Overhauling the wilderness encounter system

 The D&D wilderness encounter tables are a hot mess. That's the message from Eric Diaz at Methods & Madness, and he's absolutely right. It's a subject I've pondered a lot over the years, and I have some thoughts on the matter too. 

The post linked above covers the problem quite well, and I'd encourage you to read it. In brief, the tables are both wildly dangerous to adventuring parties, featuring ridiculous frequency of very powerful creatures such as dragons, and the encounters often aren't even terrain-appropriate. My own two cents' worth in the comments was that many of the encounters end up being utterly pointless as well, such as sea snakes on shipborne adventures.

The next step in finding a solution is to identify what we'd like to see. 

Do we want random wilderness encounters at all? I'd argue that yes, we do, because hex crawls almost by nature consist of large swathes of unkeyed territory with nothing but a terrain symbol in the center of a six-mile or larger hex. In lieu of the daunting and time-consuming prospect of keying everything (which still wouldn't come close to describing everything to be found in a single hex!) we do need some means of discovering what's there in the course of actual play. 

Next, it's probably desirable to have creature appear according to terrain. We want bears in the woods, dwarves in the mountains, black dragons in the swamp, and mermen in the sea. We don't want bears in the sea, mermen in the mountains, dwarves in the swamp, and black dragons in the woods. 

Also -- and this is where B/X and BECMI D&D fall down -- I want varying degrees of difficulty which players can anticipate and plan for. In the dungeon, dungeon level telegraphs the level of challenge players can expect; the danger increases every time you go down a level deeper. Something similar is sorely needed for wilderness encounters, too. "Low level" wilderness areas should be populated by creatures you'd expect the average town guard contingent to be able to handle with relative competence -- normal animals, goblins, giant rats, and such. Very powerful monsters like dragons and giants should rarely or never be encountered a stone's throw from the village square; in fact, they should generally be relegated to the most remote and untamed parts of the map. 

All, or almost all, of the encounters should be relevant and usable in common contexts. For instance, sea snakes and sharks, though appropriate to the Ocean terrain type, are superfluous if the adventurers are traveling by ship. (We can largely solve this issue by creating additional terrain types, e.g. by splitting "Ocean" into "Ocean Surface" and "Undersea" so as not to afflict a party on a sailing adventure with day after day of, "You see some snakes swimming in the water below. Nothing happens.")

To summarize all of the above, and echo Eric Diaz's post, we should very rarely, if ever, have to ignore or reroll a rolled encounter because it's inappropriate, superfluous, absurd, or wildly unfair. And yes, it is absolutely unfair to have a party's fate hinge on a single reaction roll when they encounter a group of four red dragons completely at random with no chance to avoid. A 1st-level party choosing to explore an area known to be red dragon habitat is on the players; encountering red dragons in generic wilderness just because red dragons are on the tables for all wilderness areas is agency-denying BS. The TSR-era stock rejoinder that "wilderness adventuring is especially dangerous" is both nonsensical (It's more dangerous traveling from one village to the next than it is to prowl about in dank, dark places where the sun never shines? Please!) and a flimsy fig leaf to cover the fact that they were either unwilling or unable to construct a viable challenge-indicating framework for the wilderness as they did for the dungeon. 

It would also be good to have tables that feature more than just monsters. As noted above, it's impossible to key all the things a party might stumble across in even a single hex, and there ought to be a lot more out there to be discovered than just monsters. There should be rock formations that look like knights from afar, enchanted trees, wishing wells, fairy rings, mysterious obelisks, magical pools, ancient statues, waterfalls, oases, quicksand pits, sheer cliffs, forgotten burial grounds, lost battlefields strewn with rusted weapons, and all sorts of other stuff. 

Finally, the whole system should be easy to use at the table, without excessive dice rolling or flipping through pages in a rule book.

Any of these goals by itself is relatively trivial. We can easily construct tables for different terrains (though TSR did a slipshod job of it, it's not hard to imagine doing it right). Likewise, we could easily divide up our wilderness map into difficulty zones, either by counting hexes of distance from safe zones like cities and towns or by doing it in a more ad hoc fashion, and create leveled tables for each zone, and we could easily rustle up tables of non-monster encounters. What's difficult is integrating all of them in a format that doesn't sprawl over pages and pages of mind-numbing, hard-to-navigate tedium. 

The non-creature encounters will probably have to remain separate, but we can do a lot to simplify the monsters. The most obvious way I can think of to do so is to abandon the nested tables of Basic D&D. To stick with that method, we'd have to create separate tables for each terrain at each difficulty level, meaning potentially dozens of tables to sift through at the game table. Instead, we can create a single table for each terrain type, with low level monsters at the bottom, graduating slowly to stronger and stronger creatures. Each difficulty level would then use a different die or combination of dice, with safer areas using smaller (or fewer) dice and progressing to larger and larger (or greater numbers of) dice. Thus, the easy zone around most settlements might start off with a d6, and progress to d8, d10, and so on as the party ventures farther into wilder, unsettled, or unpatrolled territory. Multiple dice (e.g. 2d8, 3d8, etc.) or adding pips (d8+2, d8+4...) could be utilized instead to slowly phase out weaker monsters. Perhaps the best way would be to make it completely intuitive by simply rolling the same type of die once per level of danger, i.e. 1d6 for level 1, 2d6 for level 2, 3d6 for level 3, and so on. While each such table would be rather lengthy, it would spare us a ton of repetition and make finding the correct table a relative breeze: better to have a dozen tables with 30 or 40 entries each than a hundred tables with a dozen entries each. 

As a roleplaying adjunct to tables of this sort, it would be advisable for the DM to place landmarks or other signs on the map to warn adventuring parties that they are crossing from one zone into a more dangerous one. This could be as obvious and on the nose as an actual signpost ("Danger! Owlbear sightings beyond this point!"), or an unmistakable dividing line like a river, ridge, or valley. (Remember the villagers in the original Dragon Warrior video game saying, "Beware of bridges! Danger grows when thou crosses.") It could be something subtler, like the woods becoming darker and more tangled, or the terrain becoming rougher, or even direct signs of powerful monsters, like scorch or claw marks, petrified victims, etc. For savvy players, you can provide the information in forms that have to be ferreted out through role-playing and information-gathering, such as  conversations with local NPCs ("Be careful if you go past the Great Oak; that's bugbear territory.") or in written works like journals, histories, or town records. 

One other technique we might use to mitigate the capriciousness of random encounters and augment the opportunities for player agency (whether or not we choose to scale wilderness zones like dungeon levels) is to reinterpret the rules for the surprise die: instead of surprise on a roll of 1-2, and all else being a standard encounter, a roll of 1 equals surprise, 2-3 means the monster is sighted directly as normal, and 4-6 indicates the party sees (or hears or smells) signs of the monster's presence, such as tracks, droppings, scorch marks, or whatever seems appropriate to the monster type, 1d6 turns before actually encountering it. It is, after all, much easier to spot clues, and much more difficult to surprise someone in an open, well-lit, spacious environment than it is in the dark claustrophobic confines of a dungeon maze.