Anatomy of a Puzzle: A Dicey Proposition
Back in the summer of 2019, I attended a LARP / megagame called Dragon Thrones as a player. Part of the fun of live action role-playing for me is that players take on the role and responsibility to create fun for fellow attendees. So, I thought I’d try and create an in-universe puzzle, designed for 60+ players of a faction-based game as my contribution. That framing opened up design elements and choices that wouldn’t be available in most other contexts, so I thought it might be worth walking through the design process here.
Following Directions and Duck Konundrums: Choosing a Puzzle Type
One of my favorite puzzle types are “following directions” puzzles. The puzzle type is deceptively simple: given a set of instructions, do exactly what you’re told and something magical will happen to give you the final solution.
When this type of activity is done within the magic world, it’s mathemagic. But in the world of puzzles, the results are no less magical. Within the puzzle world, Dan Katz has created a series of duck-themed “following direction” puzzles over the past 20 years that have become known as Duck Konundrum puzzles.
While the instructions may start out sedate, a hallmark of Duck Konundrum puzzles are that events get increasingly out of hand as instructions progress: for the board game themed Cones of Duckshire puzzle pictured above, players needed to track everything from which players were secretly cheating, to which seats at the table were set on fire.
Doing a fully-fledged Duck Konundrum puzzle at a LARP is too much to ask of players, who are typically juggling five different conflicting priorities at once. However, a simplified “following directions” puzzle could be fun and relatively easy…especially if there was a twist at the end.
A 2018 Mystery Hunt “following directions” style puzzle featured custom-labeled dice for a tower-building game, making me wonder how that could be adapted for LARP play.
Developing the Basic Puzzle Structure
Having settled on a puzzle type, I moved on to tackle a structure for the puzzle. The plan was to create a set of four dice for each of the game’s ten factions to try and solve. For the first few steps, teams would be told how to rotate the dice to spell out a message. After that, players would need to realize that I handed out four different colored sets of dice, and that they’d need access to all four sets to solve the puzzle.
That basic structure imposed constraints on what the puzzle could look like: one side of the dice was reserved for numbers that would give players the starting configuration of their dice. And as long as they were solving individually, I was restricted to 4-letter words. So, I settled on using three more sides of the dice to spell out instructions on how to get to the next phase.
That left two additional sides: but since I now had four sets of dice at my disposal, that left me with 16 characters for the fifth step, and 16 characters for the final solution. With that much flexibility, I could hide the instructions on how to proceed to the final step on the dice themselves.
At this point, I reached out to the GMs running Dragon Thrones and told them about my puzzling plans, and offered to let them use the 16 character solution for broader narrative purposes, if they so chose. But with the puzzling structure managed, the final solution could be deferred until later.
Building Out a Narrative Conceit for the Puzzle
With traditional puzzle hunts, you don’t really need to go too crazy constructing a rationale for a puzzle’s existence. “Here’s a puzzle — have fun solving it!” is often justification enough. But for a LARP, I needed to both justify how the puzzle could exist in the game’s world and also why my character would be seeking help from rival factions to solve it.
However, the structure of the puzzle lent an answer to that question: it’s a puzzle where players don’t have enough information to solve it on their own, and can only solve it by working together. The perfect framing for a scientific experiment! Luckily, the character I was playing for the LARP was a bit of a mad scientist engineer, so creating a scenario where I was in the middle of very important research but needed more samples to figure things out was the perfect framing.
Customizing the Puzzle Experience for the Medium
Solving the first few steps of this puzzle was relatively simple, and most factions that engaged with the puzzle cracked particular challenge within minutes of receiving their assignments. I also built in a number of contextual cues to nudge teams towards interpreting FIND MORE DICE as hunting down other teams to see if their dice were different.
Starting at step 5, factions only received instructions for other factions’ dice sets: so, teams with the red dice received instructions on what the black and green teams should do. The upper left corner of every page also had a series of four colored squares, with their color emphasized to indicate positioning for the final solve. Even the dice bag contained a piece of paper asking for help…emphasizing the fact that they only had one part of the solution.
This meant that the hardest challenge of the puzzle was figuring out which teams had which colors, and trying to figure out how to create an alliance to get all four sets of dice in the same place at the same time. I calibrated the puzzle this way because the primary challenge of the puzzle should align with the reason people came to the event: if they came to a LARP to tackle inter-personal challenges and cross-faction negotiations, I’d give them a puzzle that relied on those skills to succeed.
And players exceeded those expectations by a long shot: one faction’s leader attempted to solve the puzzle by using their character’s mind control powers on me, to compel the solution out of me. This ultimately failed because my character didn’t know the solution. Another faction snuck into a team’s headquarters, and stole the set of dice they were missing to ultimately solve the puzzle. Unfortunately, they chose to steal the dice from my faction, and then came to me with their stolen wares to solve the puzzle. That gave me the distinct privilege of getting to harangue the winning team for stealing from the very person who gave them the task in the first place in-game, while congratulating them on a job well done out-of-game.
Solving the Puzzle: Serendipity and Rewards
While the Dragon Thrones GMs didn’t take me up on the offer to use the dice puzzle for world-building purposes, they did provide another unexpected boon: shortly after I completed building out the puzzle, they announced that conflict resolution for that summer’s LARP event would be resolved through d6 dice rolls.
And since I created a puzzle centered around d6 dice, I asked them if I could give out a single-use “automatic success” roll to faction members who successfully solved my puzzle. With that confirmation in hand, I was able to find a thematically appropriate solution to the puzzle: while the fifth side provided factions with the final instruction to USE THE FINAL SIDES, the ultimate solution hidden on the sixth side confirms THE DICE ARE LOADED.
I even brought a special pair of loaded dice that were almost comically weighted to almost always roll 6’s, to demonstrate how they helped me complete the “science experiment”.
Final Thoughts On Puzzle Design
My favorite question to ask when I’m sitting down to design a challenge or puzzle is what makes the scenario I’m designing for unique. With Dragon Thrones, the answer was the presence of faction-based gameplay. For a conference I designed a mini puzzle hunt for, it was access to Center City Philadelphia and its unique outdoor artwork.
There are absolutely some events where puzzlers assemble for the express purpose of tackling puzzles. But in so many other contexts, puzzles are used as social lubricant to support a bigger purpose. And whether that bigger purpose is tabletop gaming’s collective storytelling framework or escape rooms’ focus on collaborative team-based play, deeply interrogating the context behind puzzling can add layers to the experience.
