Since last time we talked about a lot of the “DON’Ts” of puzzles and puzzle dungeons, I felt it was only fair to add a few “DOs” to the pile to balance out the brow-beating. So, I’ve collected some creative ideas for making puzzles and puzzle dungeons a fun facet of your tabletop game (or at least, not an obnoxious one).
Mechanics for Riddle-Solving
There is a model of puzzle design that manages to keep the satisfaction of solving a puzzle without the fragility of hoping someone has the proper amount of brain with them that day: I call it “clue stacking” (which isn’t quite the same as clue juggling, but the two work well together)
Clue stacking involves letting the players roll at any given time, unmodified, to get a shot at the answer to the puzzle. This works best on big puzzles that you spread the clues for across several rooms or locations, meaning it can be used on general plot intrigue across a campaign as well.
There is a very, very low base chance of success – something like 1-5%. In D&D, they’d need a natural 20, or possibly even a confirmed critical with a relatively high AC (that’s right – puzzles have armor class now) to succeed at the start. Each clue the party has found ups the percentage by a fixed amount, effectively granting a bonus on the roll until the chance of success is all but guaranteed – ensuring that your players will eventually get clued in if they’re still missing the point after enough effort is invested.
Go back to our puzzle wall notion, with the side rooms to keep things moving forward. When your players first reach the wall, they have a 5% chance of guessing the solution outright. After they clear the first side room and get to the clue at its end, they can return to the puzzle wall and try again, now with a 25% chance of success (16 or higher on a d20). If they clear the other side room, the chance goes up to 45 or 50%, and so on until the chance is at or near 100% and they don’t even have to roll. Divvy the boost each clue grants so that the players always start low and end at 100%, or very close to it. Shorter puzzles will naturally grant a lot more boost per clue, because there will be fewer of them.
Allow only one roll for the party, and only one roll before another clue is found, so that they can’t just keep hammering at it until they get lucky. They have to keep investigating the side pieces to move forward with the larger puzzle. Let any player roll, or let them elect you as the GM to roll, but it’s their option. Even if it’s random, rolling a die feels a lot like agency, which is very important to the whole model.
At any point, a player who guesses how to solve the puzzle or riddle can completely bypass the rolling mechanic – this is strictly a safety net for when no one is in a Sherlock kind of mood. It doesn’t feel like an eject button on the primary puzzle to have to still do the side challenges, but it offers the same guarantee quietly to make sure players don’t get frustrated and hit their heads against the literal wall as well as the in-character one.
For instance, a literal Rubik’s cube can simulate a hacking attempt, the solving of which is similar to the time and skill required to actually crack into the system (although that may be a little weighty an example). I recommend the Hanayama puzzles you can often find in craft or book stores. They have a level system, and levels 1 and 2 are just hard enough to fit into a gaming environment where you don’t want to derail everything with hours of tinkering.
It becomes a living symbol of the lock your rogue is working to untangle. Just remember the rule about having something else for the others to do while one is tinkering with the puzzle so that the rest aren’t simply looming over them, tapping their feet. Let them advise each other and work cooperatively to foster that sense of collaboration, and create a sense of team victory when the two pieces separate successfully.
Short jumbles or ciphers also work for cracking pass codes or puzzling out the chant to sing to summon the guardian of the temple, but they’re very commonly used and can seem rote on their own. A broken cipher is often more rewarding: Take a pass phrase and write it all out without spaces. Count the letters, and divide them into even-numbered groups. Put each group on a card – that’s a single clue for a player to find.
If the phrase is “WINTER IS COMING,” you can break it into “WINTE,” “RISCO” and “MING” to hand out as separate pieces.
Once the players have enough of the cards, they can start guessing at the missing letters in between to put the phrase together on their own, or keep searching for the remaining pieces to make it obvious. Lacking spaces is often enough to make it still feel like a puzzle, but you can also write it all in reverse to really trip them up, if you like. If you do, be sure to include linguistic clues (words with “QU” work well) to illustrate that the words are backwards so that it’s still easily solvable once they have all the pieces.
When you can, make the pass phrase something that references the plot or other fixtures in your game so that it has a sense of relevance and connectedness to the campaign world. It may also reward players paying attention to common names, phrases and motifs (see: “winter is coming”) by making it easier for them to guess ahead once they have some of the clues in hand.