You really need to read Gus L.’s blog. Across a series of informative and enjoyable essays, the man has been gradually detailing the “Classic Dungeon Crawl” style of play (and how to apply it to more recent RPGs).
I find his argument about the utility of location-based play particularly compelling:
“[T]he dungeon is first a simplification and a gamification – a game board that strips away many of the complex problems of an open fictional world… As a game ‘board’ rather than a ‘structure’ or ‘narrative’ the dungeon first creates spatial puzzles – its mechanics and principles often relate to how best to move through it as a location. With a spatial orientation in its design (the map being far more important in a Dungeon Crawl then it is a scene based game) the spatial puzzle leads to exploration, a game where it’s valuable for players to determine and understand the physical layout of the dungeon: entrances, exists, regions, locations of interest and the interrelations between them.”
I completely agree with Gus. In fact, I’ve been wondering if it’s possible to turn the entire game into a location-based game board akin to an enormous dungeon (an approach I’m calling Dungeon Mode). And, rather than looking back to classic D&D for inspiration, I’m hoping there might be lessons to be drawn from early Computer Role-Playing Games (particularly the dungeon crawler genre), although gamebooks and boardgames might also supply inspiration.
How, for example, did CRPGs like Wizardry and Might & Magic approach some of the challenges of presenting maze games to players while offering them meaningful choices, variety, and interesting risks vs rewards? The Digital Antiquarian has a fascinating post about how the strict grids of early dungeon crawlers (also known as “blobbers”) promoted “emergent interactivity”. Could we adapt that to the tabletop?
I suspect (though I have no real evidence) that tabletop gamers have been unwilling to copy the lessons of CRPGs partly because of “negative ludological influence” (i.e. the desire to differentiate tabletop RPGs from CRPGs, and instead emphasise the qualities of tabletop RPGs that are most difficult to simulate with a computer).
Tabletop RPGs are great at collaborative storytelling. Perhaps that encourages a scene-based structure instead of a location-based ‘game board’. Meanwhile, CRPGs (particularly modern CRPGs) excel at simulating detailed environments. Yet older CRPGs (i.e. from the 1970s and ’80s) were highly abstract and “gamified” in the way they simulated locations due to technical limitations.
Why, therefore, would anyone try to copy the experience of a classic 1980s CRPG in a tabletop game? Surely classic CRPGs are outdated, having been superseded by much more realistic simulations? Perhaps the relative crudeness of classic CRPGs as simulations actually makes them *more* interesting?
Gus’ classic dungeon crawl is a location-based approach. However, I assume he would still apply a distinct “Wilderness Mode” (traditionally a hex crawl across an overworld map), a “Town Mode” (typically – though not necessarily – a shopping list plus a menu of downtime activities, with the occasional random encounter), a “Dungeon Mode” (the classic dungeon crawl), and an even more granular “Combat Mode” (a tactical wargame encounter, either on a battlemat or in the “theatre of the mind”).
Rather than differentiating between modes of play, however, I want to try running everything in the same mode: Dungeon Mode. From tromping through the wilderness to exploring towns and cities, everything should be handled as if the world is a massive, seamless dungeon. Having said that, I would, I think, keep Combat Mode for tactical encounters.
Dungeon Mode needs playtesting to see if it will actually work. Here are some of the principles I’m going to experiment with:
1. Space is constant in Dungeon Mode. The aim is to have the entire campaign world be as simplified, gamified, and abstracted as a dungeon. You can see an example of what the world will look like here. My “world maze” will be composed of 10′ x 10′ x 10′ cubes of space (which could be sub-divided into 5′ x 5′ x 5′ cubes for Combat Mode), in a similar fashion to Minecraft. I would use sheets of graph paper to map the X and Y axis, whilst assigning different sheets to different depths on the Z axis. In that way, it will be possible for the party to explore a three dimensional space seamlessly, with the town connected to the dungeon, etc. This opens up interesting options in terms of tunneling and excavation, for example.
2. Time is constant in Dungeon Mode. Time passes at the rate of one turn per the party’s movement rate in 10′ cubes. I see no reason to have both a base movement rate and an encounter movement rate (particularly as smaller numbers are easier to manage). So, everything in the game would move at its encounter movement rate. I am considering allowing the party to “fast travel” between locations they already know. However, in this case they would need to make a “Losing Direction” roll based on the most challenging terrain cube they pass through, and could find themselves lost in the “world maze” having taken a random wrong turn. I’d like to avoid fast travel, if possible, but we’ll see if my players can endure crawling 10′ cube by 10′ cube literally everywhere in the world (“North, North, North, East, North, East, East, East”, etc.) without rising up in revolt.
3. The PCs move as a “blob” in Dungeon Mode. One of the lessons of the “blobber” genre of CRPG dungeon crawlers is that you cannot split the party. The party moves, fights, and interacts as a group. I’m thinking of having a single “Caller” responsible for driving the party around the world maze, calling out the direction of travel to the DM (though each party member would still be able to take individual actions within the “blob”). The position of Caller could rotate between sessions, so everyone gets a chance to be in the driving seat. However, when the game drops down into Combat Mode, then the party can still split up across the battlemat in order to fight tactically (joining back up again into a blob when Dungeon Mode resumes).
4. Everything is finite in Dungeon Mode. Small numbers are more interesting than big numbers. Scarcity is more interesting than abundance. If an object or NPC is not keyed, then it does not exist within the world maze. All monsters are individually keyed (e.g. all kobolds in the world maze would be individually designated K1, K2, K3, K4, K5, etc.). If kobold K1 shows up both in a room key and in a random encounter table, then it is the same kobold. Therefore, if the party were to kill K1 in a dungeon room, then K1 won’t ever show up via a random encounter. Likewise, shops do not have infinite stock (in fact, they will probably only stock a couple of items in total).
5. Threat level is constant in Dungeon Mode. Navigating towns is, potentially, just as dangerous as navigating a dungeon (with thieves, robbers, cultists, even traps). Think the Fighting Fantasy gamebooks City of Thieves or Kharé: Cityport of Traps. There are also opportunities to earn treasure and loot in town, and there are secrets and hidden locations to be explored. At the same time, not every encounter (in fact, perhaps not even the majority of encounters) need to end in combat. Diplomacy, stealth, subterfuge, faction play, and running away are just as important in Dungeon Mode.
In terms of running Dungeon Mode at the table, I’m thinking of borrowing from Stonehell’s approach to layout. I would compile a DM binder comprised of two-page spreads, with a graph map on one page and a dungeon key on the other. Everything needed to run the map would be included on the key, and anything not on the key would not exist on the map. To me, this would be a good example of dungeon-as-code.
Next, I need to try and playtest these concepts to see if they are actually fun.
Images from Wizardry: Proving Grounds of the Mad Overlord, Sir-tech Software, Inc., 1981, and Might and Magic: Book One, New World Computing, 1986