I’ve been thinking a lot about exploratory adventures — partly because of recent, very interesting discussions on the topic from Robert Schwalb et al., partly because I’m writing an adventure for publication and that’s one thing my assignment asked me to consider. This led me to think about PC experimentation within adventures.

What I think of as the old-school D&D tradition of having a machine full buttons and levers that PCs can experiment with has fallen out of favor recently. I’m going to assume that PC experimentation is a good thing — I think so because it’s essentially an exploratory activity, where one is given the option of testing a variety of options to see the outcome.

Now there are always going to be players that are going to want to push the button and pull the lever. My aim is to encourage the other players to want to do it too.

So why don’t some (perhaps many) players like experimentation scenarios as they’ve often been presented? Let’s go over some reasons.

The choices seem arbitrary. There’s no way to tell the difference between choosing option A or option B. This means that the players don’t actually use their intellect or creativity for this choice — they’re essentially just allowing the DM to do something to their PC. Also, they don’t get to roleplay — they don’t get to ask themselves, “Given what I know about these choices, what would my PC do?”

The choices are perverse. Let’s say the PCs are given information about the choices. What if this information is misleading? They may get the opposite of what they intended. This could make sense if an NPC is intentionally misleading, but the end result is that the PC got punished for making a logical decision.

Experimentation might not have any useful outcome, or even hurt the characters. What are the possible outcomes of experimentation? The choices might help the character; the choices might have no effect; the choices might have some effect that seem to neither help nor hurt the character; or the choices might hurt the character. Cautious players might choose to avoid harming their characters, especially if the choices seem arbitrary — it just becomes a random gamble, at that point, and one where they don’t like the odds.

So how can one design an experimentation scenario — let’s call it an X for short — in a way that encourages players to experiment?

One can always nerf the X — make it so that it only provides good things, or at least that no serious harm might come to PCs. In a way, this is what has been done with the wild magic of chaos sorcerers in 4e, where PCs get some random benefit. This has the downside of not posing any risks to adventurers, and thus not creating any tension.

One should probably allow PCs the chance of going back and experimenting with other choices. At the very least, this means that choices should be nerfed to the minimal extent of not instantly killing the PCs. (There are much fewer save-or-die situations in 4e now.)

If the choices are perverse, provide some clues that the choices might be perverse. I don’t think you want to play the game of “Acererak will always make bad choices seem attractive and good choices seem unpleasant.” What then happens is the entire, “Well, I’d know they think that, so I’m going to do the opposite and make good choices seem attractive and bad choices seem unpleasant. But I know they’d think that I’d think that, so etc.” One might want to do something like put a skeleton in the seemingly pleasant path, or some subtler hint — perhaps something that was a reward from an encounter (e.g. a clue from someone the PCs rescued), or something embedded in a puzzle/riddle/etc. the PCs had to solve.

The choices are not arbitrary, or not completely arbitrary. One either knows exactly what a choice might do, or at least get some hint as to what it might do. There’s something to be said about not knowing exactly what a choice might result in — the element of surprise can be a valuable one in an adventure. One might get hints a lot of ways as to what results might occur — either perceptions or insights at the space, or clues given beforehand.

The choices are arbitrary, but allow for trial-and-error. This makes the scenario more of a puzzle, which I’ll go into in my next post. In short, the PCs have to figure out what each choice accomplishes and then figure out the right combination of moves to get the result they want. This has the virtue of spurring players to experiment with every option, and then to use problem-solving skills to find a solution. Even if one or more choices are harmful, the players know if they figure out the right combination of moves, the puzzle can work to their advantage — a no-pain no-gain, high-risk high-reward scenario that is often attractive.

The choices are arbitrary, but mostly strange. The new 4e Gamma World alpha mutation deck seems to operate by that principle. One can easily imagine a 4e Wand of Wonder operating the same way. These strange outcomes can potentially spur creative uses of the results — I remember PCs in one of my 3e games took one of Quaal’s feather tokens to create a tree they used as a bridge.

The choices are arbitrary, but interesting. Hey, some players just like these arbitrary results, or PCs may prove so desperate they’ll try anything. The 4e Deck of Many Things fits into this category.

A neat thing about experimentation is that it could allow PCs to interact with the game world in a non-combat, non-skill-challenge related way. Of course, such experimentation scenarios might be found not just in non-encounter spaces, but as encounters unto themselves, or as parts of encounters.

Perhaps more importantly for adventure design, instead of having experimentation just consist of buttons and levers inside a machine in a space, one could have a set of doors in a room or passageways radiating from an intersection. In this manner, the same thinking that goes into experimentation can also go into the design of an exploratory adventure.


The standard adventure is a quest — the PCs have one or more specific goals, and the DM creates encounters for PCs as obstacles and aids.

One of the best things about 4e is how it created great tools and tips for designing encounters and quests. This applies to combat, obviously — how to use monsters tactically, how to use terrain, obstacles, traps and hazards, how to influence the movement of monsters and characters, how to structure XP, treasure and other rewards, how to pace challenge levels, numbers of encounters and rest periods, you name it, it’s there or is getting discussed. This also applies to non-combat encounters in the form of skill challenges.

However, I suggest that when it comes to designing exploratory adventures, one thing DMs might want to do is think more about non-encounter spaces and non-quest events.

Let’s call a space where an encounter occurs E. This includes both combat and non-combat encounters (i.e. skill challenges) and encounters that mix combat with skill challenge elements. Let’s call a space where encounters do not occur N.

In my last blog post I noted there were static spaces and dynamic spaces. I now say that *all* spaces are fundamentally dynamic, in the sense that decisions that PCs or DMs make can alter these spaces. One of the most common ways spaces are dynamic, in fact, are when PCs use combat to change an E to an N — that is, by killing off all the monsters within.

What kind of N are there, and what are they good for? Well, let’s start with the most basic kind — the empty room, which I’ll call Z. The most crucial thing about Z is that characters can rest there. Combat encounters and a number of non-combat encounters are exercises in resource management — how much damage monsters inflict, how much hit points and powers PCs have, how much treasure and XP, etc. — and Z help PCs replenish hit points and powers.

What other kind of N are there? Well, there are libraries or other repositories of information that DMs can use to convey information about the rest of the adventure, as well as potential future adventures, random information, and fun tidbits that enlarge the feel of the world the PCs are in. There are also vaults, supply closets and other spaces filled not only with treasure and plot coupons, but also tools that could potentially find use further down the line as well as just random junk that players may or may not find creative uses for. And there are just spaces that DMs can fill with flavor — strange apparitions, haunting music, mindboggling vistas, comic moments, and roleplaying moments with NPCs that are not either combat encounters or skill challenges but a way for PCs to learn more about the world or just have fun. Flavor may be fluff compared to the crunch of an encounter, but it subtly goes to enlarge the feeling of the world the PCs are in, which is what exploration is all about.

What else are Z good for? Well, they can be dynamic. They can go from Z to E, for instance, with the aid of wandering encounters. Weak arguments one could make about how wandering encounters add to an exploratory feel include them possibly doing so vicariously (if encounters are wandering the dungeon, that could mean they are explorers) and possibly doing so directly (if the PCs wandered past this room before when it was Z and returned to find it E, that means they were doing some wandering of their own). Stronger arguments one can make — if PCs see that spaces might change in their absence, they might want to investigate further or come back at some other time to see what else they might find; and a changing world again suggests a world larger than players might have thought before.

In fact, all N can be more dynamic than they’re often portrayed in adventures. There’s a long history in D&D of including strange machines with lots of buttons and levers that PCs can push. What inevitably happens, of course, is that no PC wants to toy with them for fear of something awful happening, but that kind of experimentation inside a space is a key ingredient in exploration — the feeling that players have choices that can affect the game world or their PCs. Simpler and potentially more inviting examples of dynamism can also be suggested — giving PCs the chance to destroy a bridge to divert access routes, for instance, or magic items that allow them to build or modify structures.

Another key and perhaps counter-intuitive way to make an adventure seem exploratory is to add encounters, treasure, information, etc., that have nothing to do with the quest or quests the PCs are currently on. You don’t want to carry this too far, obviously, or you’ll load down an adventure with too many distractions or dead ends. One advantage here is that one can set up story hooks for quests further down the line. Still, not everything in an adventure has to be related to the quest the PCs are on, or even related to quests they might eventually go on; this enlarges the feeling of the world the PCs are in, which is what exploration is all about.

What brought this all to my mind were the many interesting discussions conducted lately regarding the design of exploratory adventures in 4e, spurred on largely by an excellent recent blog post of Robert Schwalb’s. A followup from him here, an outside comment from SRM here, and an outside comment from the Chatty DM here. I feel incredibly presumptuous adding my two cents here, since Robert Schwalb, SRM, etc., have *loads* more adventure design experience than I do. I agree with Robert et al. that more should be done looking into Z. However, I would argue that besides resource management — how to structure encounters and rests in an adventure in order to budget how much damage encounters dealt out and how many hit points, powers, etc., that characters had — one actually has to consider the other aspects of Z and other N to create a more exploratory adventure.

In quests, PCs have specific goals they want to reach, but in exploration, PCs might not have specific goals at all — they might just want to see what they find. As such, encounters might not be the only thing they run across, and any encounters they find might not be linked to any quests they are currently on. Quests are mostly linear, while explorations are full of branches that may or may not lead anywhere — that’s what causes the PCs to explore around to look for what they want to find.

Now a problem with too many branches, as I mentioned in my last post, is that you can spend a lot of work designing stuff that might get wasted. So what do you do? One solution is that you don’t create more non-quest encounters, which take a lot of time and planning, but more non-encounter spaces, which take less time and planning. Another solution is that you create loops in branches, so that PCs can see all the branches more easily. What I think is likely the best solution, however, is that DMs create environments that players want to explore all of, as opposed to ones they feel obligated to explore.


So let’s start with stuff everyone knows, so there’s no confusion regarding terms.

Adventures are typically divided into spaces where events take place. I’ll typically refer to such spaces as 1, 2, 3, etc.

At their simplest, adventures can be designed as flowcharts. A traditional dungeon is a flowchart incarnate. It will prove useful in further discussions both as a base to build from and to move away from.

A linear adventure/dungeon typically goes 1 > 2 > 3 > 4 > 5 > 6 > etc. Pros: easy to design for DMs. Cons: players dislike having no choices. Of course, there are many, many other pros and cons one can list, but let’s start there.

A simple branched adventure/dungeon goes 1 > 2 > 3 > 4 or 1 > 5 > 6 > 7, with branch points at every single space. Pros: players don’t feel herded. Cons: a lot of work for DMs, and there’s no guarantee they can plan for every choice the players make.

The best use of branches allows players to see differences between the branches. A bad use of branches would involve two identical corridors, one going left, the other going right; a better use of branches would involve one being hot and the other being cold, etc.

One assumption made until now is that spaces are static. Of course, they might be dynamic as well. They can change without player intervention — they can switch back and forth (1/2/1/2) (say, when a patrol comes and goes), or change without repeating (1/2/3/4/etc.). They can change with player intervention, often with branch points involved (1/2 or 1/3).

One can have an entire adventure set in one space that just changes over time — see Aeryn Rudel’s “Dead by Dawn.”

One can imagine players going 1 > 2 > 3, then performing an action that changes 1 to 4, 2 to 5 and 3 to 6.

There are often loops in branches. This has significant pros — players can go back and see what the other choices might have been like without having to backtrack to the original branching point, and if they do so DMs will get to reveal spaces they designed, thus avoiding their work from being wasted. A con is that this can make adventures a bit more linear.

There may be secret or hidden branches. Pros: players feel superior for having discovered them, and their characters may receive significant knowledge, goods or services. Cons: if the characters don’t discover the secret or hidden branches, the DM will have wasted work, and the characters may not have knowledge, goods or services they might want or need.

Another possibility are locked branches. Characters typically have to receive plot coupons from one or more spaces before they can unlock a branch and proceed with the adventure. Plot coupons may take many forms — physical keys the characters must retrieve, clues they have to find that help them solve a mystery, or actions they must perform, such as defeating monsters, overcoming obstacles, convincing others or solving other kinds of problems. When done badly, players feel obligated to visit each space to receive the plot coupon; when done well, players want to visit each space to receive each plot coupon. These locked structures essentially take the form (1, 2, 3) > (4, 5, 6) > etc.

More soon.


I’m starting this blog to catalog my thoughts on adventure design. My next post will outline basics that everyone knows, and then I’ll try and go on to discuss concepts more interesting to me.