No, seriously. Here you are.
The first 21 rooms look like this;
![]()
Ugly, but functional.
Here's how it works. This is basically the Moldvay dungeon stocking tables streamlined down to a single d1000 roll, so that you roll once and get the complete room.
M means monster present, S means a special/unique present, T means a trap and a Y in the TREASURE column means treasure present.
The numbers in the columns come into play with d100 tables you plug in from elsewhere. So a roll of 007 on that table gives you a room-with-monster but also tells you that it is Monster #85 of your pre-defined table. This table is, of course, themed around whatever your dungeon or level is themed around. And you don't need all 100 monsters as you can say things like;
01-40 Chaos Dwarves
41-42 Flailsnails
43-50 Blind Cave Stirges etc. etc.
And so on. A d100 table can easily be re-factored to a d10 or d20 table. And if you don't want that many monsters, then it's easy to truncate that table so that a roll of 85+ (for example) is No Encounter.
The same is true of traps and specials. Room 005 gives us Special #19 and Room 015 has treasure guarded by Trap #64. Again, hackable to reduce the frequency.
FEAT.1 through FEAT.3 indicate d3 dungeon "features" which is basically the sort of stuff you get on Dungeon Dressing tables. Random rubbish, architectural features, natural cave formations, pools, archways, offal, chaos spikey bits sprouting from the walls etc. etc. This is where you define the "look and feel" of your dungeon.
If you want less dungeon dressing (i.e. d3-1) then simply ignore the last result on the list.
Now, the obvious use of this is for stocking pre-created maps. But, your Boy Coop here is an idle fucker at the best of times and hates stocking maps. So kilodungeon is designed from the ground up for seat-of-the-pants improv. In other words no map drawn out in advance, just generated in play. This is an exercise in determining what you really, really don't need sorted in advance to start to play (which means that it owes a considerable debt to the "lightbulb moments" engendered by the One Page Dungeon concept and the superlative Vornheim).
How big are these rooms? What shape, where are the exits?
They are as big and as small as is required for the action contained within them. If it's a Dragon lair it's big. If it's an empty room it's probably not important, if it's a temple to Prahmfaaze, Mother of Bastards then it's probably big enough to house an altar and large congregation. Exits are, well, there's the one the PCs entered and the rightmost column tells you how many there are. Shape is not important unless it's important in which case draw something down. On a piece of paper somewhere. At whichever point it starts to matter.
There is nothing to stop you deciding once a result has been ascertained (posh way of saying, rolled on d1000) that it applies to a series of rooms or caverns so stops becoming a cave with Morlocks in and becomes a complex of linked caves that the whole Morlock tribe inhabits. This increases the efficiency of kilodungeon considerably and might more to the tastes of players who demand realism in the dungeons they ravage while pretending to be busty Elf tarts in scalemail knickers.
Corridors
When planning out the dungeon (which means assembling the M, T, S and Dungeon Dressing tables) decide upon the percentage of rooms to corridors. Not necessarily footprint ratios or anything like that, just the flat percentage change that leaving a room leads to a corridor. Then roll for each exit used. If it comes up corridor, you can roll above for number of exits from that corridor on the d1000 table or simply roll d4 (which is all the little bit of VBScript I wrote to produce the above does).
Corridors go wherever and twist and turn wherever. Make it snap up to the rest of the already-drawn map or head off in another direction. Whichever seems most useful at the time. You're the DM after all.
Secret Doors.
Again, flat percentage of rooms/corridors that may have secret doors. In the interests of saving effort don't even waste time thinking about them until if and when the PCs start looking, then do a secret roll. (Or pre-game knock up a list of 20+ results and use them in order).
Level Entrances/Exits
You probably saw this coming - flat percentage chance of each room. Roll it etc. etc.
With the last three constants you can tweak the snakiness of the dungeon, density of secret doors and passages, and ease of egress from the level. It strikes me that you could also rip an existing module apart, cannibalizing it's monster list, trap list, specials and dungeon features and play it using kilodungeon.
Obviously, the contents of the PDF were generated by pressing refresh on my VBScript. I could churn out dozens of those charts at a drop of a hat.
Thoughts and criticisms welcome!
The first 21 rooms look like this;
Ugly, but functional.
Here's how it works. This is basically the Moldvay dungeon stocking tables streamlined down to a single d1000 roll, so that you roll once and get the complete room.
M means monster present, S means a special/unique present, T means a trap and a Y in the TREASURE column means treasure present.
The numbers in the columns come into play with d100 tables you plug in from elsewhere. So a roll of 007 on that table gives you a room-with-monster but also tells you that it is Monster #85 of your pre-defined table. This table is, of course, themed around whatever your dungeon or level is themed around. And you don't need all 100 monsters as you can say things like;
01-40 Chaos Dwarves
41-42 Flailsnails
43-50 Blind Cave Stirges etc. etc.
And so on. A d100 table can easily be re-factored to a d10 or d20 table. And if you don't want that many monsters, then it's easy to truncate that table so that a roll of 85+ (for example) is No Encounter.
The same is true of traps and specials. Room 005 gives us Special #19 and Room 015 has treasure guarded by Trap #64. Again, hackable to reduce the frequency.
FEAT.1 through FEAT.3 indicate d3 dungeon "features" which is basically the sort of stuff you get on Dungeon Dressing tables. Random rubbish, architectural features, natural cave formations, pools, archways, offal, chaos spikey bits sprouting from the walls etc. etc. This is where you define the "look and feel" of your dungeon.
If you want less dungeon dressing (i.e. d3-1) then simply ignore the last result on the list.
Now, the obvious use of this is for stocking pre-created maps. But, your Boy Coop here is an idle fucker at the best of times and hates stocking maps. So kilodungeon is designed from the ground up for seat-of-the-pants improv. In other words no map drawn out in advance, just generated in play. This is an exercise in determining what you really, really don't need sorted in advance to start to play (which means that it owes a considerable debt to the "lightbulb moments" engendered by the One Page Dungeon concept and the superlative Vornheim).
How big are these rooms? What shape, where are the exits?
They are as big and as small as is required for the action contained within them. If it's a Dragon lair it's big. If it's an empty room it's probably not important, if it's a temple to Prahmfaaze, Mother of Bastards then it's probably big enough to house an altar and large congregation. Exits are, well, there's the one the PCs entered and the rightmost column tells you how many there are. Shape is not important unless it's important in which case draw something down. On a piece of paper somewhere. At whichever point it starts to matter.
There is nothing to stop you deciding once a result has been ascertained (posh way of saying, rolled on d1000) that it applies to a series of rooms or caverns so stops becoming a cave with Morlocks in and becomes a complex of linked caves that the whole Morlock tribe inhabits. This increases the efficiency of kilodungeon considerably and might more to the tastes of players who demand realism in the dungeons they ravage while pretending to be busty Elf tarts in scalemail knickers.
Corridors
When planning out the dungeon (which means assembling the M, T, S and Dungeon Dressing tables) decide upon the percentage of rooms to corridors. Not necessarily footprint ratios or anything like that, just the flat percentage change that leaving a room leads to a corridor. Then roll for each exit used. If it comes up corridor, you can roll above for number of exits from that corridor on the d1000 table or simply roll d4 (which is all the little bit of VBScript I wrote to produce the above does).
Corridors go wherever and twist and turn wherever. Make it snap up to the rest of the already-drawn map or head off in another direction. Whichever seems most useful at the time. You're the DM after all.
Secret Doors.
Again, flat percentage of rooms/corridors that may have secret doors. In the interests of saving effort don't even waste time thinking about them until if and when the PCs start looking, then do a secret roll. (Or pre-game knock up a list of 20+ results and use them in order).
Level Entrances/Exits
You probably saw this coming - flat percentage chance of each room. Roll it etc. etc.
With the last three constants you can tweak the snakiness of the dungeon, density of secret doors and passages, and ease of egress from the level. It strikes me that you could also rip an existing module apart, cannibalizing it's monster list, trap list, specials and dungeon features and play it using kilodungeon.
Obviously, the contents of the PDF were generated by pressing refresh on my VBScript. I could churn out dozens of those charts at a drop of a hat.
Thoughts and criticisms welcome!