Edit: this is not a dupe because the other question asks for an everchanging maze that is built with pre-historic technology and that is impervious to engineering (i.e.: immune to drilling). Mine allows for backtracking until reset, must be built with current methods, and involves no magic/alien technology to protect it against heavy machinery.

Roguelikes are videogames in which the player must navigate through a procedurally generated network of rooms (and possibly tunnels) to reach the end of the game. Popular examples include Diablo, The Binding of Isaac, and Rogue itself.


To build an underground labyrinth that changes its internal configuration based on a procedural generation algorhitm, which can be reset at will by the maintaining staff (the time it takes to reset is irrelevant). The maze may reuse rooms, but it should be statistically unlikely (< 0.00001% chance) to have the very same network twice.


Technology level should be what we expect to achieve by 2030.

Arbitrary goals for the labyrinth measures are 0.5 km2 per level, for a total of six levels.

Disregard/handwave away how the labyrinth is going to be staffed or populated.

Suppose no expenditure limits.

Finally, the deadliness and level of challenge of the labyrinth are not part of the question. The goal is to determine the engineering feasibility of the changing labyrinth, and only that.


Suppose a villain has played very hard games such as Rogue and concluded that those are nearly impossible to finish, so they think that hiding in such a dungeon will make them harder to defeat.

Alternatively, some TV channel may be one upping the reality show business, or a theme park may be trying this as an attraction (in which cases the traps would not be deadly and the monsters would not harm the guests).

  • 1
    $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – James
    Commented Oct 8, 2018 at 2:08
  • $\begingroup$ As this is eminently feasible, there are lots of answers to the affirmative, so it would be great if you would sharpen your goals a bit: As the answer from Keith Morrison suggests, you do not need all that many rooms: 49 rooms in a gridlike pattern (or any other pattern preserving that graph) will result in half a Billion different ways throught the maze - Are there any other goals that might thin the herd? pathlength? sub-patterns of visited rooms? ...? $\endgroup$
    – bukwyrm
    Commented Oct 8, 2018 at 12:38
  • 1
    $\begingroup$ @bukwyrm all my goald have been met beautifjlly already. Also I don't like to make edits that invalidate answers. $\endgroup$ Commented Oct 8, 2018 at 13:08

19 Answers 19


Well, you can go the operable wall way:

Tactical Ground

police use this Tactical Training Grid to construct distinct scenarios to train policemen. I think you can do the same but with automatized walls in a grid, that moves and deploy according to the desired pattern. To compose the maze i suspect a lot of movements will be required, so this in action could be like a giant 2-d rubik.

The trickiest part will be the connection between floors, but you can place a lot of doortraps closed and concealed and only open one per level.

All the other stuff (chests, furniture, staricases, fountains, etc) can be placed by robotic Forklifts.

  • 8
    $\begingroup$ +1 - and the trapdoors could be covered by a layer of moving floor, with only one room having a moving floor with a square cut onto it. $\endgroup$ Commented Oct 4, 2018 at 13:40
  • 8
    $\begingroup$ Like that movie, "The cube"? $\endgroup$ Commented Oct 5, 2018 at 0:52
  • $\begingroup$ I think a raise/lower wall instead of a sliding one would simplify composing it and automatically changing the layout. $\endgroup$
    – Bobson
    Commented Oct 5, 2018 at 17:39
  • $\begingroup$ As long as we have to deal with multiple levels (ala roguelike maze) if you need to raise/low a wall, a significant space between levels are required (a interlevel to acomodate the walls and manage the positon, then low/raise it to the current position in the "playable" level). Why is this more efficient than the sliding? $\endgroup$
    – Onoper
    Commented Oct 8, 2018 at 10:53

Thick floors with hydraulic pillars

Given a grid made up of 30 to 50cm square pillars, each with an independent hydraulic piston and controller allowing it to either act as a full height pillar, even to the point of acting as stairs to the room above, or sit flush with the "floor", everything else is software.

You'd need to define regions on each level that are always open space to allow for removable panels for the stairs to the lower level, but these regions aren't necessarily always available to the adventurer, they can just never be walls or stairs to the higher level.

While crude, and slightly limiting in where you can put the level changes, it will fulfill your basic requirements.

It also leaves open the option of crushing your intruder against the ceiling if he gets too close/boring

  • 13
    $\begingroup$ And it gives the super villain an easy out, if he remembers to bring his badge, the walls will just drop in front of him, allowing a straight line out. $\endgroup$
    – Martijn
    Commented Oct 4, 2018 at 14:27
  • 16
    $\begingroup$ “This was a triumph. I’m making a note here: Huge success”-Glados, Portal. $\endgroup$
    – Joe Bloggs
    Commented Oct 5, 2018 at 8:07
  • 1
    $\begingroup$ What about doors? $\endgroup$
    – M.Herzkamp
    Commented Oct 5, 2018 at 11:57
  • $\begingroup$ @M.Herzkamp there are no doors in Rogue $\endgroup$
    – Separatrix
    Commented Oct 5, 2018 at 12:18

If reset time (and cost!) are not relevant, just tear it down and rebuild from scratch on every iteration.

You can procedurally generate the blueprints by computer just like the actual games, perhaps using one of the games as-is.

  • 15
    $\begingroup$ Not the answer I want, but I'll give you a +1 for being a smartass. Ideally I want the mase to be reset automatically - but I am not editing the question now because I'd hate to invalidate an answer. $\endgroup$ Commented Oct 4, 2018 at 16:58
  • 6
    $\begingroup$ On a similar note, a huge 3D printer would do the job. Would take some time to generate (build), and there's no expenditure limits, so all you'd need to do would be to remove the old one at the end. $\endgroup$
    – gabe3886
    Commented Oct 5, 2018 at 12:25

You should check out the movie Cube: https://www.imdb.com/title/tt0123755/ The maze keep changing, with almost infinite possible combinations. In a sense it's a rogue like maze.

It consist of many rooms that are shaped like cubes with different sets of traps (some with no traps at all). Each room has six hatches that leads to other cube like rooms. The structure is basically i gigantic cube, made off smaller cubes.

Spoiller alert - As the time passes by, the rooms move, like a sliding puzzle, modifying completely the structure of the maze.

how to build it The fictional Cube device in the film was conceived by David W. Pravica, a mathematician. It consists of an outer cubical shell (the sarcophagus) and the inner cube. One side of the outer shell is 434 feet long. The inner cube consists of 26x26x26 =17,576 cubical rooms (minus an unknown amount of rooms to allow for movement, as shown in the film), each having a sidelength of 15.5 feet. There is a space of 15.5 feet between the cube and the shell. Each room is labelled with three identification numbers, for example, 517 478 565. These numbers encode the starting coordinates of the room and the x, y, and z coordinates are the sums of the digits of the first, second, and third number respectively. The numbers also determine the movement of the room and the subsequent positions are obtained by cyclically subtracting the digits from one another. The resulting numbers are then successively added to the starting numbers.

  • 7
    $\begingroup$ I'm afraid this doesn't answer the OP's question, wich is how to build such a maze. This is a comment, identifying a pop-culture reference to such a maze. $\endgroup$
    – JBH
    Commented Oct 4, 2018 at 14:02
  • $\begingroup$ Not completely. Instead of the Hypercube construct like in the refered movie(s), you could make it static, configuring the cubes themself. Like a fixed 3d maze $\endgroup$
    – Martijn
    Commented Oct 4, 2018 at 14:40
  • $\begingroup$ @L.Dutch just added the detais to built it $\endgroup$
    – Lupino
    Commented Oct 4, 2018 at 15:38
  • 1
    $\begingroup$ The answer is now not only good, it is very good. $\endgroup$ Commented Oct 4, 2018 at 16:40
  • 1
    $\begingroup$ @MrScapegrace I also remember it this way. It works simply like a sliding puzzle. $\endgroup$ Commented Oct 5, 2018 at 12:48


We’ve already made huge gains in VR in the last few years. By 2030, the medium will be considerably more mature. For both the TV channel and the theme park, that should be more than good enough.

As for the super villain, security through obscurity is terrible. Shame on them.

  • $\begingroup$ It would be good to expand this, with examples (or mentions) of VR being used currently with positional awareness. (such as this, centimeter-accurate VR tracking for samsung gear: vrscout.com/news/centimeter-accurate-virtual-reality). And this method of making a small area seem much larger: (newatlas.com/eye-tracking-vr-infinite-walking/54824) - likely inspired by Star Trek's holodeck. I think a lot of people don't realise quite how advanced the bleeding-edge has become over the last couple of years. $\endgroup$
    – Bilkokuya
    Commented Oct 5, 2018 at 9:17
  • 1
    $\begingroup$ Another great example of actual-in-practice VR doing this sort of thing is The Void, where they do a whole bunch of stuff to mess with peoples' expectations even in multiplayer VR contexts (for example, "cloning" party members in VR and shuffling them off in different directions in reality, allowing each half of the party to see the other half get eaten by a monster). $\endgroup$
    – fluffy
    Commented Oct 5, 2018 at 18:45
  • $\begingroup$ I remember a story waaaaay back (like pre-pre-2008 and my storyteller was like 70 so probably even farther than you'd think) were he was recalling some research done where they were testing VR for the military and knocked people out and threw them in VR to see if they could tell the difference between sleeping and VR and whether their conscious choice on things would change. I imagine doing this would work splendidly (testing in Room 2). They only get out of their VR tank to room 3 if they win. If they lose knock them out again and put them in Room 1. $\endgroup$
    – Black
    Commented Oct 7, 2018 at 11:12

The TV show version is easy. There, contestants are following rules, so any of the modular room design patterns in the other answer apply. Anyone who tries to break through a wall they aren't supposed to break through gets penalized.

The evil villain case is a different story. If you're filling the room with deadly traps and monsters, the heroes assaulting your lair are not going to play by your rules. They'll break through walls, go over boundaries, and do whatever it takes to get to you. As such, you are going to want much stronger materials. Typically you would do a lot of work in concrete, which is hard to demolish and rebuild. You would likely need to be refurbishing one or two rooms at a time, and adjust your maze such that that is useful. Indeed, one small change was enough in The Matrix.

However, you mention 2030 tech, and that's terribly convenient because I just saw a video of a robot that can hang drywall itself. A lot can happen in 12 years, and it may be reasonable for 2030 robots to hang steel plate, and bolt it or even weld it in place. A steel plate can be cut down, given effort, but it is decidedly not easy, so it could be a reasonable basis for your mutable labyrinth lair.

Of course, we all know labyrinths wont be the same without David Bowie. RIP

The Goblin King

  • 3
    $\begingroup$ Houses are being 3D printed in China as we type (takes 12 hours to get the framework done). Totally possible to 3D print a maze as well. $\endgroup$ Commented Oct 4, 2018 at 17:21

Gigantic 3D printer

  1. Build an empty underground structure which will be no more than a gigantic 3D printer.
  2. Fill the first 10 cm of your cube with sand.
  3. Let your gigantic laser (or any very hot beam) melt the sand to created 10 cm height “glass” walls following your maze pattern. This can be archived thanks to rails fixed to the ceiling, allowing the laser to move freely.
  4. Add another 10 cm of sand and repeat the process…
  5. Once the whole cube is printed, removed the un-melted sand and there you go, you have a 3D maze :

enter image description here

The process in video.

But don't do it with sand !

Of course, removing the un-melted sand will not be easy if you have a true maze with lots of levels and dead ends and removing the whole structure when the maze will have to be reset will also be difficult, but here is the solution. Don’t do it with sand but with microbeads of a special polymer that melt when exposed to very hight temperature and then harden when cooled. Then, to remove the un-melted microbeads, just flood the cube with a selective solvent that will dissolve un-melted microbeads, but not the hardened walls which will have changed of molecular structure during the hardening step. Then... Flush it by opening some bung ! And when you want to reset the whole thing ? Just turn the heat on and let your structure melt...

Other option

If you don't want to have problem with flame-thrower-equiped-adventurer (Thank M.Herzkamp), you can harden your microbeam polymer with UV light and reset your whole maze with solvent specificaly designed for the hardened polymer. Of course, keep the polymer and solvent formulae secret in a very strong vault in your base !

  • 1
    $\begingroup$ So, if the adventurer has a flame thrower, every wall is a door to them $\endgroup$
    – M.Herzkamp
    Commented Oct 5, 2018 at 11:59
  • 3
    $\begingroup$ @M.Herzkamp Well, a "NO FLAME THROWER ALLOWED" sign should do the work... More seriously, flame thrower would melt the walls, the floor and the ceiling, flooding the adventurers in some boilling hot polymer ! $\endgroup$ Commented Oct 5, 2018 at 12:14

Other answers have focused on the mechanisms of how to have a changing maze. I'm going to answer your other question: how many rooms you need so that there's less than a one in ten million chance a randomly generated maze repeats?

Surprisingly few.

For the number of rooms you need, where people going through the maze do not enter the same room twice, you need a self-avoiding rook walk.

Here's an example for a 2x2 and 3x3 array of rooms. Think of the rooms as occurring at the intersections of the lines. The entrance is in one corner of the array, exit in the opposite corner.

So in the top row, in the 2x2, there's only 2 ways to get form start in one corner to finish in the other. In the 3x3 array of rooms, there's 12.

enter image description here

  • In a 4x4 array: 184 possible paths.
  • In a 5x5 array: 8,512 possible paths.
  • In a 6x6 array: 1,262,816 possible paths.
  • In a 7x7 array: 575,780,564 possible paths.
  • In an 8x8 array, a chessboard: 789,360,053,252 (yes, over 789 billion possible routes).

You certainly don't need multiple floors, and you don't need to reuse rooms on a given single run through.

(Numbers from here)

Now, the thing is that the rooms certainly don't have to be identical, only that a given room is only connected to between 2 and 4 other rooms at most, depending on where it falls on the array network (and how many doors you allow to be open). So some might be small rooms, some long tunnels, whatever.

There is a downside to that: suppose the starting position is in the southwest corner. Someone entering your maze will know, if they've been through it before, there's only two possible rooms that come next, one to the north, or one to the east. If the path is to the north, they'll know that in the next room they can only go either north again or east. If enough groups make it through and compare maps they made, they could probably figure out the general concept of your maze.

However, that wouldn't necessarily do them any good. Look again at the possible paths through the 3x3 network. They might know there's only 9 rooms, but they wouldn't know which of the 12 possible routes they'll end up on. They might need to only go through 5 rooms. Or they might have to go through all 9 to reach the end.

For an 8x8 network of rooms, it would be a horror show. They start off, they don't know if they're going to be fighting traps and whatever in only 15 rooms, the minimum, or 64. With mapping, they might might have a good idea how many rooms they have left near the end of the run, but still...

And if you do only two floors, so two 8x8s on top of each other, with ladders connecting rooms so they might have to go up or down a level, and not knowing how many times...

So: an 8 x 8 network array of rooms, on each of two floors. Each room has between 2 and 4 doors, depending where it is in the network, and a trapdoor either on the ceiling or floor, depending what floor you are on, but only two doors/trapdoors can be opened: the one you use to enter the room, and the one you use to leave. Not only won't they know what you might have set up in a given room to deal with, they won't know if they'll only face 15 rooms (shortest route, single floor) or have to go through all 128 of them.

And to add even more fun, why assume they start and end on the corners? Consider the entrance to the maze itself consists of a long, winding tunnel which makes it difficult to determine how far they've traveled, when it suddenly comes an an abrupt end with a door. The only thing is, the end of the tunnel is moveable like an airport bridge, so when they open the door, they don't know what room on the west side of the maze they're entering. It might might be any one of the rooms on the west side.

And imagine that the exit consists of a long tunnel that wraps around the other three sides on both floors, that any given room on the north, east, or south sides might have the exit door. Now that screws up their mapping again because if they end up in one of those room, they don't know if the next door is the way out, or if they're going to be turned around and sent deeper into the network.

Another thing about thinking of it this way is that it doesn't have to be procedurally generated: for your 8x8 array of rooms, the computer can be programmed with all 789 billion possible routes; despite the number of routes, each can be coded easily, for example NNEESWSEENN... (describing what door in each room opens to the next). When a run is ready to go, the computer randomly selects one of the paths and it's ready to go and flagged so that it's not included in the next random selection.

This also allows you to set up "difficulty". Coded like I suggested, it's trivial to tell the computer "select one of the routes that has between 15 and 23 rooms" or "select one of the routes that's between 50 and 64 rooms", giving you difficulty levels through shorter or longer possible mazes.

By the time you get through 789 billion combinations, it's pretty safe to assume that someone who's been through once is never going to see the same route again, no matter how much of a human lifetime they keep running it.

Fun times, fun times.


I don't see any particular problem here from what you've said so far. You want a six story building (or 5 stories and a basement) with no load-bearing interior walls. You can have load-bearing posts or pillars. You want movable walls for the interior. You might settle for two stairwells and block off every other level on each of them, so you have to get from one to the other. That's easy and cheap, though you could use, say, ladders anywhere you're willing to open up the floors.

You didn't specify that you want the interior walls to be resistant to maces and battering rams.

Getting robots to do the work is trivial in theory.

.5 square kilometer. say a building that's 1 kilometer by 500 meters. About 3300 feet by 1650. 500,000 square meters. The Boeing Everett factory is 400,000 square meters, maybe 10 floors. Close. There are various other great big buildings. So this isn't completely unprecedented.

There are various commercial solutions for the actual walls, here's one of them.

movable walls


You could do it nowadays if you really wanted to. Consider a square room with 4 walls, a roof and floor. Above each room, is a secondary room where a winch and motor are attached to each of the walls. Given a command, each of the walls could be lifted, creating an opening and a new passage way. This way you can create your maze structure fairly easily, and it can be maintained by people walking over the roofs of the modular rooms.

Modular Rooms

The next key point is set pieces. These are rooms that are designed to be different from your standard passages and contain fixed puzzles. You can change the way your victims travel to each set piece, and you can also move the set pieces yourself. How? Overhanging crane system.

If you've ever seen videos of Airplane construction, you will notice that large parts of the planes are moved using a warehouse wide crane system. You lower it the crane down. Attach it to the modular square rooms you want to remove and off they go. Then simply lower a set piece into the game and open up the corresponding passage ways on the modular rooms outside.

And there you go... a changing and customizable maze

Edit: To get multiple levels, you would use a special set piece in fixed locations that would cover altleast 6-7 stories. It could be an elevator, a pit or a long stair case. You have the original Level they are on (1), the overhead for the changing walls (1), overhead for dangling set pieces/rooms (2-3), Room for the crane (1-2), the next level (1). You could combine these with rooms that spin to disorientate your victims.


This shouldn't be an issue even today (aside from being horribly expensive.)

Take inspiration from Minecraft, and use a grid structure of "blocks". I'd use 3 "floors", and just fake the floor traversal. Because you're not getting to kill me (the villain) just because you managed to traverse 6 floors...

  • Bottom floor is essentially just hydraulics and cable management to lift the various blocks up/down to create the room and passages. It can even include pitfalls by going further down than the rest of the floors. The randomness is just an algorithm, they do that already in the various rogue-like games.

  • Middle floor is the maze itself where the people get lost. The floor traversal could be a winding stair that just slowly move up/down as the people traverse it, giving the impression of changing floors while the rest of the maze re-arranges itself.

  • Top floor is all your traps, monsters and goodies being shuffled around by warehouse robots, then dropped down in the respective rooms by a hatch in the ceiling. Simple enough. Upon re-arranging the maze, the stuff is pushed back up again by the rising floors.


Trivially a grid or cube of e.g. shipping containers, with doors between every one -- so each cube has 6 doors (4 walls, top, and bottom). If you offset the lines of rectangular containers (so the front of one container is adjacent to the back of the one next to it), then 2 doors on each of the long walls, so eight doors per container.

What makes it configurable is just that the doors are locked electronically (and networked) -- so maintenance staff can press a button to reconfigure which doors are locked and which aren't.


Roguelike games consist of lots of rooms. It’s the random nature of the rooms’ connections and contents which gives it the random maze feature.

So you don't have to build a massive building with no load-bearing walls. You can pre-build lots of identical rooms in a grid, in a number of sizes. So working with a 10m base grid, you'd have rooms such as 10x10m, 10x20m, 20x20m, 40x40m... (larger rooms will be a little larger than 2x because of the wall thickness). Doors would be in the middle of each standard grid wall (so a 20m wall would have 2 doors at 5 and 15m).

Yes, the position of the smaller/larger rooms would be fixed, but given the ability to rearrange the connections inside, your maze qualities are preserved. With foreknowledge of a previous run, users might know that the next room will be larger or smaller, but that doesn't really give them much advantage.

You then have automatable doors – you could do a fancy thing retracting into the ceiling, but sliding would be sufficient. You can then change routes through the grid of indistinguishable rooms. Some rooms can be left empty, to serve as corridors.

Contents for the rooms can be delivered either via a service floor above, or more simply by sliding all the doors open whilst you rearrange the maze, giving you a wonderfully accessible area for an army of robot forklifts or similar to drive through.

You can build your walls and doors out of something nice and solid like re-enforced concrete to prevent damage, particularly if you're having more lethal activities inside it.

You can build your levels above each other. Stairs can easily be deployed where you want them as a movable set of steps to a trapdoor in the ceiling. You can have a number of pre-set locations with doors to the floor above; on a 0.5km2 level you could have a hundred or more, and with only 2 open (one up, one down), knowing the potential locations of the stairs provides users with very little help.

Similarly with any traps that require installation into ceilings/walls/floors. Provided the rate of them being active is very low, knowing that there's the potential for one in a location is not useful – in rogue, any step is potentially a trap, but knowing this provides no help.

A service floor between each level (and at the top&bottom) would be sensible – this would allow you to route power cables, etc. You'll also need substantial air-con; you could probably use small impassable ventilation grilles between rooms, so that air can pass even when the doors are shut, and some clever design could use passive techniques to encourage airflow.

This should be entirely buildable with current technology, though the army of robotic forklifts might need some work to make them reliable; that's likely to be the hardest part. The cost would be fairly high, but it's hard to imagine it'd be beyond the pocket of someone like Musk – the issue is that to build an investment of that size, you'd need a plan to make returns on it, and it's hard to see how it could make a meaningful ROI.

Trying to get an estimate of costs is hard. The largest building by floor area is the Boeing Everett Factory at 0.4km2. I don't know the cost to build it, but they built it to deliver a 525m USD contract; assuming they had a healthy profit margin and that the bulk of the costs were to build the planes, we can surmise it cost probably no more than 50m USD.

Obviously a multi-floor building is more expensive, but normal height ceilings and smaller rooms should make it a lot cheaper. So you're looking in the region perhaps of $500m?

It's worth noticing that the Boeing building has lots of its own support infrastructure, including canteens and a fire department. You'll need to add in that kind of stuff too. And a lot of staff; they'll need to be fed etc.

I'd recommend a 'staff room' for monsters at regular intervals, particularly if you need to administer first-aid also. Or you could give monsters passes so that they can open doors and take a direct route in/out as long as the adventurers aren't too near.

  • $\begingroup$ "I believe activity only occurs in rooms, not corridors." - in games I have played, this is false. Thus, as a general statement, this is also false. $\endgroup$
    – Mołot
    Commented Oct 5, 2018 at 14:51
  • $\begingroup$ Ah, it’s been many years. In any case though, I don’t think corridors are integral to the concept. $\endgroup$
    – Dan W
    Commented Oct 5, 2018 at 14:52
  • $\begingroup$ Updated answer in response. $\endgroup$
    – Dan W
    Commented Oct 5, 2018 at 14:55
  • 1
    $\begingroup$ There are ones that are mostly corridors and ones that can do without corridors at all. You can have it both ways, and any way inbetween. $\endgroup$
    – Mołot
    Commented Oct 5, 2018 at 14:55
  • $\begingroup$ "(...) fire department. You'll need to add in that kind of stuff too." Only for my personal room, the heroes may burn for all I - I mean, nice answer, +1 :D $\endgroup$ Commented Oct 5, 2018 at 15:16

You can get the same general result easier if you think about it in a slightly different way. The rooms in your maze are normal, static rooms. The connections between them, however, aren't.

The simplest example would be to imagine a tower with one room per floor. Floors are accessed via an elevator (fully automated, no accessible controls). When you leave a room, you get in the elevator with no idea where it will take you. You can visit rooms in any order, perhaps visiting some rooms more than once or not at all. Rooms can be reset and re-stocked while they're idle. The end result is a completely unpredictable experience, even if the rooms themselves are mostly static.

You don't need a fancy automated elevator system for this. Connecting rooms via stairwells/hallways/lobbies with heavy access doors and remotely-controllable locks would also work. You'd likely want some kind of signage or signal light to direct your victim hero to the next open door, to avoid the hassle of having to check a hundred doors just to find the one that's unlocked.

I did something similar back in the late 90's when working a haunted house that was built in an old office complex. We would adjust the flow by opening or blocking doors while visitors were in a room, allowing us to dynamically route traffic flow based on which rooms were ready and which were still resetting. We did it manually (so no fancy technology needed), but you can make it as high-tech as you want. I was very surprised at how well this technique could turn what was originally a bland set of rooms along two parallel hallways into something that felt like a sprawling, disorienting maze. Change up the décor a bit while the victim is in a room and it's even hard to detect that you're walking through the same hallway again.

For the theme park angle, this also makes it easy to dynamically adjust the difficulty by swapping out difficult rooms for easier ones and adjusting the number of rooms required to finish. For enhanced replayability, you can have hidden puzzles that change the room sequence when solved. An attraction that follows an overarching plot could thus have multiple endings.


Moving walls that are also mirrors.

Something like this
mirror maze

But on system like this moveable walls

Add the fact that you can gave 3 floors with such system and you're set for all Bruce's Lee's that may want to hunt you.

Just remember to invest in those cured, hard to broke mirrors. 7 years of bad luck for each.


Thing is, you have plenty of time to prepare, because procedural generation is not random (at least in this case). You can map an arbitrarily large part of the labyrinth before it's actually built or used.

Assuming we have a set of 10 type of rooms, a dozen or so of every variety, all ten are in some nice shape to be tileable (note: we don't need the mathematically significant concept of covering the full plane), and you can build whatever your want with simple cranes right now.

With a moving platform you can even keep the "player" in the center of your arena all the time, so you don't run out of place.

I don't see any engineering problems with it. You want the generation to be more fine/detailed, less blocky?

  • $\begingroup$ "You want the generation to be more fine/detailed, less blocky?" That should be a comment on the question. Anyway, level of detail is irrelevant. $\endgroup$ Commented Oct 4, 2018 at 14:02

Pop up walls.

Consider a piece of graph paper. Each block has 4 sides. Each of those sides is a pop up wall. The hoisting (pop up) device maintains the integrity of the wall, so you don't need rails in the maze space (you could use rails beneath the floor).

The walls would hoist using wheels/rollers, and then pins would secure them in place.

There will necessarily be gaps between the walls. If this is a problem, the walls could have pop out edges that expand when the wall is raised to cover the gaps.

You can make the walls out of any stiff material you could hoist (metal, wood, cloth flats).

You might have a select set of panels that are actually doors.


People have already given a lot of good answers on the how to do a procedural maze in real life.

My answer is more about "do you really need procedural generation" ?

I think one of the most important part in a roguelike is the permadeath. That's why I don't think that Diablo can be defined as a roguelike and why it isn't really relevant to talk about roguelike in real life. We are already living in a permadeath world. If your maze kill people when they fail, you don't need to change it everytime, dead people don't try again nor talk.

It's only useful if they don't die and just get flushed to the start or if it is a TV show and people can see other tries before going in. But in case of TV show, you could decide that the broadcasting is done when every participant already tried. You film the whole season in one go and you have a year to change the maze manually.

  • $\begingroup$ This is interesting comment, but how is it supposed to answer the question? $\endgroup$
    – Mołot
    Commented Oct 5, 2018 at 9:49
  • 2
    $\begingroup$ He's building a world where he felt that a roguelike maze was useful/relevant. My answer point to the fact that it may be really hard to justify a roguelike concept in real life. And if I read a book were the author invent a very convoluted solution to a problem that doesn't exist, it can weaken the story. I feel like my answer can either encourage him to either : - Not needing a real life roguelike anymore. - Be sure that its implementation is well justified in his world, else it feels forced. $\endgroup$
    – Jemox
    Commented Oct 5, 2018 at 10:12
  • 1
    $\begingroup$ While we do allow what are known as 'frame challenges', it's a grey area and such answers can very easily slip into non-answer territory; for me this one is too far from actually answering the question. $\endgroup$
    – walrus
    Commented Oct 5, 2018 at 13:24
  • $\begingroup$ To be more direct, if someone came and asked "How could we make bird fly using modern technology", my answer would be the one that point that bird can already fly without any help. While it's not answering the question, it felt like a valid point to make. Moreover in the case of roguelikeness, where the simple answer is less obvious than the flying capabilities of birds. $\endgroup$
    – Jemox
    Commented Oct 10, 2018 at 12:23

Just open/close doors

On the surface, just opening and closing some doors sounds like a lazy solution, but it actually opens up a wide range of possibilities you don't get with a modular design like those previously suggested.

Modular designs are boring grids of spaces made up of boring walls and boring floors, but what if you want each room in your dungeon to be it's own work of art? One room is a jungle biome with a tiger in it, one is a scenic waterfall overlooking a peaceful meadow, one room is a narrow bridge over a smoldering lava pit, while another is a frozen wasteland, etc. Many of these rooms would need complex machinery to maintain that would be too much for a temporarily or movable room. So instead, you just make a giant dungeon that is several times as large as any one adventure will pass through, and you open up and close off passageways using simple locking doors (preferably that blend into the environment when closed). In doing so, you can direct adventurers to visit each room in whatever order and from whatever direction you wish.

In some cases, you may walk straight from one room to the next, and in others, the DM might make you journey down a 100m passage past several sealed off rooms to get you your next challenge. Either way, it is only the path, and not the points of interest you need to change.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .