we can separate elements like so "This is a peasant that no individual PC can lift" and "This is peasant that a PC could lift individually."

That way, when setting up Conflict in the Scene, you decide directly, is "lift without help" an option, or not, rather than having to either in preparation or arbitrarily decide during gameplay what the weight of the peasant is.  I find that even in D&D, which has explicit mechanics for lifting certain weights, we often ignore the rule anyhow.

Perhaps this can be done with everything, so that every Conflict in every Scene can be broken down into simple options such as this.  These simple options work within in-game-world logic, because the scenes were designed that way.

Sets of typical ways of manipulating the Scene from within itself could be covered in system, and others could be creatively made up (preferably in advance) by the GM.  Anybody who exceeds the GM's creativity here by thinking of an option that the GM didn't should be rewarded just for that with success.

This idea kinda came from computer text-based RPGs (Yay, LORD) which had simple logic.  Also, I don't like making arbitrary in-game decisions against the players. 

I think that maybe Callan's idea for generating conflicts through elements of setting is great for thinking of ideas, but for setting up the conflicts themselves, it's probably easier if we have a simple handle on most of the variables.

I don't know much about Beast Hunters, but from what I've heard, its mechanic doesn't appear to get the context-sensitivity David might be looking for in making Challenges.

Or maybe this whole post is just too obvious...

-Jason Timmerman

[David Berg]

You say the challenge needs to conform to the laws of the world, but do the game mechanics actually have to simulate those laws?

Well, in certain areas they do, and in certain areas they don't.  Lendrhald physics work like real-world physics, and no mechanics are permissable which fail to make resolutions work out accordingly.

In terms of generating in-gameworld content (e.g. Challenge situations), I don't really care what the process is as long as the results make sense given the laws of the world. 

I got all excited about Callan's idea on the hope that he has some thoughts toward implementing it that he will soon share...  If your point is that the process itself may not get me anywhere, then yeah, agreed.

Beast Hunters' solution is to have the mechanics be open, with the specifics determined by the narrative. Those mechanics don't simulate in-game causality, that's left for each action and the GM's adjudication.

Resolution outcomes must mirror real-world causality.  So either the system has to handle that, or the GM does.  I would feel like a lazy designer if I just said, "GMs, use your own knowledge of How Stuff Works to arbitrate everything ad hoc."  Is there some other option I'm missing?

the damage that's inflicted is generic and can represent a multitude of things, it does not model any particular injuries

So then a player or GM decides what happens?  I am fine with players and GMs adding color to outcomes.  I am not fine with letting them determine what the outcomes are -- see previous point.

If you want a system that actually models the laws of the world, which also simultaneously allows a wide range of tactics and a balanced creation of challenges, you've got a very ambitious project ahead of you, and one we can't tackle without knowing the specifics of your system.

Yeah, it's very ambitious.  Let me reiterate, though, that as long as everything comes out opaque and plausible, I'm happy to have the GM get as Meta as he wants during his prep.  Roll-on charts, how-to steps, checklists, whatever will help in building the types of challenges I've described (i.e., allowing players to choose between multiple ways to Step On Up).  Intricate connection to the processes of the world is just a nice bonus.

I'm not sure which specifics of my system are actually relevant here.  As for what the players can do while the game is running, just imagine yourself with a sword and the ability to use it, walking into an unfamiliar place and seeing a monster in the distance.  Supposing you were as brave as the characters we play, what would you do?  Whether you decide to track it, kill it, take a tentacle, or use it to terrorize your estranged father, you're welcome to try.  My current effort is to provide as much assurance as possible that those options be present (though I am okay relying somewhat on player resourcefulness) and (most importantly) fun/rewarding to tackle.

If that all makes sense to you and you still need to know how our mechanics attempt to model reality, then ask away.

[David Berg]

Jason,

I definitely think we're on the same page about a lot of stuff.

making context-dependent options available and necessary regardless of specific situation.  Here, I'm not talking about the more broadly defined Situation, simply about having mechanics adaptable to different types of conflict.

Sounds good... just want to make sure, though, that we're talking about mechanics that determine what's in the world (i.e., nature/identity of Challenge situation).  I don't intend for this thread to discuss my task resolution system any more than absolutely necessary.

start with axiomatic notions of in-game-world logic, but only those notions that apply to the outcome of conflicts.  Think directly about how these apply to Player options.  For instance, does the weight of an unconscious peasant matter?  Or is it simply enough to say that we can separate elements like so "This is a peasant that no individual PC can lift" and "This is peasant that a PC could lift individually."

Sure, only worry about what you actually need to worry about, no need to go charting and graphing every last environmental detail...

The thing is, with the players determining what gets played, the list of potential things the GM needs to worry about is infinite.

That way, when setting up Conflict in the Scene, you decide directly, is "lift without help" an option, or not, rather than having to either in preparation or arbitrarily decide during gameplay what the weight of the peasant is.

"When setting up Conflcit in the Scene."  Please elaborate (with an example if you don't mind). 

Agreed 100% on not wanting to include more work than necessary during prep and on avoiding on-the-spot arbitrary determinations when conveninetly possible.

Perhaps this can be done with everything, so that every Conflict in every Scene can be broken down into simple options such as this.  These simple options work within in-game-world logic, because the scenes were designed that way.

And the GM can be helped to create this how exactly?

Sets of typical ways of manipulating the Scene from within itself could be covered in system, and others could be creatively made up (preferably in advance) by the GM.

When you say "covered in system" do you just mean that the book provides the GM a list of ways in which PCs might tackle situations?  (I'm already working on one, in fact.)

Anybody who exceeds the GM's creativity here by thinking of an option that the GM didn't should be rewarded just for that with success.

Success must depend on whether or not it would actually work in the gameworld.  The GM's job is to arbitrate this to the best of his ability.  If in-gameworld logic is totally mute, then sure, what the hell, be nice, indulge the players.  But that's won't come up often.

for setting up the conflicts themselves, it's probably easier if we have a simple handle on most of the variables.

A time-efficient way to achieve that would be awesome.