Mining Dungeonaday.com

Monte Cook is one of the greats in D&D, and I have a lot of fun reading Dungeonaday.com, where he lists a new section of a dungeon, well, daily.

In going over puzzles, one classic idea presented here (paywall) and in more than a couple of computer games is the notion of pushing a button in one room and having a door open in another room. This might involve finding what door the button opened; it might involve running as fast as you can to get to that door; it might involve teamwork to have someone keep that door open.

Just another wrinkle to throw in. There’s a lot of good stuff there! I highly suggest you subscribe, if you haven’t already.

So I talked about conundrums in the last post — what I call puzzles that players solve by figuring out the right action or combination of actions to take. Let’s see how we might be able to design them using the guidelines for traps or skill challenges or not. Two kinds of conundrums will be analyzed — one, posed by Steve Winter, involves PCs trying to figure out how to get an object on the other side of a chasm; the other, posed by me, will involve a machine with several levers and buttons that PCs have to operate properly.

It became evident to me that puzzles represent a different way of having fun than other types of encounters or encounter elements, and as such have to be designed different ways. The difficulty then seems to come in how to scale them properly — can one have a level 6 riddle or an elite leader crossword puzzle?


One can either to get to the other side and back, or find some way to retrieve the object without crossing the chasm.

One can imagine PCs using mundane equipment they have, such as rope for a lasso, or a sunrod thrown into a dark corner to see what’s there, or food to lure an animal over. One can also imagine manipulating one’s environment in some way, such as pushing a log on over to create a bridge, using a ladder to reach a ledge, swinging on a chandelier, or using a telescope to view a distant inscription. One might also use a plot coupon of some kind, such as a key.

There might also be buttons or levers that can extend bridges, lower ladders, reveal secret passages, etc.

In many cases, in this scenario it’s more interesting to use equipment that is not a weapon or a magic item to solve a puzzle, or to use a weapon in a non-combat way or a magic item in an unintended manner. For instance, one might use an arrow to hit a distant button to extend a bridge.

When it comes to crossing the chasm, one can imagine just using a power, ritual or physical skill, but that seems too easy, and take the fun out of things. The most fun option would seem to be for players to figure out the right way to use a skill or power in a clever way — say, to go to the right place and jump, or to use a blast of fire to set something on fire at the right moment.

To make it more challenging, one can make it capable of inflicting damage in some way — say, by using a pit of fire instead of a bottomless chasm.

So is this a skill challenge? One can make it one, but that would seem to make it boring — it seems to me that you don’t want to give PCs a limited number of chances to pass, but allow them to try a couple of different strategies with the luxury of failure. Is it a trap or hazard? It’s similar in that one has to consider all the “countermeasures” a PC can take to solve it, but it doesn’t have to damage or otherwise inconvenience PCs, and it doesn’t necessarily matter how well a PC does in a skill check or a combat check. Puzzles such as conundrums represent a different way to have fun than combat, skill challenges and traps — the fun in it is not in exercising the skills or powers the PCs are best at, but in figuring out the proper course of actions to take.


Let’s say there are a number of buttons and levers PCs can use to operate, say, a crane to pick up an object. One button raises the crane, one button lowers the crane, one button grasps an item, one button lets it go, and several levers move the crane forward and toward, left and right.

This isn’t a very exciting puzzle just by itself. One can make it more exciting a number of ways:

* One or more of the controls releases an electric shock, is hidden behind a curtain of fire, releases a monster, starts lowering the ceiling, etc.

* Two or more controls need to be operated simultaneously, or in a certain order, and PCs have to be in different locations and coordinate their efforts.

* There is a penalty for failure. This could mean damage, summoning of monsters, not getting a vital resource, the fact that they have to go around the machine into dangerous areas or a longer path, etc.

* There are a limited number of moves the PCs can make before they fail the puzzle, or a time limit the DM gives players.

Is this a skill challenge? Potentially — instead of having players try each control, you can just have them run Arcana, Thievery, etc. checks to simulate them using their knowledge to figure out what the right combination of skills is. Is this a trap or a hazard? Again, potentially — one can use a number of skill checks and dare a number of risks as “countermeasures” to “disarm” the device. However, treating this as either a skill challenge or a trap might remove the fun of figuring out the right order in which to carry out actions. Both skill challenges and traps are fun, but there are more ways to have fun than either of those choices.

So what actions can PCs take with either challenge?

* Minor actions and movements such as flipping a switch, pushing a button, throwing a rock, opening a door, swinging on a chandelier, lifting or dropping an object, etc., which don’t usually play much of a role in either traps or skill challenges. It can also involve using equipment other than a weapon, armor, or a magic item — say, a rope, a ladder, a ball.

* Ability checks could be more appropriate in certain situations, more so than skill checks or combat — say, to move a wooden log or bend a metal bar.

* Roleplay might work as well — one might imagine a sentient trap, for instance, that talks about clues.

* Story elements such as instruction manuals or codes in poems could be vital.

* When it comes to skills, Perception is usually key with traps, and often takes center stage here with conundrums. Insight might work as well — for instance, to figure out the purpose of a machine and what actions people might want to take with it. Thievery is often key to solving any trap or machine. Knowledge skills such as Arcana and History could either help fix this machine, or lead to clues on how to fix it. Physical skills could help, say, reach certain parts of the puzzle. Social skills such as Bluff or Diplomacy might not play a role in many puzzles, although that just begs some DMs to come up with ones that do.

* Combat is always an option, and while it is crude, it could be a way to get combat-oriented PCs or players involved.

In short, the fun of puzzles seems in large part to be in finding out the right choice of actions to take as opposed to being good at trying to do your best at skills you are good at. It also involves trying out actions different from ones normally used with combat encounters, skill challenges or traps and hazards.

The chasm and the machine seem to represent two different kinds of puzzles. The machine seems to have one best solution, while the chasm could have many different solutions. Puzzles with one best solution seem harder to design to me than ones with many different solutions, since the former involves limiting player choices, and in games like D&D, it can be very hard to limit player choices and figure out every single option they might take. (How many times has the power of flight made a D&D challenge absurdly easy?)

At the same time, if there is only one best solution, it can be fun for players to try and figure out what the right choices are to solve the puzzle — you just can’t make the process too infuriating. Similar kinds of puzzles with just one choice that players have to guess include riddles, rebuses, passwords and so-called lateral thinking puzzles. I especially love the former (the albatross puzzle being my favorite), and they could provide vital clues on how to design similar single-choice puzzles well. Make the practice of exploring possible answers fun — whenever the players explore what choices they have, don’t just say, “No, you can’t do that,” but also give clues as to what the right answer might be, and give an interesting bit of color or something else nice. In other words, reward players even when they make a “wrong” choice, so they feel motivated to find the “right” solution.

The problem I run across again and again is whether or not such conundrums can or should be fit into encounter building mechanics that exist for combat, skill challenges and traps. What is the encounter level of a puzzle? What is the XP budget? Is there a grade of complexity as with skill challenges?

I think it’s at least possible to figure out a basic level for puzzles — look at DMG2 p. 65, figure out if you want to make it easy, medium or hard for PCs, and set the level appropriately to get DC values and damage expressions. At every step in a puzzle, offer PCs ways to get a piece of the puzzle with a die roll as a “success,” and if the PCs accumulate enough successes, you can consider them to have solved the puzzle — but each step of the puzzle can be solved without die rolls as well with sheer player brain power. Perhaps the number of successes they need can be determined using the skill challenge complexity system (DMG2 p. 80, Rules Compendium p. 159). I do think that puzzles should be more like traps than skill challenges when it comes to failures, tho — certain puzzles can involve a lot of trial-and-error, so players should be offered an unlimited number of chances to fail, or at least a large number. Perhaps some of the successes could be determined by skill tests or group checks instead of single skill checks?

What do you think? Is all this analysis of puzzles worth it? How might one come up with a better and/or more unified set of mechanics for puzzles?

I’d like to talk about what I call “conundrums” — what they might be, and how we might design more of them. Let’s first go over some basics, such as what goes into encounters and puzzles.

The basic actions that a PC can perform in any encounter include:


Skill checks.

Roleplay: basically, conversations of any sort. There can be overlap here with skill challenges.

Story: this includes reading of texts, etc.

Move actions: not just the PCs moving from one place to another, but also pushing buttons, flipping switches, carrying objects, etc.

Ability checks.

Combat encounters obviously involve mostly combat, and skill challenges obviously involve mostly skill checks. Story encounters involve mostly roleplay and story.

So where do puzzles fit into this? And what about scenarios such as the following proposed by Steve Winter? “A dungeon room bisected by a bottomless chasm, with the desired object on the other side. Players need to figure out how to get over there, recover the object, and bring it back again, using their powers, their gear, objects they can find in this room, and objects they recall having seen in previous rooms.” Is that a skill challenge or a puzzle or something else?

I’ve talked a bit regarding puzzles. When one thinks about puzzles, one usually thinks of mazes or riddles. If a puzzle does not fall into any traditional category, let’s call it a conundrum, especially if it’s a puzzle that players must solve using actions. In that box we put in Steve’s scenario, where PCs are given a desirable solution but no clear actions they can perform to achieve it, and the trial-by-error scenarios I mentioned here, where PCs are, say, shown a machine with buttons and levers and have to figure out the right action or combination of actions to perform to achieve the desired solution. (Mysteries can be seen as conundrums as well, but they’re really more a form of adventure than a single adventure element.)

Now puzzles may be designed as skill challenges — DMG p. 84 gives an example. Still, not everything that involves skill checks is necessarily a skill challenge — alternatives include group checks and skill tests, as DMG2 pg. 83 points out. Let’s talk about the puzzles that can’t be designed as skill challenges.

The hard part about designing a puzzle is that one can’t give, say, a crossword puzzle a level value. However, since conundrums involve actions, the virtue of a conundrum is that one can at least assign DCs to various rolls, and use skill checks and ability checks as a way to provide hints if the players are stumped. Mechanically, conundrums may resemble either traps or skill challenges or both, and incorporate elements such as skill tests and group checks.

Here’s how one might design a conundrum, tips that might carry over to any puzzle:

Is the conundrum an encounter? That is to say, are there benefits if one solves the puzzle and risks if one fails to solve it? The DMG would argue that not every puzzle is an encounter. While I would say that’s true, I would have to argue that every interesting puzzle has to be an encounter — otherwise, why do it?

Is there a way to continue with the adventure if the players cannot solve the conundrum? This is a consideration every adventure element must face — there must be a way around the puzzle if the players cannot solve it. If the players cannot solve the puzzle, consider that a failure.

What are the consequences of failure? PCs might then have to go into combat, or engage in a skill challenge, or take on a penalty, or not have a benefit available to them.

What action or actions must PCs take to solve the conundrum? These include speaking the answer to a riddle aloud, pushing buttons, moving objects and the like. When designing the conundrum, DMs should consider every action a PC might take to solve the conundrum — not just skill checks, but combat, roleplay, etc. — just as they might when designing traps. If, during the course of designing or even running the conundrum, a DM finds that a lot of skill checks are being made, the DM might want to evolve the conundrum into a skill challenge.

How does one determine failure? The most obvious answer is, “If the players fail to carry out the correct actions successfully.” If, however, the DM finds the puzzle is more like a trap or a skill challenge, then perhaps the rules for those might be considered instead.

What action or actions can PCs take to get clues on how to solve the conundrum? Once a DM figures out what actions PCs can take to solve the conundrum, the DM can then figure out what actions PCs can take to get clues as to what the right actions are. Figure out what clues might work, figure out where they are placed, and then figure out what skill checks or other actions PCs might perform to help get them those clues. This includes the “get a clue” Intelligence ability check mentioned in DMG p. 81. Now, just because skill checks are involved, that doesn’t mean this is a skill challenge — conundrums (or at least parts of them) could be resolved by things such as group checks (everyone makes checks using the same skill, at least half must succeed) or skill tests (a number of successes are required, but unlike with skill challenges, there is no penalty for failing three checks).

Assign the conundrum a level value. Follow the mechanics used for designing traps on DMG2 p. 64-5.

Does the conundrum have the capability to inflict damage? If so, follow the mechanics used for designing traps on DMG2 p. 64-5.

Why is this conundrum here? This isn’t strictly necessary from a design standpoint, but is good to understand from a story standpoint, just as it is when placing a trap or a hazard in an adventure. Was this conundrum intentionally placed here, like a trap? Or is it more naturalistic, like a hazard?

Most if not all conundrums are in practice unbounded — although the DM might in theory have one or more solutions for a conundrum in mind, players inevitably seem to come up with solutions the DM did not foresee. That’s all right — it’s part of what makes conundrums fun for everyone. (The DM still should have at least one solution in mind, however — there’s no point in creating a puzzle without a solution.)

In many ways, conundrums resemble traps and hazards — perception can play a key role in solving them, and DMs must design a list of countermeasures PCs can employ. I suppose the difference is that there’s not really a trigger, and the price of failure is probably often not damage. Perhaps one can think of conundrums as hybrids between traps/hazards and skill challenges?

Oh, the same rules for making traps fun (DMG2 p. 70) also apply for puzzles and conundrums, I think.

Well, that was fast. Let me ditch the clunkily-named “non-encounters” and call them “story encounters” instead.

So let’s come up with a taxonomy of encounters to shed light on how one might better design them. We have combat encounters, story encounters and skill challenges (or “skill encounters,” if we wanted everything to fit neatly together). Of course, these divisions are arbitrary — each encounter type can include elements of the other.

I’m probably the only one who’ll agree, but I would argue that empty spaces are themselves a kind of encounter type in and of themselves — the “empty encounter” — or at least are vital in terms of thinking about planning encounters. At the very least, they’re needed to replenish resources such as hit points and powers. They’re also potentially combat or other types of encounters, if wandering or recurring/scheduled encounters are taken into account. They also allow players the chance to roleplay only with each other, which is interesting in and of itself. If empty spaces aren’t encounter types, they at least have to be considered as encounter elements.

So what encounter elements are there then?

Monsters and NPCs.

Skill challenges.

Terrain, obstacles, and lighting.

Story elements.

Traps and hazards.

Puzzles and what I’ll call curios. (More on curios below.)

Treasure, rewards and other items, including healing, as well as useless, apparently useless or harmful items.

Let’s try and develop a taxonomy of story elements, as well as one of puzzles and curios.

What knowledge, goods, services, monsters, NPCs, places, etc., might serve as common story elements?

Plot coupons unlock certain branches of the adventure.

Hints and clues shed light on upcoming encounters or adventures or on past encounters or adventures. Rumors are kinds of hints and clues that PCs might gather or overhear. Exposition essentially consists of long blocks of hints and clues that give vital context to PCs as to what is going on and what has happened or might happen.

Flavor and fluff sheds light on the game world at large, and may serve as hints and clues. Although fluff and flavor might otherwise seem purposeless, a clever player can use them as ingredients in strategies to use in encounters, and DMs can use them as the foundation for future story hooks.

Enigmas, which include strange apparitions, weird sounds, bizarre NPCs and other mysterious elements. Enigmas could be hints and clues, could be flavor and fluff, or could turn into other types of encounters.

So what kind of puzzles are there? And what is it that I’m calling a curio?

I would argue there are bounded and unbounded puzzles.

Bounded puzzles are ones where PCs are given a specific number of choices. These choices may be helpful, harmful, neither or both, or have no apparent effect. The right choice or combination of choices will help the PCs reach desirable solutions. The PCs will be given varying amounts of information about the nature of these choices in order to help them decide between different choices. Bounded puzzles include trial-and-error puzzles, cryptograms, word searches, quotation puzzles, crossword puzzles, logic puzzles, and mazes, as well as certain kinds of mysteries.

Unbounded puzzles are ones where PCs where PCs do not have a specific number of choices available to them. The right choice or choices will help the PCs reach a desirable solution. The PCs may or may not have hints or clues as to what the right choice or choices might be. Unbounded puzzles include riddles, rebuses and password puzzles, as well perhaps as certain kinds of mysteries.

When designing puzzles, one must not only consider what the choices, hints, consequences and solutions are, but also their creators and the reasons for their creation. They could be machines without instruction manuals, creations of madmen, clues inadvertently left behind by criminals, combination locks, or tests of worth. I’ve done some rumination on what makes a good puzzle here and here.

And what are curios, then? They’re pretty much just sets of choices and effects that occur for no express purpose. These might include malfunctioning devices, odd mutations, space-time warps or creations of madmen. They’re just strange curiosities. I’ve done some ruminations on PC experimentation here, which might come to bear on both puzzles and curiosities.

So I spent a post talking about non-encounters, and they’ve been cropping into every post here since. I’m seeing a greater push towards including such elements into 4e books lately — MM3, for instance, although one could argue that it began at least in PH2 and PH3, when story elements were integrated more deeply into class back-stories.

I do think that “non-encounter” is kind of a sucky name, though. I could go the acroynm route and call all non-encounters “non-encounter material” or NEM. There’s certainly a long history in D&D of acronyms (THAC0, anyone?), and some have still stuck around — PCs, NPCs, DMs, and of course D&D itself.

Non-encounter material has often been called fluff or flavortext. Although I think non-encounter material includes fluff and flavortext, other non-encounter material includes puzzles, libraries and the like, all of which can serve as vital adventure elements.

I’m stuck in terms of what else to call non-encounter material. Any ideas?

Bruce Cordell let me know he linked to this blog from Twitter. Thanks, Bruce!

In the course of conversation with Bruce, an interesting question came up — how does one drop in a puzzle without them disrupting the flow of an adventure?

There are probably a couple of ways of covertly slipping in a puzzle structure into an adventure and an encounter, Mysteries are one way to slip puzzles into an adventure, while machines that have several buttons and levers that can affect terrain are probably a way to slip them into an encounter.

If you want to be overt about a puzzle, maybe you can let PCs know that a puzzle is coming, and that there are clues they can find to solve it or bypass it. They can then spend a few encounters or non-encounters doing so.

Another way to include riddles into an adventure are with the sphinxes reintroduced into 4e in MM2, or with some other riddle-telling monster, like a dragon, Gollum, etc. In this way, players have the option of either a combat encounter or a non-encounter puzzle. Again, one can provide clues well beforehand as to what the answer of the riddle might be.

There’s always the natural question, “Why did someone spend all this time making a giant magic crossword puzzle?” I do like the logic used in the Tomb of Horrors — “Because Acererak is nuts.” There are two basic sets of reasons why someone might intentionally create a puzzle — rational and irrational.

There are all kinds of irrational reasons one might have to create puzzles. Acererak has a prove-he’s-better-than-everyone-else personality. Other irrational puzzle-builders might have sadistic cat-and-mouse, play-with-your-food mentalities.

One rational reason could include the puzzle essentially serving as a combination lock, and the clues as to what the correct combination is were seeded throughout the adventure up to that point. Another common rational reason for a puzzle in adventures is the desire to test a PC’s creativity and reasoning skills to see if they are worthy.

A few thoughts on puzzles

So I mentioned puzzles in my last blog post. Let’s examine them further here.

There is some support in 4e for designing puzzles. There’s not a lot, though, as far as I can find — just a few pages in DMG1.

I’m writing an adventure for publication, and puzzle-solving is one thing my assignment asked me to consider. I’ve been researching puzzles on Wikipedia, in computer-based RPGs, on some ideas as posed by a UK mathematician, and in some old D&D work, such as that of Johnathan Richards, who wrote classic adventures such as “The Challenge of Champions” series or “Gorgoldand’s Gauntlet.”

Let’s define a puzzle as one where one must figure out the right combination of actions to reach the desired solution. (I think there are more kinds of puzzles out there, but for purposes of adventure design, let’s stick to this one first.) One might solve these kinds of puzzles by recognizing patterns or by relying on inquiry and discovery.

A maze is a kind of puzzle, in this sense. One is given a number of intersections, and then one has to figure out the right combination of moves to make to reach the end. However, the choices one makes in a maze can often seem quite arbitrary, and they lack the kind of details that encourage roleplaying, so mazes have fallen out of favor in 4e, and perhaps rightly so.

Other kinds of famous puzzles include river-crossing puzzles, such as how one brings a fox, a goose and a bag of beans across the river safely, or the hell-heaven puzzle where one has to figure out who is lying and who is not to get where one wants to go. In computer RPGs, PCs are commonly asked to push blocks around in the right order, or pull the right combination of levers, or open and close the right combinations of doors, or put objects in the right groups or order. Again, these all involve figuring out which action or combination of actions is the right one to choose.

Many mysteries are probably puzzles in this sense as well. In many, PCs are given a specific range of suspects, and through inquiry figures out which suspect has the right combination of clues to match the perpetrator. In others, PCs don’t have a specific range of suspects, but by finding the right combination of clues, they at least get set on the correct path to find the perpetrator.

Word and number puzzles such as crossword puzzles, word searches and quotation puzzles all have found their way into D&D adventures, but I’m not such a huge fan of including them myself, since as with mazes, there’s not a lot of role-playing involved there. Cryptography of all kinds involves puzzles as well, if you look consider finding the right combination of number-and-letter substitutions a kind of puzzle.

All these puzzles offer PCs clear choices. They differ from, say, riddles and rebuses, where PCs don’t have an obvious choice and must figure out the answer some other way. (All puzzles where one has to try and figure out a password are riddle puzzles, I think.)

How does one design fun puzzles? Let’s look at the problem-solving process.

First, as mentioned in the last post, one can provide clues so as to keep choices from being arbitrary. PCs might discover these from additional perceptions or insights at the space where the puzzle is located, or in spaces that PCs visited previously. An old tradition in D&D is to provide some kind of poem/song/riddle or map or the like that provides clues, if PCs can decipher it. There’s also an old tradition of D&D of knowing what might happen from meta-knowledge — for instance, if there’s a frog, they might know that kissing it can turn it into a prince. (One can replace that route nowadays with a skill check.)

It’s also important that PCs have the option of trial-by-error to figure out the right action or combinations of actions. This means that PCs get to make several choices, as opposed to getting only one. At the very least, this means that choices can’t be fatal or crippling. If you want a trial-by-error puzzle, it also means that you want to avoid situations such as “Is that your final answer” or “Is that the door you are choosing.” Such “final answer” situations might be appropriate where PCs solve the puzzle by process of elimination, e.g. by narrowing down suspects in a mystery.

Some good points brought up here include the fact that, just as with an encounter, there has to be a way for the PCs to progress even if they can’t solve the puzzle. (Often times in adventures, it seems PCs choose to progress past puzzles and find out the solutions afterward.) Also, puzzles should try and incorporate the whole group, or at least give those who don’t want to solve the puzzle something else to do. In this sense, there’s a lot of intersection between puzzles and skill challenges, but not every puzzle is a skill challenge, and not every skill challenge is a puzzle.

Riddles and rebuses seem difficult to incorporate into adventures — they don’t offer clear choices. The best one could do is provide a number of clues as to make them easier to solve. One might also create the illusion of a riddle or a rebus by providing information on a number of topics beforehand, and the PCs then must essentially draw on this knowledge to answer the riddle or rebus — in this case, the riddle or rebus is actually a choice-based puzzle in disguise.

Puzzles can be spread across more than one space. This sense of movement and inquiry is something valuable for both quest-type and exploratory-type adventures.