Degrees of encounter-ness?

edited June 28 in Story Games
Short version:

Maybe we could roll a lot for what happens as characters travel, with some rolls providing nothing and others providing color and others providing difficult encounters, with uncertainty over which will be which.


Long version:

Sometimes it's fun to play through the journey. Sometimes, when you're getting from interesting location A to interesting location B in the fiction, you don't want to just, *poof* arrive at B.

Maybe you want some possibility that something might have happened en route that was interesting in its own right, or affects what condition you're in now, or something.

Maybe you don't like showing up at point B with either no idea what happened en route, or some idea that you just made up after the fact.

Maybe "covering" the journey helps your sense of immersion or character identification or plausibility, realism, world-modeling, causality, etc...

So, what techniques do we have for that? Well, we have a ton of options, but I think the main categories are:
  • Roleplay the journey out minutely, fast-forwarding only when it's totally obvious that nothing interesting will happen just now.
  • Let the GM control the pacing. They simply decide "here's where you are when X happens" and introduce stuff they expect to be worth playing through.
  • Rolls. Especially wandering monster rolls and random encounter rolls.
The pros and cons and how-to's and best practices for those first two options could generate their own long threads, so right now I'm just thinking about the rolls. I think rolls can contribute a lot here, taking some potentially difficult work and weighty decisions off the GMs plate.

The rolls I'm most familiar with during character travel look something like this:
  • Per unit of time or distance travelled, roll to see whether you encounter danger or what danger you encounter.
Compared to the full range of things that might happen when one travels, this strikes me as a bit limited. Ideally, I think I'd like travel rolls to cover more bases. As long as we're relieving the GM's burden to come up with stuff, why not include more than just monsters and dangers? How about free loot and weird people and bad weather and ancient ruins and pretty rainbows also?

But before we start adding "weirdos" and "rainbows" to our encounter tables, there's a problem to consider.
  • If we're actually going to slow down and get into character and interact with weirdos and rainbows, that may actually mean we're increasing the GM's burden in providing content.
  • Or, if we never get into character and interact with stuff, then I don't think the rolls are adding all that much to play.
What I think we need is a tree of rolls. We need a broad table, and then some sub-tables. If the broad table roll produces something the players don't interact with further, cool, we have a bit of additional color for this part of the journey and nothing more. If the players do interact, though (either because they choose to or because they must (e.g. combat encounter)), then instead of the GM improvising to fit the rolled prompt, then GM just rolls again on a narrower list for any additional content that's needed.

So, as a player, you're moving along toward point B, and you encounter a spooky well. Wouldn't that be cool if it were neither a mandatory encounter nor immediately obviously irrelevant? Wouldn't it be fun if the uncertainty of travel persists through the point when you decide to poke at the well or not? Wouldn't it be neat to know that, even if this time you're just going to speed to point B, there was a chance that poking the well might have brought death or riches or a cool story?

That's the kind of table I'd be excited to roll on.

Just pondering how this could work, I could see the second level of roll-on tables being either concrete fictional details like 7 kobolds or broader stuff like types of threat/opportunity. In the broader case, there could be a third roll for 7 kobolds or whatnot.

All thoughts, anecdotes, or brainstorms along these lines are welcome!

Comments

  • I have some old school D&D thinking about related topics in my desk drawer, let's see what's in there...

    About color encounters

    Something I did in a certain sandbox hex-crawl was quite simple: encounter rolls were on a d6 per traveled distance or elapsed time as usual, but with two target number ranges:

    A '1' would give you a real encounter off the encounter table, with potential for high stakes like death or adventure hooks or treasure or whatever.
    A '2' or a '3' would give a "color encounter", meaning inconsequential setting detail. My sandbox had a list of examples for these, but the GM was encouraged to ad-lib according to their imagination.

    The GM would presumably roll these in secret for the most part (or however secret you do - with us it's more "roll in the open but the players don't care to pay attention"), and not tell the players explicitly in advance whether a given encounter is a real encounter or a color encounter.

    This has a few desirable features for where I'm coming from:
    * It breaks the meta-level expectation of encounters being serious; you can't assume that every time the GM starts talking about noises coming from a bush that it's something consequential. Many real encounters overlap with color encounters in terms of superficial tropes: in this system you can't know if this merchant guy you meet on the road is actually important or not, for instance.
    * It's a regular signal that encourages the GM to wax poetic about setting color. As we know, the color is what by the process of emergence becomes new situations and meanings later on, so it's important to have a regular routine for piling that on.
    * It cleans up dud encounters from the encounter chart. I am mildly infamous about having had some really weirdly anti-dramatic encounters in my games at times. It gives better sandbox control if the "wandering cow" encounters occur more often and more freely than the actually important stuff.

    Moving to a cumulative distribution

    In a different context from the above, I've been mulling over the way random encounter checks behave in longer journeys between discrete points. Referees often use the same exact process for a "we want to travel from London to Istanbul" as they do for "we want to wander around exploring for the city of gold". The issue with this is that the constant grind of multiple random encounter checks over time in what amounts to monotonous travel on a road isn't the most pleasant process to play, as there's usually not much or anything to say about the hexes the travel passes through: you're just rolling the dice until you either get there or you get an encounter.

    I developed an interesting alternative dicing method for these situations: instead of rolling for a binomial distribution by rolling one die per unit traveled, I would roll the dice to find out how far the journey gets to progress before an encounter occurs. This effectively turns the probability of a random encounter into a cumulative function similar to the cumulative effect of the repeated encounter checks, but with less dicing and less swingy probabilities.

    For example, in a given sandbox campaign context we might say that a party on foot has an encounter every d6 leagues of travel, with '6's exploding. If they're travelling to a place 10 leagues away, they may expect to have two or three significant encounters on the way on average, but they could also potentially get there with no encounters at all due to the exploding die.

    We could also say that a party on horse-back, or even in a wagon, rolls 2d6 instead to account for less time spent on the road and the likelihood of passing over potential encounter opportunities. And we could of course have various modifiers for places where encounters are more or less likely.

    This method works well if you have a sort of boardgamey game board on which to track the party's progress. It is just like a boardgame at this point, in fact: you roll the dice, move forward as many steps as the dice indicate, and then have an encounter in the "space" you moved to. The GM could use encounter tables specific to the place where the party ends up with the die.

    (Note that although the die in this method indicates the movement of the party over geography, it doesn't tell you how fast they're moving - it just tells you how far they got with routine travel before they were interrupted by a special event.)

    In a more travel-intensive context, though, you don't really need a map - you just need to know how far it is to the destination, and how you're going to determine the kind of encounter that occurs once you've figured out where it's going to be.

    Considering my original concern over how long, relatively safe travel sequences are handled in comparison to dangerous hexcrawl exploration, the encounters in this system should either be mostly minor travel occurrences (bad weather, broken equipment, etc.), or they should be dramatically exceptional, but occur only relatively sparsely. For example, my personal aesthetics find it reasonable for a 19th century adventuring party to have a couple of random encounter adventures on a rail trip through Europe, as one is wont to do in a period adventure novel. Having an interesting occurrence every ten miles would too much, as most of the time you'd expect travel to be rather routine. So you need a different style of encounter tables than you do in wilderness exploration travel, even if both are technically travel encounter schemes.
  • Firstly,
    Several years ago I remember being in a discussion about random encounter tables, and trying to push the idea that they didn't necessarily have to imply conflict.

    That is to say, if you encounter 10d10 orcs and roll 96, you're going to see that column of orcs from a long way off, and you'll probably notice them before they notice you. Where are they headed to (or from?) and why? It's not an encounter, it's a new story hook in an evolving world. Same thing with a 2nd level party encountering a dragon. You notice a massive shadow pass overhead; the dragon doesn't care about you, but damn, now you know there's a dragon doing SOMETHING in this part of the world you're adventuring in.

    Secondly,
    The AD&D 1E Oriental Adventures book, for all its problems, has great random events tables for Yearly, Monthly, Weekly, and Daily occurrences. These range from things like major political assassinations, natural disasters, and the appearance of new cults, to someone getting sick, being challenged to a duel, or stumbling across some ruins. The majority of the daily events are not monster encounters, for what it's worth. Absolutely worth checking out if you're into that sort of unplanned world building thing.

    For what it's worth, I think MechWarrior 2E also has really cool random campaign events tailored to its specific milieu, only some of which pertain to actual combat in big stompy robots.

    Thirdly,
    I think like you said, the idea of whether to slow down at all is obviously important, but I also think that you don't need to interact with the results of any single roll for a significant amount of time to make it meaningful. I worry what I'm about to say might be anathema to some people's play styles, but not everything needs to be a hook, not everything you mention in the game needs to be explored.

    Of course, I know some players will take the opportunity to try and exploit everything and inevitably bog down your game and accidentally trigger TPK's because they just had to know what was in the spooky well, but that's hardly the random encounter tables' fault.

    Fourthly,
    Regarding the player who has to peer into every well and poke their hand into every hole, I think that part of why this kind of encounter design fails is because of this idea that everything must be meaningful or else it wouldn't be there. It's a hard metagame issue to get past because it generally conflicts with a social contract point about wasting people's time.

    This is a really tricky part of messing with time compression and focus that I don't have a really good answer for, but I suspect that a play group culture of being up front about negotiating stakes and player buy-in to "optional" content might help. To that end, your formalized "tree" is probably a really good way to help prompt those discussions. Having explicit levels of engagement is some really sound thinking.
  • Some great ideas here.

    I've given this kind of thing a fair it of thought, as well, and I like the idea of further exploring the techniques and approaches which might be most helpful for any given game.

    I think there will necessarily be differences in application based on the purpose of "encounters". For instance, in old-school dungeon crawl mode, the "conflict or colour" nature of encounters comes from Reaction Rolls, not the encounter itself. (Meeting a group of monsters and then rolling a "they ignore you" result, for instance.)

    In that style of play, of course, the danger of encounters must be preserved, though, since it's a pacing mechanism which needs to be dangerous, or it stops being a necessary source of time pressure for the players.

    Eero's approach seems to be more fostering a sense of verisimilitude and allowing opportunities to truly "explore the world", bringing it to life, and shading with various nuances. (I have some follow-up questions about that!) There's a certain aspect of naturalism at play here, it seems to me.

    Dave's is, if I'm reading it correctly, to give a sense of passing time, to create context, and to add Colour to the events the characters are experiencing. Allowing the player to perceive passing events better from the character's experience is on his mind (I think).

    Each will need a slightly different application. Eero seems to be assuming that the GM is capable of and enjoys improvising Colour on the spot, whereas Dave is talking about creating a tool which makes that a non-issue in the first place. All very interesting stuff!

    Eero, can you tell us more about this? It caught my eye, and I'm not 100% sure what you mean:


    * It cleans up dud encounters from the encounter chart. I am mildly infamous about having had some really weirdly anti-dramatic encounters in my games at times. It gives better sandbox control if the "wandering cow" encounters occur more often and more freely than the actually important stuff.

    What are "dud" or "weirdly anti-dramatic" encounters, and how does this process help "clean them up"?

    What is "sandbox control"?
  • Dave, that's how the encounter tables in 5e already are. You sometimes come across of a silent spot, or a well, or a dead tabaxi, or…
    It's not all monsters on there.

    Like, if you want to do hexcrawl do hexcrawl. I love hexcrawl. If you want to fast forward then fast forward. That's also good. Or pointcrawl. Every prep has it's own "scale".

    We're doing a (Tomb of Annihilation) Chult hexcrawl now where just a couple of hexes is the size of all of Barovia, one our previous hexcrawls in Curse of Strahd.

    So taking one "step" in Chult is like crossing half of Barovia. The map is more zoomed out. And instead of rolling for encounters every couple of hours, it's just three times per 24h period. This is awesome, how some modules are at one zoom level and others at another. You can even do multi-level; perhaps a short journey is done on a 6 mile hex map and another is done on a 100 mile hex map. And what's hours per hex on one map is weeks per hex on another. (Using the movement rates in your game of choice ofc.)

    Learning hexcrawl procedures has been so good. It's a substrate for all sorts of games. Like, our Corsair Council game was a political game set in the city of Hawa. But the ship went on a disastrous expedition and since we had the hex map & encounter tables for the region Hawa was in (thank you, 2e box sets) we could hexcrawl it out. They only met a giant, and was able to barter with it. The journey wasn't a big part of the game. Although they all almost starved

    There's nothing wrong with just going "ok so after three days of walking, you're at the dungeon" if you don't have a wilderness map+enc table. And if you do have one, just use it. Learning to use it is a good investment
  • Paul_T said:

    What are "dud" or "weirdly anti-dramatic" encounters, and how does this process help "clean them up"?

    What is "sandbox control"?

    "Sandbox control" is simply the procedural framework you use to maintain a sense of order in what you do. A sandbox campaign consists of many types of facts and dynamic processes that need to be kept in some sort of order. In this case, for example, I think that distinguishing between "color" encounters and "real" encounters on the GM side makes maintenance easier even if the actual refereeing processes are identical in both cases. By keeping these two types of encounter ideas in different boxes (different lists, tables, whatever) I can add to them more analytically and riffle through them quicker than I could if they were all in one big box.

    As a concrete example, writing up the sandbox with the concept of "color detail" on hand was easier for me because I could categorize certain types of content as too incidental to deserve careful statting in advance. By treating the encounter concept of "wandering merchant" as a color element I save myself the work of giving one a stat block and frequency of appearance and other similar properties that I would give to violent orcs.

    As for weirdly anti-dramatic encounters, imagine if you had something like this in your random encounter table:

    "A lost cow wanders on the road."

    While I have nothing against having the party in a hex crawl encounter a lost cow, it is also clearly not a substantial encounter likely to ramp up into consequential play. Putting it in the random encounter chart gives it undue weight - I might as well put "The sun shines particularly prettily today" on there if that's the direction I'm going. And I would need to be most careful about the distributions if I didn't take care of it with a formal structure like I posit here - what exactly prevents me from free-associating a random encounter table with like a hundred color encounters and two actually substantial events if I'm just throwing all the ideas into one table without categorizing them?

    By establishing the concept of "color encounter" I clean up my mental space of the wandering cows, allowing me to focus on creating a random encounter table with teeth, while also maintaining a definite window in which the cows are allowed to wander.

    The actual random encounter tables in that particular sandbox are very sharp, I think. If it wasn't in Finnish I'd share it here as my state of the art in that matter. (I think we discussed random encounter table construction a few years back, it's based on the ideas I was spit-balling around then. Clearly distinct geographical areas, each with a punchy d6 table of common encounters and a couple of rarer special encounters, with color encounter ideas on the side as a free list. Optimized to give a gamey, distinct feel to each different region the adventurers travel over so that it really matters which region you are in.)
  • But... but... shouldn't we avoid teleological thinking? ;-)

    (As long as we're perfectly willing to let that cow gore, trample and kill a character when provoked, we're fine, I guess.)
  • Yeah, I mentioned that in an aside - this whole conceit of some things being color and some things being something else is an organisational principle for indexing game content while doing prep, not a predetermination of what kills characters and what doesn't. Once the stuff is in play, it does what it does.

    In other words: wandering cows definitely should kill characters in unfortunate circumstances, it's just that I propose that instead of having a shared 1/6 of hexes where you encounter cows or orcs, perhaps the pacing would be richer if you had 1/6 per hex for orcs and a complementary 2/6 for cows, for a total of 3/6 hexes that trigger the GM to specifically add content (on top of general travel description, of course - you describe the terrain and weather and whatnot for every hex anyway). Cows are much more common than orcs in many places, after all, and they're quicker to deal with as game content.

    My older method for doing this stuff has basically been to talk about color content in an undisciplined way whenever the inspiration strikes, plus having rather uneven random encounter tables where some entries are more like orcs and some entries are more like lost cows. This is methodologically rather messy, as there is no significant reason why the party couldn't end up encountering lost cows via either process: I might get inspired to say something of the sort for the heck of it (as per my right to spout "neutral color commentary" on a whim), or I might not, or the party might encounter the same thing as a formal encounter. I don't like this, as there are some vague corner cases to this sort of messy process that need extra attention. Better to clarify it to myself what exactly the purpose of the minor color encounter is, and then regularize the procedure towards that.
  • What I'd do in that case, Eero, is to have the encounter table itself to be like "10–40: color". Set the encounter check to 3/6, and make two-thirds of the encounter table just read "color".

    (Note to my players: this isn't something I've done! I put wandering goats, lost cows etc etc on the real encounter table!)

    As far as prettiness of sunshine I have a separate table for that, unrelated to encounters.
  • edited June 29
    2097 said:

    Dave, that's how the encounter tables in 5e already are. You sometimes come across of a silent spot, or a well, or a dead tabaxi, or…
    It's not all monsters on there.

    Oh, I didn't know that! Looks like 5e and I are somewhat on the same page. :)
    2097 said:

    Like, if you want to do hexcrawl do hexcrawl. I love hexcrawl. If you want to fast forward then fast forward.

    I think I'd like to try a randomized mix of both. With minimal prep.
    2097 said:

    This is awesome, how some modules are at one zoom level and others at another. You can even do multi-level; perhaps a short journey is done on a 6 mile hex map and another is done on a 100 mile hex map. And what's hours per hex on one map is weeks per hex on another.

    Great examples! "Map scale" is something I'd like to be emergent in the moment (though possibly within broader constraints). I've been pondering whether "roll for how much time elapses before the next Thing" might be more efficient than "roll every X unit of time/distance, with some 'nothing' possibilities". I definitely want "stuff happening" to not be predictably paced.
    2097 said:

    There's nothing wrong with just going "ok so after three days of walking, you're at the dungeon" . . .

    Yeah, I've done that just fine. Sometimes I think it's clearly the best option for the real people at the table, dealing with time constraints and whatnot. Ideally, though, I prefer to make those 3 days come to life a bit more, with at least the apparent possibility of changing position (gaining loot, losing HP, etc.).
    2097 said:

    . . . if you don't have a wilderness map+enc table.

    I think I'm talking about a system for building those maps/tables. Or at least for informing what should be on them, so they can then be built to spec. That seems to be one option, anyway. Or I guess I could collect just the right set of pre-existing ones. :)
  • Eero, looks like we agree on the upsides of "roll to see when the next thing happens". I like your exploding 6's as a way to potentially stretch that timeframe out.

    Also agreed that it can be handy for the GM to put encounters into boxes for organization's sake, regardless of how those encounters appear to the players. I think "potential encounter" might deserve to be its own box. That way you can summarize the color encounters ("as y'all walked to town, you heard a bunch of rustling in the bushes, but whatever investigation you did revealed nothing") without playing through fruitless investigation, and play through only stuff that at least might go somewhere.

    An alternative might be to have every color encounter be potentially something more... until the next die roll. If that die roll comes up "nothing to see here", then the GM can summarize and end the encounter.

    I think I said something like that in my opening post...

    I'm not sure which of these approaches is preferable, or whether some hybrid might be best.
  • edited June 29
    All very interesting ideas.

    Dave, I really like the idea of a handful of randomly-rolled prompts informing the Colour narration of a journey; it could he really helpful for players into a certain kind of immersion or play of a more "literary" style.

    Eero, I'd love to see some of your tables. Perhaps Google translate would do enough of a job to get the idea across?

    Sandra, are there any examples 5e-style encounter tables to look at out there? Everything I can quickly search for is just a list of monsters.
  • No, it's in the DMG, the XGE, and in modules.
    But in OSR stuff such as the One Page Dungeon Contest there often are cool enc tables
  • I'm just surprised to hear that there is this kind of technique or technology in 5e encounter tables... it's certainly not something I've heard anyone talk about before.
  • edited June 29
    Well, I'll try to cut a little sample slice for demonstration's sake. I think this is pretty cool stuff, even if only tangentially related to David's topic:

    First, here's the hex map of the sandbox. It depicts the coastal region of the land of Nod (of biblical fame), an antediluvian plain, in the final years before the great flood. The wider context is that this is to be used as a sort of a pocket universe (think Ravenloft) in a 19th century mystery archeology campaign - characters time-traveling to pre-flood times will end up on this map. The antediluvian world technically speaking continues outside the map, but I'm playing with neat prep here, so I just prepped something that looks nice on the premise that it's very likely enough for any immediate needs.

    The map consists of several multihex bio-social "regions" indicated in different colors. There's a sea, a primeval forest, a marsh, some farmland... the usual stuff. Each of these regions has their own write-up in the following format:

    [Region Name]

    [General description, transmitted to the players once the characters discover the new region.]

    [Local rules for navigation and getting lost; travel rates by locally applicable means; visibility from hex to hex; local odds for foraging and barter.]

    Color encounters: [A list of notes events and sights particularly typical to the region.]

    Random encounters: [A random table of the local random content.]

    Special encounters: [A couple of particularly remarkable/complex/rare/unique local encounters.]

    Hex contents: [A bird's view description of each individual hex belonging to the region, by coordinate. The description lists the local feature highlights in order from those most easily found while exploring the hex to those least easy to find.]

    Let's pick one region to sample - the sparsely populated marsh region, for instance:

    Perspiring Marsh

    A foggy, boggy marshland that turns saltier and wetter towards the northwest, and finally ends in a wishy-washy manner on the coast. Home to a rich ecology of dinosaurs and lesser beasts. The entire region is covered by countless pools, streams and waterways of the river, dry land only exists on the occasional mound, which tend to grow thick groves that act as landmarks and sanctuaries for all kinds of animals. The population consists of independent families of bog people, as well as mysterious lizard men.

    Getting lost is easy (3/6) and there are no roads; travel by foot is slow (1 hx/day), quicker by boat (2 hx/day), and even quicker along the river (4 hx/day downstream). The sea is visible from the higher mounds of the coastal hexes on a clear day. Hunting and gathering are only possible for characters familiar with the region.

    Color encounters: Exotic birds, different fogs at various levels, strange noises from weird directions, shapes moving under the water surface, humanoid observers in the distance, alluringly colorful flowers, giant ferns.

    Random encounters:
    1d6
    1 A greater dinosaur 3d8 HD - 20% a terrible orcalizard, 80% a calm neckosaur
    2 Fog - 50% magical (random low level spell)
    3 3d6 hunting lizards 1 HD or 1d6 crocodiles 3 HD
    4 3d6 lizard men
    5 3d6 stirges
    6 Special encounter

    Special encounters:

    The Great Hippoursinus: A 10 HD river-dwelling sweet-water bear, its coat distinctly green from algae. Basically a huge bear that lives like a hippopotamus.

    Lamech: the patriarch and cultural hero has his home in hex [2|4].

    [followed by the list of hex write-ups for the region]

    The above scheme is additionally supported by a separate special encounters table that is generic to all the regions. The intended procedure is that the initial random encounter roll determines whether it's a color encounter or a real one, and a roll off the real encounters table picks between the five common encounters and the 1/6 chance of a special encounter. If you get a special encounter you go to the special encounters table, and only if that indicates one of the encounters listed here (such as the elusive hippoursinus), then you get to pick one of them.

    Here's what the special encounters table looks like - as you can see, it is essentially a collection of various causal schemes that produce a massive variety of encounters, so as to balance out the fact that 5/6 of the local encounters are simple and characteristic to the local ecology:

    Special encounters:
    1d6
    1 Changing ecology: select a new normal encounter for the region (e.g. from a suitable monster book), and add it to the regional encounters table. You can also remove an old encounter, switch up the dicing, etc. alongside the revision. The party encounters the new arrival right away.
    2 Wandering monster: roll or pick an encounter off the closest nearby other random encounter table; this may be from another region, city or dungeon.
    3 Monster lair: adventurers stumble upon the headquarters (or main herd, or a map/guide that leads there) of an encounter. Roll or pick to find out which encounter's "lair" it is; if it's a normal monster encounter, there are more of the monsters and their leader is an elite (+1 HD or such).
    4 Old grudge: A familiar encounter has found the party again, by accident or design. If there are several possibilities, roll or pick. If there's nothing in party history, there is no encounter.
    5 A travel accident: A remarkable difficulty of travel. Roll the type from subtable; the problem may be trivial or inapplicable due to advance preparations or circumstances.
    6 Named special: either one of the regionals, or one off the common list

    As the table indicates, there's also a region-free list of named special encounters in addition to the ones listed for the region, so the GM has a few options there. I would probably randomize a bit between the special encounters, but they're all so distinctive and contextual that it's not a bad idea to make it a semi-intentional choice once you get to the point in the dicing where one of them is indicated.

    The travel accident subtable is just a list of humdrum travel issues that are technically speaking possible, but most of the time essentially just color; their presence on the special encounters table is because their becoming so bad as to require attention should be rather rare. I guess I'll just put up that table, too:

    Travel accidents subtable:
    1d6
    1 A bad storm
    2 Equipment breakdown
    3 Accidental injury or illness
    4 A quarrel among companions
    5 The road is cut off
    6 The path is misleading

    In summation, the random encounters scheme explained here features roughly 33% minor color encounters, 14% regionally typical normal encounters, 2% rare variations, 0.5% hilariously unlikely travel accidents and 0.5% real special encounters with fabled beasts and legendary heroes. Or in other words, a party travels an average of 216 hexes in the marshlands before they might stumble upon the Great Hippoursinus by chance.
  • Thank you! This is very much appreciated.

    So the random encounters include...

    1 One of two archetypical big-ass creatures (rather than a whole array)
    2 A spooky, terrain-appropriate weather effect
    3 A bunch of small predators or a few large predators
    4 A band of monstrous humanoids
    5 A bunch of fantastical monsters
    6 A special encounter

    1-5 are totally terrain-appropriate (even the stirges who get across the "in a swamp you'll be plagued by bloodsuckers like leeches and mosquitos" vibe). I like it!

    (Seems that the orcasaurus is the greatest danger (at least nominally, with around 13 HD), though as a predator, it might be satisfied with one PC and dinosaurs are typically dumb anyway, so I'd totally send my character into this swamp. It's only a 1-in-30 chance anyway. Ill-tempered lizard men seem like the greatest TPK risk to me.)
  • Yes, I would expect the orcasaur to be an overpowering threat for a party that just has to get into a tangle with it. In practice it's like 1/180 to step into a hex and stumble in its path, and even then it's going to be however-likely to encounter it in such a ways as to cause combat - about 1/3 at my table I would expect. Possible, but I wouldn't let the risk sway my hexcrawl decision-making as a player.

    In practice the velociraptors (as Sandra has clearly attested) are the guys who deal the TPKs, of course. Lots of attacks with plenty of teeth. Merciless and aggressive pack hunters for the win.

    Reading this stuff anew, I'm not extremely fond of those HD values for the dinosaurs - I can see what I was getting at there, and this was for publication in a Finnish fanzine, so I needed to be cross-compatible, but I think that standard gygaxian HD math can get rather awry with dinosaurs - they get inflated HD to account for large size. At my own table I would prefer benchmarking those big dinosaurs at something like 4+d6 HD rather than 3d8 (so more like 8 HD average than 13), plus I'd give them a +50 HP bulk bonus or bigger hit dice or something similar to account for size. HD needs to align with strategic significance, thematic significance and technical complexity rather than the amount of meat in a monster, IMO. Same thing as with statting whales, it's just vapid for a whale to rank at the same level as Orcus the Demon King just because it has a lot of blubber and you gotta make sure those 1d6 sword-bites don't sting it to death in the combat simulation.
  • You could do something like roll 1d6 each day you're traveling. On a 5+, you get an "encounter."

    1: Fellow Travelers (1-2 friendly, 3-5 neutral, 6+ hostile).
    2: Landmark (1-2), Vista (3-5), or Sign of Civilization (6+).
    3: Change of Scenery (1-3 weather, 4-5 biome, 6+ season).
    4: Hazard.
    5: Dungeon (1-2), Lair (3-4), or Ruins (5-6).
    6: Monster!

    Then you could separate them based on environment and external factors. You might then have something in the vein of roll a 5+ (encounter), then a 4+ (hazard), then on another 1d6 table to determine its nature.

    Environment: Forest
    Season: Summer

    Sample Hazards:

    1-2: Heat wave. Everyone must make a Constitution save or gain a level of exhaustion.

    3-4: Humidity. Streaked with sweat, everyone must consume a full skin of water or become dehydrated.

    5-6: Insect swarm. Biting flies, gnats, mosquitoes, etc. inflict 1 HP of damage and slow pace by half.

    In practice, you roll a fistful of d6s and interpret accordingly.
  • edited June 29
    Hey those raptors only downed one PC not the party


    I gotta admit I'm not sold on the separation you advocated. As Johann noted, it seems teleological. While heaven knows I do advocate prepped 'salient' components and improvised color, if you are going to do sublists I'd rather do… hmm, how to explain it… this is an idea I had a while back, three years or two, but haven't tested yet.

    'Overlays', that is to say you have an encounter table. Then the region might overlay some of the entries with region specific. Then the hex in turn could overlay even more, with hex specific. Randomly rolled but post-hoc fixtures, like 'lairs' or other rolled geographical features, could have its own slot as well. In the end I burnt out on the project.
  • The fundamental issue is that the game itself has a teleological base reality: it is being played for a purpose. This causes a situation where an absolutist commitment to neutral objective simulation is not as useful for successful play as a controlled golden mean is. That is to say: the question is not how we can remove all intentionality from an adventure game sandbox, but rather how we can put that intentionality to work in service of the game.

    In the specific, the difficulty I face constantly in D&D world simulation comes in the fact that the game-related demographical statistics that an adventure game needs have almost nothing to do with realistic distributions, and therefore "how many bandits live in a square-mile of medieval forest" is not truly an useful factoid for sandbox design. In a realistic world you do not have three ruined manors within 20 miles of each other, and those ruins certainly do not all have anything even resembling treasure in them. For the purposes of D&D, however, this is all completely reasonable, and that's because the game's goal has embedded teleological intentionality to it: what you prep has to be "gameable", whatever that means for your campaign. Usually it means having the adventuring party encounter much, much more adventure-relevant content than an average tourist ever would.

    I often phrase this base reality by explaining the game's downtime and preparatory activity phase as a stage of negotiation during which the GM and the players provide intentionality to the game table: they selectively provide interesting game content in the form of adventure hooks, adventurous characters and curious happenstances. These are often randomized in many ways, yes, but it's not an attempt at neutral world simulation, as we would only find that 1% of outcomes that actually contain adventure potential to be acceptable. Rather, the random elements are there to provide surprise and unexpected combinations over a content base that has already been selectively prepped to be adventurous, no matter the recombinations.

    This is what's going on in the way I categorize sandbox content in that particular project: by formalizing the content into functional categories (as per expected relevance to adventurers) I enable certain useful overarching rules to exist: 50% of travel steps will trigger the GM to open his mouth, and 66% of those times will be essentially low-stakes atmospheric discussion, while 33% will be serious business of some sort. This is no more teleological than your choice to sit down to play a given adventure module, or the choice to design a module that actually has monsters and treasure in it, or the choice to skip to the next morning because nothing interesting is happening to the adventurers today. It's just content management, pacing the arbitrary amount of swamp fantasy imagery and situations against the number of moves the players choose to make on the map.

    There is a hygienic advantage to not separating color encounters from serious ones during play, as a GM who doesn't have those categories explicitly in their notes will be less likely to have biased conception of how the encounter should go. However, there are serious disadvantages as well: you either have to choose to not have color encounters at all, or you have to write them all out in advance into the encounter tables. You cannot improvise the color encounters from short-hand notes, either, without having the GM be aware that this is what they're doing.

    Ultimately this seems like a subjective technical issue, as the best way to go about preparing and delivering the material depends on the GM's skills and inclinations. For example, if somebody simply does not provide color descriptions at all, then this entire concept of "color encounter" is just a distraction. I couldn't claim that this or that way is objectively the best approach, but we can make note of the advantages and disadvantages of various techniques.
  • edited June 29
    Sorry, still catching up on earlier comments! I'll get to the more recent stuff in a bit.

    @yukamichi those are fantastic examples!

    Armies and dragons avoided or glimpsed in the distance -- that's exactly the kind of stuff that I love to emerge from travel. Story hooks in an evolving world, yes!

    I have a tangential thought that I'm putting in its own block:
    I'm not sure how to generate world-significant content randomly. If, as GM, I roll something very local or transient, I'm allowed to forget about it. If I'm rolling army movements, though, then I feel like I need to keep track of that situation. I've never liked that "keeping track". So if I'm going to use a randomizer that might impact the world, I'd also like a tool that makes tracking such stuff easy.

    I suppose one solution to this is just to use whatever sticks in people's memories, and not worry about what doesn't. If my group remembers the dragon but forgets the orcs, well, maybe those orcs are fighting somewhere very far away that doesn't impact us.
    As for random tables for different timescales, some of which include shifts in the larger issues of the setting (factions, govt), I like the sound of that.

    To be honest, I've always had pretty solid grasps on the worlds I've GMed, so the idea that the world itself could emerge from random rolls over time is foreign to me. It sounds fun, to share that element of exploration and discovery with the players! It'd probably take some getting used to, though.
    yukamichi said:

    not everything needs to be a hook, not everything you mention in the game needs to be explored
    . . .
    I think that part of why this kind of encounter design fails is because of this idea that everything must be meaningful or else it wouldn't be there. It's a hard metagame issue to get past because it generally conflicts with a social contract point about wasting people's time.

    Yeah. I think the ideal is to have a clear and simple way to communicate to the players what their options are and why they'd pick one over another. Honestly, I don't think most players have much trouble (a) soaking up a colorful description for its own sake and moving on, or (b) investing in a random challenge with an obvious level of threat or reward. I think it's the in-betweens that get tricky. The "there might be something here" encounters are the key thing to support, IMO.
  • Eero, what I meant to say that effort spent on this teleologically based separation doesn't seem to pull its weight, compared to other layering.

    And as I did say earlier, if writing them out in advance is the problem, just put "1 to 30: color" on there. And if you do come up with some good or important color, that's where they can go.

    Advantages of this:
    1. Solves that 'the players can't look at the initial D6' problem
    2. Makes it more compatible with other layering
    3. Gives you room to add a random and non-arbitrary chance for recurring or re-occurring color, like those goats they liberated from someone's pen

  • Every day they need to drink water and deal with insects, VSK, not just sometimes. That's part of the routine: roll for weather, roll for diseases, roll for navigation, then encounter rolls. Then in the evening get exhaustion for those that didn't drink enough. Then night time encounters.

    I was just saying the other day… it's been 13 sessions and zero progress in the module. They haven't found any major sites. Just jungle daily life (+ intra-party soap opera that puts Lost to shame). But there's been a lot of progress in their skill. Like, after the first night they discovered the hard way that they needed tents. So they headed back to town, and got lost on a weeklong detour. That's how they realized that having a bard to inspire those nav rolls are probably a good idea. (Now they have a ranger, even better.) Same for water, diseases, insect repellents, clerics, tanks etc. The jungle is a challenge and… there is constant progress in them developing routines and strategies to survive it.

    And, the players get better, not just the PCs. PCs are dying but the new PCs are more savvy. (Unlike my glitchworld campaign where the players sent three parties to their deaths in the same trap because they "didn't want to metagame".)
  • The problem I'm trying to solve is more about the timing of when it is appropriate for the GM to bring content to play. For immediately significant encounters this is carefully regulated by encounter frequency, which is set in stone at content design stage, same as dungeon room contents and whatnot. For less immediately pressing types of content the procedure is vague: you might infer that the GM is never allowed to improvise fun tidbits unless provoked by e.g. players asking questions, or you might decide that the GM is always allowed to do it freely. Both extremes have subtle problems and require using judgement. For me the tendency is that I don't tend to do enough of color content over time (in my own judgement), which is why I like the idea of a procedure that reminds me to do it with appropriate pacing.

    I'm sure that's not a problem for everybody.

    And yes, I would get almost the same benefits from a e.g. d100 random encounter table that just happened to include a lot of empty "ad-lib some color stuff here if you feel like it" slots. If I determined in advance that slots 1-33 are real encounters (maybe just five of them, as in my example) and slots 34-100 are color encounters, it would be precisely the same arrangement of odds, but worked into a single table instead of a couple of consecutive rolls.

    I'm not fond of that one big fat table thing, but that's strictly an issue of mechanical aesthetics - it feels messy to me compared to smaller numbers more deliberately used. Probably something to do with playing MERP during the '90s, that game has some very ugly d100 tables. A cascading process through multiple small tables feels prettier to me - easier to fill with meaningful content, resulting in ideas coming through more clearly.
  • Your last two paragraphs show that you know it's the same odds so your first two paragraphs feel kinda non sequiteur.

    To me, hiding the initial d6 roll goes against Transparency of Method.
  • I suppose I could do a random encounter check in the open and then another secret check to see if it's really a random encounter [grin]. That would achieve the desirable effect of curtailing the player meta knowledge about whether there's a real encounter here or not.

    Or, we could just use random number sheets like real competitive gamers, allowing post-game review of the GM's rolls. No need to roll in the open when the players can check the accounts afterwards.
  • I suppose I could do a random encounter check in the open and then another secret check to see if it's really a random encounter [grin]. That would achieve the desirable effect of curtailing the player meta knowledge about whether there's a real encounter here or not.

    Or, we could just use random number sheets like real competitive gamers, allowing post-game review of the GM's rolls. No need to roll in the open when the players can check the accounts afterwards.

    I started a new thread about this. But my hypothesis is that open rolls increase buy-in.

  • To be honest, I've always had pretty solid grasps on the worlds I've GMed, so the idea that the world itself could emerge from random rolls over time is foreign to me. It sounds fun, to share that element of exploration and discovery with the players! It'd probably take some getting used to, though.

    It's definitely an enjoyable and exciting way to play! I've played in a handful of games like this, though I've never GMed this way myself for an extended period of time. Very much worth exploring! (Perhaps we could do some of this with our online group.)

    I think you hit the nail on the head when you mention the difficulty this poses: how do you handle the "switchover" between randomly-generated content and a more and more defined and/or solid "game world", where the boundaries of what's possible or plausible grow smaller and smaller?

    You write about that here:


    I have a tangential thought that I'm putting in its own block:

    I'm not sure how to generate world-significant content randomly. If, as GM, I roll something very local or transient, I'm allowed to forget about it. If I'm rolling army movements, though, then I feel like I need to keep track of that situation. I've never liked that "keeping track". So if I'm going to use a randomizer that might impact the world, I'd also like a tool that makes tracking such stuff easy.
    I agree that this a bit of a challenge, at least in some cases. Are there any good tools or techniques for working with this kind of thing out there?

    Honestly, I don't think most players have much trouble (a) soaking up a colorful description for its own sake and moving on, or (b) investing in a random challenge with an obvious level of threat or reward. I think it's the in-betweens that get tricky. The "there might be something here" encounters are the key thing to support, IMO.
    Yes! This is the other major challenge.

    For Eero and others who have done this extensively: do you have any at-the-table techniques to make navigating this smoother or easier?

Sign In or Register to comment.