Wednesday, 28 August 2013

Random Zonal Map Generation for Random Encounters

A while ago (it's actually over a year ago; fuck me, I'm getting old) I posted an idea of how to use Diaspora-style zones in D&D or other games. It's easier if you actually go back and read the post than have me precis it here, but I expect that you are too lazy to do that, so here is the main body of the post reproduced for your convenience:

When combat looks likely, the DM draws a map, and then divides it up into zones. Zones are not a grid, and do not represent strict distance. Rather, they represent a combination of space, ease of travel, view, and time to pass through. Thus, an open field might be one zone, but a nearby cottage with three rooms might be made up of three separate zones, because the time it takes to sprint across the open field is the same as the amount of time it would take to move between rooms cluttered with furniture. Likewise, a thick forest full of boulders might be divided up into three zones, whereas an area of light woodland of the same "size" might just be one zone - representing the speed with which one can move through each area.  
With me so far?  
If there are borders between zones (a hedge, a wall, a ditch, a door) these have numerical ratings indicating how long it takes to pass through them. So a hedge might have a rating of "2", indicating that it takes 2 "zones" worth of effort to cross over. (You can climb over in the same amount of time it would take to cross 2 zones.) This is called the "pass value".
Pass values can change if there is a doorway. If there is a doorway, the pass value is 0 - unless the door is shut, in which case it costs "zones" to open it, indicated by a number, called the "opening value". (A wall would cost 2 "zones" to climb over, so it has a pass value of 2, but it has a gate which would cost "1" zone to open, or an opening value of 1. So the hedge has a pass value of 2/1.)  
Each turn, the player gets one zone of movement and an action. The zone of movement might consist in eroding pass value (e.g. getting half way over a hedge with a pass value of 2), or opening a doorway (e.g. eroding an opening value from 1 to 0.) The action would be the usual sort of thing (cast a spell, attack, whatever). Or he can give up the action to make two zones' worth of movement.  
A turn is however long it would take to cross a zone. This means that combat can scale up and scale down to suit the situation. A fight taking place in an area covering a square mile of countryside, or taking place in a few rooms in a hotel reception, would follow exactly the same pattern - it's just that the zones would represent different levels of abstraction. The zones in the former would probably be bigger by area on average, and a turn would be longer (maybe two minutes in the former as opposed to around 10 seconds in the latter). But the rules are exactly the same. Only two things would really change. First, the larger the zones and the longer the turns, the more abstract attack rolls (the attack roll would represent a period of maneuvering and trading blows), and the smaller the zones and shorter the turns, the less abstract attack rolls would become (the attack roll would represent literal single attacks). The other thing that would change would be pass value. A hedge separating two open fields, each about 100 square yards and represented by a zone each, would perhaps have to have a higher pass value than a hedge separating two small gardens of 10 square yards each.

Did you spot the deliberate mistake? The last sentence of the last paragraph is clearly me talking out of my anus. If the approximate average size of a zone is 100 yards, then a hedge separating two such zones will obviously have to have a lower pass value than a hedge separating zones where the average zone size is about 10 yards. If you can tell me why, you win the prize of one coconut, delivered for £27.99 P&P.

Anyway. Talysman's excellent subhex crawl posts got me thinking about ways to generate an immediate wilderness area for wilderness encounters. One thing that I've always found dissatisfying about random encounters in D&D is that they never really pan out in a manner which I think remotely reflects reality. I do quite a bit of hiking, so I know that wild places are extremely interesting and varied, and full of gaming potential that is largely untapped.

By which I mean: I know how to run a random encounter in the wilderness. I just think there could be much, much more to it than D&D currently envisages.

So let's mess around with some ideas. I've not devoted a great deal of thought to this - I'm just ruminating out loud - but bear with me.

1. When a random encounter is rolled, grab a piece of paper and a pencil as you normally would.
2. Pick up some d6s and thrown them on the paper. Four or five will do. These represent areas of the same terrain that are going to be subdivided into zones. The d6 position indicates the rough centre of the area on the map; the number on the dice represents the number of zones contained within it. From this, you can extrapolate terrain type and size, with a large number of zones indicating "denser" terrain (e.g. 1 = totally clear and unbroken terrain, 6 = dense deciduous forest) or a larger area (3 zones of flat grassland will be much larger than 3 zones of thick swamp).
3. Roll a couple of d4s. (I'm stealing this from Talysman.) Light coloured d4s indicate elevation above the average value. Dark coloured d4s indicate lower elevations.
4. Now everybody at the table (DM and players) roll 5d6 each. If they get a pair, they get to place a terrain feature: 1-2 - Hedge, wall, lake or river, 3-4 - Gate, bridge, cave, fallen tree, etc. 5-6 - Building or special terrain-appropriate feature. If they get three of a kind, they can place two features. And so on.
5. The players then roll a d4 to determine which quadrant of the map they begin the encounter in.
6. The DM decides where the encountered creatures are, based on encounter distance and surprise. 

It feels fiddly. But I think you get the aim, which is to create an actual physical wilderness environment on the fly which doesn't just come down to what I always do, which is jotting down a shitty map off the cuff. Jotting down a shitty map off the cuff is okay, but there has to be a better way. Moreover, I like the idea of giving players input into the map itself - the crucial caveat being, the players won't know where they are on the map until after all the terrain is placed, and they don't know where the enemy is either.


  1. Yeah, I throw dice or just draw random junk for a wilderness tactical throwdown. A picture of the surroundings usually comes pretty readily. Players can maneuver if they want to look for a specific feature.

  2. Permit the chuckle over what is in your wilderness. Hedges?

    I too am a bit dissatisfied with wilderness,but mostly because it has an absurdly cheerful idea of cross country travel, as if all woods are created equal or for that matter as if all streams/gullies are as well. Sometimes a stream is a path, other times it's barrier. I admit it, I once took some people on a two mile hike to prove my contentions about roads, types of woods, and grasslands. Because waist high grassland in the wrong season is a major chore, while heavy woods makes it easy. That a creek can be the fastest way to move, but a bog can take hours. That denser woods means faster travel and light woods means slower travel, because it's about undergrowth. And that is without dealing with grade, as even a lightly travelling party is burdened with more than a backpacker. I wasted a lot of time trying for a detailed system, which ended up being more work than just spending an hour drawing up random encounters ahead of time. and pulling them at random.

    I think the only way to really do it randomly is to ignore getting descriptive at all, looking at it as purely mathematical. Rolling the abutting hex movements as you enter each hex, without trying to qualify this is a thicket, a stream or a steep hillside.

    Roll an eight-sider for each hex, with a one being an "avenue" that is a game path/creek/hollow. 2-7 being set difficultly to cross and an eight being a barrier, a wall/swamp/deadfall, which stops the party. On a 1 or 8 you roll a six-sider for orientation. For a avenue, 1 being spot on, 2 left, 3 right, 4 extends in one direction, 5-6 in both (rerolling until you get a 1-3, thus you might draw several hexes of avenue at once). For barrier, you draw from the points, 1 horizontal, 2 left, 3 right, 4-5 extends in one direction, 6 in both.

    So as you walk into a new hex, you roll at most five 8 sided dice and say four or so six sided (figuring 12.5% will need an extra, 6% two extras, 3% three,etc...) you write down the adjoining hexes, drawing in barriers and avenues. It takes the time to roll a handful of dice and write roughly seven or so figures.

    1. I use wilderness in the traditional sense of overland hexes - i.e. distinct from dungeon.

      I feel like somebody needs to do what Patrick S is trying to do for caving for overland travel.

  3. Ah, yes I see what your saying; you'd want to have smaller barrier ratings because you're measuring the rating in "zones", meaning that as you scale up a hedge would be as hard to cross as a field.

    This means that certain barriers would have to be handwaved completely, as they'd feasibly be in the region of 0.05 zones worth of movement. But even if you get rid of that idea, and have a certain number of points for passing obstacles or something, then a larger zone means a larger time scale, so you'd just naturally have more time to deal with obstacles, meaning that the ratings would have to be lower from that side too.