Nameless Bodies, In Unremembered Rooms

Gustave Doré (1832–1883), “The Fortress of Fear”

illustration from “The Days of Chivalry; Or, The Legend of Croquemitaine” by Ernest L’Épine (as Quatrelles), 1866

Gustav Adolf Mossa (1883–1971), “The Satiated Siren”

I think it would probably have general dungeon-crawler vibes, although skewing towards sci-fi flavored (while the source material doesn't explicitly satet it, something about the name, the fake game cover, the megastructure nature of the labyrinth, and the general vibe of the story make me feel like it's more of a sci-fi than fantasy game)

The only thing where I see myself leaning toward the spooky-dookies is maybe leaning into a weird an intentionally inconsistent tone? Like my idea would be to have the way the mechanics are described range from standard military wargame tone to coldly clinical to occassionally subtly ritualistic.

I haven't really given it much thought beyond the "I wanna do this someday" level until recently, but I kinda have some ideas about it.

The end result would be more of an art-project than an actual thing you'd probably want to olay, it should be playable but it's explicitly meant to be an extremely archaic ttrpg from 1965 and The text explicitly mentions that Rolemaster was a "simpler" game to play by comparison (although this might be a tongue-in-cheek remark from our point-of-view character and not completely literal) so it should be overcomplicated, overengineered, and extremely cumbersome to actually play.

It should also eschew not only modern design sensibilities, but also a lot of *traditional* design sensibilities. Since it's supposed to be a game that predates the original edition of D&D by almost a decade, any concepts invented by D&D should be either entirely absent (preferrable), or visibly independently invented from first principles.

It should be almost entirely table-driven, with most actions resolved by rolling on a table, and probably with most tables directing you to a different table. The text explicitly mentions "hundreds" of tables modified by bonuses from everything from surrounding rooms to recent player actions, so that's going to be fun to figure out.

The creepypasta mentions that the game uses a 6d7 roll plus a coin flip to generate results with an "inverted bell curve" probability distribution. My first idea for this was something like rolling a 6d7 and flipping a coin, on heads keep the highest result, on tails keep the lowest result, but I realized this was probably going to be hard to work with, because it would result in the roll *very rarely* showing anything other than 1 or 7, and a probability space of 1-7 is kind of limiting for a table-driven game. So my idea rn is doing the same but instead keeping the 3 highest or 3 lowest dice depending ok the coin flip, with maybe some special effects for special results such as matching numbers (1 1 1, 2 2 2, etc.) and consecutive numbers (e.g. 1 2 3)

The thing that I think is going to be hardest to pull off is the spawning of the "exit hex", the story explicitly mentions that there are strategies for maximizing the chance of generating it, but that the reason no one has been able to do it is because any attempt to optimize for it also increases the chance of intervention from a sabotaging presence that the game's play culture has collectively personified (and probably unintentionally brought into existence by themselves through decades of houserules) as "the minotaur".

This necesitates that 1) the chance of spawning the exit hex is influenced by a complex web of factors affected by player actions, and 2) that the same web of factors also influences the "minotaur"'s activity.

E.g. to get the exit you might need to roll a 24 on the Primary Hex Generation Table, which takes you to Hex Generation Sub-table x on which you need to roll a 19 to get to the infamous d777 Special Hex Table (which gives you an unmodified 1-in-343 chance of getting the exit). In order to maximize your odds of getting a 24 on the PHGT and then a 19 on HGS-T x, you need to get as far from the entrance as you can, collect a trinket from each hex-row you've visited, play a reed pipe before you open the door, etc.

But if you get a 23 on the PHGT, you go to HGS-T w, which is likely to lead to a trapped hex if you're far enough from the entrance. And if you roll a 17 or a 11, you go to HGS-T q or k respectively, which both have rooms with a chance of being dead-ends because of a locked door or blockaded passage—more likely every time you collect a trinket. And if you roll a 28 you go to HGS-T β, which is pretty safe—unless there's food scarcity in too many map sectors, in which case it's likely to generate a hex with a frayed rope bridge over a deadly chasm or with every surface covered in broken glass. And of course each time you play the reed flute, there's a small chance to increase Rodent Disturbance in that sector, which increases the likelihood of generating food scarcity when you roll on the Vermin Migration Table you mentioned.

although. my conception here is probably a bit difficult to manage with the player physically rolling tables. wouldn't be too crazy to have a system of rolling on a bunch of tables to generate the dungeon as you go, but doing that in a way that prevents the player from just looking through the tables to find what generates the exit is the hard part. obvious answer there is that it could be a game with a GM. but that's a bit of a standard design convention, so i have a second proposal:

Take all the duties normally assigned to a GM and distribute those duties between several players. I don't mean "you run the monsters, i'll roll the tables, you play the NPCs" type shit. I mean like "you own this batch of tables, i own this batch of tables, you own this batch." each facet of GM stuff gets divided between all the players, so that no one person is always in charge of the same sort of thing. in a fight, the players are playing both their characters and a monster each sorta thing. when it comes to tables, i think you could maybe even let each table have its own rules for what about that table is and isn't allowed to be shared with the other players.

another idea is copying the NPC emotions and roles system from Exanima. in that video game every npc has a current emotional state and a bunch of 'roles' assigned to them, both of which effect each other. the roles come together to dictate behavior. its conceptually pretty flexible. roles can be pretty simple or they can be more complex. so you could, in theory, have NPC behavior be entirely dictated through a series of enormous tables where various roles determine where 1 on that table starts. or maybe just add or subtract from your final result but they often have really big numbers so like a d100 table has results going all the way to 1264 or something. For example, in the Combat Decision Matrix (Ready) (ready being their current dominant emotional state), the lower end of results could have stuff like "flee" and "avoid the fray" be more common, and the NPC role "Coward" lowers your result on the CDM(R) table by some amount. But if that NPC also has the Knight role, maybe that increases the result, so that NPC usually winds up with results that show a more average amount of bravery.

I do want to say that things such as "being difficult to manage with the player physically rolling" or "requiring way too many tables" are not dealbreakers for me for any idea I might implement in this project, I'm walking into it with the idea that the end result of this project won't be anything other than *very* intentionally cumbersome to play, and I fully intend to take the "several hundred reference tables" part of the source material extremely literally (to the extent that, rather than having any centralized resolution mechanic, I intend for any action the player character might attempt to be resolved through its own specific dedicated table).