"Rando" adventures

Malrex

So ... slow work day? Every day?
Was reading the recent review of Garden of Bones where its a 'rando' adventure.
I wanted to explore some of the comments:

"Emmy Allen’s Gardens of Ynn and Stygian Library (IIRC-correct me if I’m wrong) follow this broad template and were rather well received, so I’m curious as to what worked with those but failed here. Is it just a matter of better setpieces excused the overall random generation aspect?"


"There’s some similarity but significant differences. In those products, you generate a map as you go, but once generated, it remains somewhat static. There’s a lot more system to it, one you can interact with meaningfully."

"The difference there I guess is choice. Head into the unknown or retreat t the familiar. What has _changed_ back at the familiar place as a result of your actions also enhances interactivity/choice/consequence.
The limiting factor on both is lack of ability to plan ahead or scout for what’s coming next- although if denizens or encounters can provide clues that the giant likes beans or something around here breathes fire, you can ameliorate that somewhat."

Reason I'm interested in this is that I've been working on a 'dungeon' where one level is a "rando" dungeon--it's definitely not the focus of the adventure, but I believe it makes sense for the type of 'dungeon' it is in this particular case and I was experimenting with something new. It seems some other comments believe that this type of dungeon is lame or boring. I can certainly see some frustration and lack of interactivity as the GM just rolls dice to see what's next, but I also feel that if its a short area, it could be fun? It feels like the random tables in the DMG for when a DM is creating a dungeon but the PCs are along for the ride...

My editor said my area was a decent idea but didn't quite work and so I decided to do it the 'regular' way, yet include an alternate way to run the level in an appendix. I think one way I can make the alternate way better is to give the option for PCs to cut their way through and make their own passage if they really just want to get the hell out. I think that would reduce the 'frustration' as the time it took could be hand-waved, but if there is that component then truly what is the point of the 'rando' dungeon?

I've seen some discussions about the quantum ogre or whatever...wouldn't the 'rando' dungeon completely negate that? Is there a benefit to it?

I'm rambling...but just curious what other people think about 'rando' dungeons and if they can be saved, if they are just a total failure, and how they could be improved. I'm also trying to determine if people will feel 'ripped off' if they get a 30 page dungeon, but 5-8 pages of it is the same level of the dungeon but an alternate way to run it---or if it would be better to somehow combine the rando tables into the regular dungeon way--like an events table similar to Irridated Paradox that 1Tru did. Please share your thoughts.
 

DangerousPuhson

Should be playing D&D instead
Rando is feasible - Castle Gargantua, Gardens of Ynn, Vornheim, etc.

Randomly-generated places are fine, because ultimately it comes down to what the DM does with the space that affects how it interacts with the rest of the dungeon/adventure, not something inherent to the system of randomization itself. If your random rooms are boring, then they're boring rooms... that's not a problem with random generation, but rather a lack of creative input into the randomizer that would generate an interesting output.

For instance, a 1d6 random encounter table with:
1) giant rat;
2) giant spider;
3-6) 1d4 orcs

...is boring, not because it's a table, but because everything on that table is boring.

Here's a sample of a random item that would impact the rest of the dungeon:

1) Hive Chamber: a bloated queen sits atop a throne of shed exoskeletons, attended to by 2d6 drones at all times. In one day, the hive will absorb all immediately adjacent rooms, using them as a combination hatchery/larder. At the end of the week, if not exterminated, the hive will have a presence everywhere in the complex (triple all drone random encounters rolled, halve all others).
 
Last edited:

bryce0lynch

i fucking hate writing ...
Staff member
Of course it can be saved! But, maybe, we don't know how yet ...

Double Penetrated noted one issue, keeping the procedurally generated from being boring and brining it to life.

In contention with that is the prep work. Do you need to generate it before the game? DOes the process DURING the game slow things suffeciently to break the DM/Party flow? (I don't see how it can't, but I'm open to it)

A major decision point would seem to be "once rolled does it stay that way?" and dealing with the frustration factor and hinderance to the exploration game that implies.

Another would be: why? What impact does randomness have on the players? What if we CALLED it a random dungeon, and all the flavor text said it was, but it was just laid out for the DM like a normal dungeon? Maybe, examination of this question would reveal some Truths(tm) that can then be focused on instead of the distractions and assumptions we bring?
 

Two orcs

Officially better than you, according to PoN
I've run four sessions of Ynn and found it good when I used an internet tool to quickly generate the rooms and tables, otherwise looking up, rolling and generatingon the fly was too clunky. Of of course, the entirely open ended # of paths in Ynn broke my immersion, I would much rather there was a fixed number of alternate paths from each location. I think there should be included in the randomness limits to digging in place and retracting your steps to make the world more solid.
 

TerribleSorcery

Should be playing D&D instead
I think Gardens of Ynn is a genius module.

But something about random dungeons nags at me. I was never really able to articulate it, but it has bugged me since I read the Stygian Library. I think it is a sort of intangible thing - you can run games in a random dungeon that are great, obviously. But certain things will never happen, can never happen. Tactical combat is borderline impossible. The monster in the next room will never hear you causing a commotion and come to investigate (unless the DM has rolled up the next room in advance, and then the discussion devolves to "what counts as random?" so I'll leave it aside for now). There won't be choke points or themed areas. The territory of a big & dangerous creature is not distinctive - there is no telegraphing of threat, except through setting lore. These questions occupy me more than concerns about speed.

In Ynn, these are alleviated by it being a weird, freaky dimension. So I can give it a pass. But certain avenues are closed off nonetheless.
 

Malrex

So ... slow work day? Every day?
Ya...I agree...I think interesting areas def. need to be in there.

My reason for doing it is to bring in a heavily influenced environmental factor to the dungeon. Without giving away my idea....here is a made-up similar example:

Imagine characters are going to adventure through a gigantic mop that is underwater. Currents in the water are moving the mop yarns too and fro..sometimes opening up areas, and sometimes closing them off--all random. Within the mop are special areas, or interesting areas that DP pointed out and can be discovered if the dice rolls make it happen. But traveling through the mop can be random with wandering encounters, or maybe the mop yarns have a chance to crush you while you try to move through the passage, or entangle you, or have glass fragments within the strands that try to cut you or other environmental factors....all random and determined by tables. You can swim up, down, or wherever and the tables provide multiple entrances to new areas or one...or sometimes none and time is needed to wait for something to happen (and you hope your Water Breathing spell or other resources don't run out). There is even a chance that all the mop yarns open up perfectly to provide an exit out...or the PCs could cut their way out if desperate (or frustrated). So maybe one difference is there is a set area--the size of the mop.

So to answer Bryce's question with this example--once rolled, no, I don't think the dungeon would stay that way unless it was a special area. Perhaps creatures within the special areas have trimmed or tied the tendrils to create a "room". So those are solid locations...but the way back to them may be different or they may never be found again. That may be difficult for the DM as there would need to be some sort of map. Or rather...for this example--a circle representing an area of the mop, and depending on what the dice do, a map is created within the circle during play. Once a special location is discovered, it's locked in place, but all hallways and regular passages that may lead to these special areas change after a period of time.

No prep beforehand...you merely roll and glance at some tables to say what's going on. I can see how that may break immersion a little for the players, but I might argue, a little, that a rando dungeon, if there aren't too many rolling options, might take about the same amount of time as a regular dungeon mainly because a player wouldn't be asking details for mapping purposes. Would the timing be about equal? maybe. The DM would be drawing the map in that circle, but that's quicker than a player trying to figure it out.

I guess I'm wondering if the 'frustration factor' can be turned more into a 'panic, terrifying vibe'...you are underwater, you don't know how to get out...you are running out of air...mop yarn keeps trying to entangle you or hide strange creatures...That's the impact on the players I would be going for. PCs could still have some measure of control (they could start cutting the mop yarns), maybe that would reduce the frustration factor? But the panic comes from not necessarily being able to go back a familiar way.

Tactical combat...imagine you find a monster in a passage within the mop yarns..you start fighting it...a few rounds pass, and suddenly the hallway has other exits out to another passage or a different room...maybe there are monsters in there as well...or maybe the passage is starting to close with no exits and everything gets tangled up unable to move for a few rounds...you could potentially change the terrain to increase the number of tactics during a fight.

Finally...the PCs may choose not to explore the mop, and instead go explore the shaft of the mop which is the main adventure anyways...is something like that more workable then a completely rando dungeon?
 

DangerousPuhson

Should be playing D&D instead
But certain things will never happen, can never happen. Tactical combat is borderline impossible. The monster in the next room will never hear you causing a commotion and come to investigate (unless the DM has rolled up the next room in advance, and then the discussion devolves to "what counts as random?" so I'll leave it aside for now). There won't be choke points or themed areas. The territory of a big & dangerous creature is not distinctive - there is no telegraphing of threat, except through setting lore. These questions occupy me more than concerns about speed.
I think all these things are in the realm of possibility.

If you know how many rooms will exist beforehand (but not necessarily anything else about the rooms), then you can divide those rooms for the purposes of area control (ie. factions and connecting themes). You know that the general areas around the players are part of a larger territory, and simply apply whatever specific encounter rules you'd like to that territory. Simple extrapolation and management of all the variables can let you apply encounters/themes/situations to the adventure almost like layers; some being visible when others aren't, some hidden and some hiding others, each able to be blended with the ones around them to seem as one. And yes, these layers can all be randomly generated - your job as the DM is not to tweak the layers, but rather to tweak the relationship between the layers so as to make one cohesive piece.

As a demonstration:

I have given myself the following randomized keywords for things you'd find in a dungeon environment - gore, vermin, fountain, beast, undead.

The party is in the Tomb of the Randomized Cube, fighting their way through interconnected rooms and meeting challenges along the way. As they progress down this one particular passage (which branched off from an earlier staircase at the bottom of an empty pool, etc.), they notice that further rooms are made not of drab stone, but rather of bone and mummified sinew. There's sand all over the floors, and in a few branching rooms, some sarcophagi can be found/looted. There's a blood-filled fountain that, if fiddled with, spurts flesh-eating scarabs. There's a huge, moose-sized scarab beetle sleeping soundly on a bed of ancient bronze coins. In the room beyond, there's a headless mummified corpse sitting completely motionless atop a throne of golden asps.

This area of the dungeon is very clearly not the damp grey cobblestone of earlier. A theme pervades, even if all the guidance given to work with was person, woman, man, camera, TV "gore, vermin, fountain, beast, undead".

So let's say the players are now fighting a headless mummy who commands a swarm of golden snakes (because of course they would touch it) - how weird would it be if, at some point during the fight, a moose-sized stag beetle burst through a patch of wall and starting rampaging around the room, attacking all parties indiscriminately? I don't mean "how weird would it be" in the rhetorical sense, I mean literally on a scale from 1 to 10 how weird would that development be to the players? Would the players stop the game and say "whoa, no... sorry, but that's just not possible, and you've killed the immersion"? I would hope not.

Perhaps the appearance of the beetle will add intensity to the scene, or perhaps it will make the fight way too hard - that doesn't matter. What matters is moving a thematically-appropriate creature from another area as a reaction to the events transpiring in the current area. There is sparse guidance that can be squeezed from five random words, but there they've gone on to allow for a dynamic encounter through blending several elements together (vermin + beast + undead = gigantic beetle fighting a mummy).

Now, we've already got some established areas of the dungeon, right? This sinew place was, after all, discovered at the end of a passage, down some stairs set into the bottom of a pool. So it follows suit that unless the party is somehow blocked from going back the way they came, they still have that area to work with. They decide the moose-sized scarab is way too strong, and want to fall back to the narrow staircase from earlier, noting that the great beetle wouldn't fit in its pursuit. Hey, just like that, they're using a chokepoint.

Maybe they got the idea when they entered the areas inspired by the a previous batch of random words like "staircase", "narrow", and "Tucker's kobolds". Maybe they're just always on the lookout for something to exploit. It doesn't matter. What matters is that a series of five innocuous words has spun off a themed area, a dynamic encounter, and a tactical plan.

Yes, I had to improvise most of that, but that's where all that mumbo-jumbo about layers comes in. If instead of five random words per area, you were given five clusters of words... clusters like "gore[sinew/bone]", "vermin[insect/carnivore]", "fountain[blood/trap]", beast[sleep/unstoppable]", and "undead[ancient/tempting]"... I think you could come up with something similar to my example, or at the very least something dynamic, tactical, and thematic in its own way.
 

Beoric

8, 8, I forget what is for
Looking at your mop example, and thinking about what I would accept as a player, I think the randomness has to be random in a predictable way; that is, there needs to be a way for players to try to mitigate the randomness. Wandering monsters work this way, because while you may not know what is on the wanderers table, you know that screwing around or making noise increases your chances of encountering something.

So using your mop dungeon as an example, let's say that your map has a chance of changing every X number of turns, so that once a change is made the players know (or eventually figure out) that they have a certain amount of time before the pathways change. Lets also say that there is some precursor to the change, so the players get a warning of a few rounds and can try to hurry up what they are doing/where they are going. Let's also keep many parts of the dungeon static (major rooms and landmarks, for instance), so the players can make meaningful guesses after a change to head toward their destination (eg. they know the exit is to the south, but have to guess at the new route for going there).

Probably the easiest way to do this would be with some sort of gating system, using fixed positions for the rooms. When a gate is encountered you check to see if it is open or closed. Once that is determined it stays that way until you roll a "2" when you are checking for wanderers, at which point the change in environment is telegraphed and the players have 2d4 rounds to get where they are going before everything changes. Once everything changes, you have to re-check for open/closed if the PCs encounter the gate again; it would be a bad thing if that happened when the party was divided on two sides of a gate.

In most circumstances I don't see any value to randomly placing the rooms when you first start the dungeon, when they will thereafter remain fixed, since that makes more work for the DM and changes nothing from the players point of view. The only way I could see it having value is if you expect the PCs to return there later in their careers, at which point you reassign some of the room locations; or if you expect to run the dungeon more than once for the same players and need it to be replayable, in which case it had better be damned good because very few modules are good enough to replay with the same players.

If you do move things around, to make it easier on the DM I might suggest setting it up as a number of dungeon tiles which are assembled on a grid which differentiates between positions that are allowable for rooms and positions that are allowable for passages and intersections (which then rotate at random intervals).

EDIT: you could also have some discoverable static relationships, so that for instance there is always a secret passage from the Lounge to the Conservatory even if the position of those rooms changes.
 
Last edited:

squeen

8, 8, I forget what is for
I do think the idea. as said above, of randomly connected vinettes or set-pieces that start spill over into each other after a "mix" time could work.

Also, agree with Beoric: just pre-roll and establish the first set of results as a concrete example for the DM.
 

Malrex

So ... slow work day? Every day?
Ok..I'm taking some notes on this. I'm still on the fence if I want to take this chance and go all out and take a chance, or just have it an appendix as an alternative method of running this level. Definitely don't want a frustrating adventure. I appreciate this discussion.

"The territory of a big & dangerous creature is not distinctive - there is no telegraphing of threat, except through setting lore".
TerribleSorcery's point here stands out to me. I think I can telegraph this with my idea as this isn't magical fog or simply magic that makes the place random but a natural environment...so some random passages that appear could have some clues that something big passed by (tufts of fur or whatever).

"Lets also say that there is some precursor to the change, so the players get a warning of a few rounds and can try to hurry up what they are doing/where they are going. Let's also keep many parts of the dungeon static (major rooms and landmarks, for instance), so the players can make meaningful guesses after a change to head toward their destination (eg. they know the exit is to the south, but have to guess at the new route for going there)."
Beoric's comment here...this is a good idea and I can easily incorporate that in my idea. Sticking with the mop example, maybe it shudders every turn and PCs know change is coming in 2d4 rounds or something. Plus, if they do try to do a quick map...they may realize they are only a few mop strands away from a static room and can cut their way towards it.

The Gate idea--so I understand...are you talking like a Gate to a different dimension that leads to a static room? Like a permanent Rope Trick spell or something?

If the area of the randomness is known (the size of the mop head), and static/interesting rooms are fixed and unmoving within the mop....so all that is random is passages/rooms/wandering encounters/random things leading to these areas...with the option of cutting your way out or towards something...would this be interesting or just frustrating?

For a map, the DM would have a circle (of the mop area), static rooms, and the circle would have some dots and he could connect the dots as he rolls the random passages/rooms so he knows about where the party is. There could be player maps of the circle...so if they do mapping...and discover some of the static rooms, they would have an idea of how to get back there despite the random environment that makes up the dungeon.

Predictable randomness...does that sound about right?
 

Beoric

8, 8, I forget what is for
The Gate idea--so I understand...are you talking like a Gate to a different dimension that leads to a static room? Like a permanent Rope Trick spell or something?
Sorry, I should have defined my terms. All I mean is a point that may/may not be passible, be it a locked door, a sliding wall, a potential cave-in, or any other thing that blocks passage. Disappearing doors and corridors also work. Strands of a mop, I suppose.

On the DM map you identify points where passage might be open or blocked, so rather than changing the whole map whenever the mop moves, you only check the "gates" that the PCs encounter. The ones that are encountered stay static until you have a random movement of the map, at which point you check the ones immediately visible to the PCs to see if they have changed, and any others they encounter from then on, including those they have encountered before. Until they are seen they exist in a quantum state, neither open nor closed.

Say there is a room with four quantum passages that apparently enter it, each of which has a 2 in 6 chance of being open. When the PCs enter the room only one is open. While they are in there the mop moves and by random chance where they entered may or may not have closed, and the other exits may or may not open. If some characters are in a passage when it closes, they are either whisked to another adjacent open location (if one is available) or stuck in the strands of the mop until the passage opens up again (subject to creative ways of dealing with the situation). So the party has to be careful to make it to stable points when the alarm goes, and not get separated.

If the area of the randomness is known (the size of the mop head), and static/interesting rooms are fixed and unmoving within the mop....so all that is random is passages/rooms/wandering encounters/random things leading to these areas...with the option of cutting your way out or towards something...would this be interesting or just frustrating?

For a map, the DM would have a circle (of the mop area), static rooms, and the circle would have some dots and he could connect the dots as he rolls the random passages/rooms so he knows about where the party is. There could be player maps of the circle...so if they do mapping...and discover some of the static rooms, they would have an idea of how to get back there despite the random environment that makes up the dungeon.
I think we are talking about the same thing. In this example the random elements may or may not be traversable at any given time, but for the sake of managing it the places where those elements might be is static.

So say you have a circle with a number of rooms. Most of the rooms that have content (monsters, tricks, traps and/or treasure) are static. One or two rooms with content move around; each has four points where it might be, and when the PCs come to a place where it might be there is a 1 in 4 chance that it is actually there (unless it was established to be somewhere else and there has not yet be a reset). You also have a number of "empty" rooms which are placed at static points but which may or may not exist when the PCs arrive; these are effectively intersections, and are the "dots" in your example. There are also a number of passages connecting the static rooms, points where roaming rooms might be, and places where empty quantum rooms might be, and each of those passages may or may not exist when the PCs come to the place where it might be.

When the PCs come to any of the random locations (passage, moving room or quantum room) the DM rolls to see if the element is there. If not, all they see is mop. Either way, whatever they find stays static until the mop moves, at which point all random points reset and you roll for their status when the PCs come to their location again. Maybe put a checkbox beside each location; empty mean sit has not been encountered, check mark means it exists/is open, cross means it is closed; reset means you erase all your marks.

To reward exploration, add a couple/few secret passages that are always open in some way, so the players can find tricks to move around more predictably.

Predictable randomness...does that sound about right?
That ... is pretty much most of the good bits of D&D. Certainly every time you roll the dice, when the players know or can figure out the odds.
 

The1True

My my my, we just loooove to hear ourselves don't we?
I tried dealing with this issue in the Tesseract by drawing a map and then leaving the rooms open. Admittedly, the constraints of geometry are a pretty shaky thing in the dungeon and the rooms can bend and stretch to fit the descriptions. Also, as commented on, it needed more stuff. But I guess that's the thing here; with a rough map or even just a Point-map laid out, you're wide open to put more time and effort into rando room-dressing and encounters instead of worrying about generating geometry.
 

squeen

8, 8, I forget what is for
What if the rooms and geometry was static but the time-period in which you entered the room was random?!
:devilish:
 

DangerousPuhson

Should be playing D&D instead
In my eyes, there are only two real reasons to implement randomness into a module:

1) Replay-ability (only if employed properly; if 95% of your module is static, then the randomness won't help)
2) To fill a gap in the creative process ("I don't know what to put here, I'd better roll on the table")

Every other implementation tends to be a gimmick that adds more complication than it ameliorates, IMO.
 
Last edited:

Two orcs

Officially better than you, according to PoN
Now I've run several more sessions in the Gardens of Ynn. The randomness irks me from a game perspective, the players tried to be clever and plan from what they encountered last time and there was a severe lack of grounding (+ way too frequent random encounters, one per turn! running by the book instead of the more reasonable 1/6 per two turns). I got the best results pre-rolling locations and the riffing on the results, adding specificity and drawing proper maps.

What I liked a lot is how useful the custom class was. They picked up a "native" mutant of sorts last time (more than a year ago) and his odd abilities really felt at home in the environment.

Anyway, my new plan, in case they go back, is to ground it harder in reality. I already didn't use the "door" entry but placed in physically in the wilderness and had the passage be a thorny wall where you crawl through a badger's path that gradually closes up. At the center of the garden, visible from vantage points, is a magnificient Chateu (Castle Amber) which instead of the BS fogwall will have the garden as insulation. I'll draw a proper map where the locations are arranged in concentric circles, higher depth values the closer you are to the center. Whenever they enter they will from a random direction, so the random aspect is still there, but they can stumble into familiar territory so to speak.

tl;dr meaningful game needs meaningful decisions, randomness robs the players of that.
 

squeen

8, 8, I forget what is for
Improvisation is necessary evil for a DM, to my mind---but an evil none the less. Random tables help with improv, and maybe also as design seed generator, but that's the only use I can see for them. I don't think they have a proper place in a pre-written adventure, except in minor instances. Designers that imagine otherwise are probably deluding themselves it will work out better than it actually will.
 

Malrex

So ... slow work day? Every day?
Thanks for all the feedback.
For the adventure I'm working on, I may make the 'hallways' more permanent in the 'mop' but have random things going on, instead of having the DM roll up everything. That way it could be a simple, 1 roll.--Maybe its like a Wandering Encounter table, but happens more frequently. For this part of the adventure, I feel strongly that it makes sense that something is going on in the 'hallways'. It doesn't all have to be bad, but I want to try to capture that the environment (i.e. the hallways) could potentially be a threat.
 
Top