Wanting advice on how to puzzle them without making puzzles

Hi folks,

Looking for advice on ways to cause puzzlement/consternation/perilous obstacles in dungeons that challenge without a lot of prep work. Like, I don’t want my players having to meta game to figure out a secret code or sort of thing…

1 Like

Awkward constraints from random tables ? Troika is basically that, also Details (found on itch). Roll for colours, as in Sign in Stranger. Pull a sewing machine and an umbrella and lay them on an operating table.
Then ask your players to paint the scene.

I like how The Between handles this. Give them a bunch of clues and let them spin a good theory about how they fit together. There’s no right answer to the puzzle, just a cool answers. I know that approach will some heads explode, but I’m warming up to it.


That’s a really nice idea! Never heard of it before.

Noirlandia just blew my mind in this aspect a month ago, the system is made to impro a whole crime scenario complete with clues and have the players come up with the whole case and the solutions. There’s even a time constraint, the longer the players take to figure it out, the more chances for the criminal to escape.

I put a similar mechanic to create puzzles in my game, but still haven’t tested it.


What are the basic techniques used in this approach? How would you characterize them?

In Noirlandia, players build a very basic setting and society: important places, groups and people that also help to define an atmosphere and theme. This all pays later when you need the players to come up with stuff. It’s really hard to come up with things from nothing, but once everyone is in the mood and the same page, things run a lot easier.

Next, the crime and subsequent clues come from simple and quite general random tables, that everyone helps to give shape or discard. Clues pile up until players come up with any idea that relates 3 of them to answer one of the usual questions on a crime: who, why and how. There’s a caveat that as a bare minimum the first clue found by the PCs must lead them to another location, and either they get that one for free when they visit a location or roll against a really low difficulty to get it (I think it vas the former, can’t remember well)

This way of finding “clues” and connecting them also solves the problem of being able to leave out clues that don’t fit anywhere, ending up as red herrings without ruining the “investigation”. The whole thing works the same way conspirational theories are made: find a way to connect anything and you get “proof”.

The game itself reccomends to use strings and pushpins to keep account for the whole process, which helps everyone getting into the noir detective mood.

Actual puzzles like “solve this riddle and do a thing to get to the next room” may require a different approach. As players tend to overthink stuff and there’s always a great chance of communication issues when explaining it, it’s best for this type of puzzles to be somewhat simple, have ways to circumvent or have a way to be solved by trial an error where the latter isn’t punished too much or too fast (you could use a flood trap or other slow trap to add tension)

If you want to use the “let players come up with it” approach, the data provided needs to be vague enough to allow different interpretations, and you need to offer a high amount of data so players need come up with leading questions to properly interact with it. Like, in Indiana Jones and the Last Crusade final puzzle, Indy must find out which of the dozens of grails in the room is the holy one, reasons “it’s a carpenter’s cup”, looks for a wooden one and finds it.

It kinda limits how many types of puzzle the system can handle, but if there was an easy way to generate fair puzzles, RPGs could be totally different beasts.


I love puzzles, but the rest of my table doesn’t care for them. There’s also the old chestnut of “but my character would know even if I don’t” kind of thing.

For major puzzles I allow for hints after a certain amount of time/frustration. I used to write them down and reveal them in a particular order-- starting from the most general of hints to pointing out a key word. Usually three are available. They cost something, like a Drama Point, or in-fiction something like time spent, a contact burned, a lost chance, if relevant. When this was for Ashen Stars (Gumshoe), I could ask for increased point-spends.

So first I think there’s a distinction between “meta-gaming” and player problem solving with non game-world knowledge. For example, I’ve always thought of meta-gaming is using knowledge of the rules to overcome in game difficulties – e.g. a player knows that their PC can survive a 40’ drop due so they simply walk of a cliff rather then trying to figure a way down vs. a player who makes the argument about how the party could braid and weave ropes from vines individually too weak and use them to climb down the same cliff.

If we’re on the same page here, I’d say that dungeon adventures are generally about a fair amount of design and prep work on the part of the referee. That’s one of the main reasons that adventures containing complete dungeons are a common product in location based exploration games. The reason for this is because designing and preparing interesting, open-ended obstacles and puzzles that aren’t simply a space for an appeal to a dice based skill test is tricky.

I’d say you best bet if you don’t want to design your own adventures, or more don’t want to design your own puzzles/obstacles is to look at the sorts of puzzles and obstacles people have been designing for dungeon adventures over the past 50 years: Early D&D adventures are full of spatial/navigation puzzles, faction intrigue puzzles, environmental puzzles, complex traps and similar puzzle rooms that are ripe for borrowing.


I try to avoid situations where there is a specific solution. I find that presenting challenges with no direct answer and then being receptive to the players’ creativity is more engaging and rewards creativity. Things like crossing hazardous terrain (a gorge, river rapids, acid pit) or performing with constraints (move this obejct without touching it, climb a 50 ft tower without magic, bring a snowman through a volcano)

We’ve all had that puzzle where the solution is to push three buttons but we’ve spent an hour trying to disassemble the device because someone hyperfocused on the word “red”.


I think the most important consideration is:

What purpose does the existence of the puzzle play within your game?

Is it to challenge the players? The characters? A distraction from the real danger? A way to “gate” away secret content (like secret doors in video games)? A necessary challenge the players need to overcome in order to progress in the game?

That last one is the usual stumbling point here; if you have a puzzle that MUST be solved in order for the game to proceed, then you’re in a tight spot. That kind of adventure design is really, really difficult. And if you choose to go this route anyway, you need to have a backup plan of some sort: what happens in your game if the players can’t solve the puzzle? Or aren’t interested in it? Or if the characters fail (e.g. it’s based on a roll of some kind)? Or if it takes too long? You need to have some principled answers to these questions, or your game is likely to crash in one way or another.

Second, it’s really nice to have a sense of what role the puzzle plays in the game world. Why is it here? Who made it? Why?

My preferred approach is to understand the mechanism of the thing in a solid and clear way. How does it work, and why does it exist? What happens if someone interacts with it like so?

Then you can release any expectations about what will happen or what the players might do, and just play. Describe the puzzle, have the players or characters try things (depending on the style of game), and describe the results.

If you don’t know how and why the puzzle works, then you’ll often be in the awkward position of describing what happens but also trying to decide whether “it’s time for the puzzle to be solved” or other concerns which don’t mesh well with the process of trying to figure something out.

The most obvious example is if the puzzle’s purpose is to withhold a secret until a specified time. (e.g. You’ve placed a cunning lock in room 11, because you don’t want the door to room 12 to open until a specific event has taken place.) This kind of thing tends to be terrible, since you can’t really “play” the puzzle at all; instead, you’re lying to the players if they try to solve the lock/puzzle, making up excuses why everything they try fails, instead of playing a game together. There are no clear stakes, and it’s a sure recipe for frustration.

If your puzzle serves some teleological purpose like that, much better to be upfront about it.

Another principle I go by:

It’s ideal if you understand how the works and why it exists, instead of planning a specific solution. This opens you up to giving the players the benefit of the doubt when they try clever or unexpected things. Since your internal imagination of the space, puzzle, or challenge doesn’t match theirs perfectly (we’ll all be imagining things a little differently, all the time), it’s usually better for the integrity of your game that you go with proposed solutions and consider them possible if it makes sense. A player will be suggesting a course of action because it’s plausible to them, based on how they’re imagining the situation. Sometimes it’s necessary to undercut that to maintain the integrity of the game, fiction, or puzzle, of course (e.g. a player tries to push a door open because they forgot that it’s locked), but whenever it’s not, you actually strengthen the players’ connection to the game when you accept what they’re trying to do as plausible.

1 Like