PF2 Structure Part 5.1 - Dungeon Structure

Go to Part 1

Hey folks!

From Shadowdark RPG

This is a weird one, because it's part 5a of a series for which I have not yet published part 1 - the overall series is about what I see as a need for more thoroughly fleshed out Scenario Structures in Pathfinder 2e. I have part 1 about 90% done, and the other parts are well under way as well, but I decided to go ahead and publish this one first because I think folks might find it useful!

Today, we'll be outlining a core dungeon procedure for PF2 that aims to fit as neatly into PF2 as possible without changing any rules. Ideally, it should work even if you implement it in an existing AP without doing a whole bunch of extra work. I'll also include some extra credit for if you do want to put in the work, or you want to change more about how you run dungeons.

Let's jump in!

Exploration Subsystem: Dungeon Crawling

When carefully combing through a dangerous location, time is divided into dungeon rounds, each of which is 10 minutes of time in the game world. Exploration is more free form than encounters, and within a single dungeon round a creature might ask questions, cast a spell, and search the room, and then cast more spells! They might even do one thing, wait for other players to declare their actions, and then jump back in. The following is meant as a guideline for how much can be accomplished each round, rather than a restriction on your creativity or 

Structure. Exploring dungeons or similar locations is played out in a series of 10 minute dungeon rounds, during which participants act in whatever order they want to.

Travel time can be ignored in most situations, as time is tracked in 10 minute increments and even the slowest party (performing activities like Search and Detect Magic) travels 150 feet per minute, or 1,500 per ten minutes. In cases where a journey is long enough not to be ignored, round to the nearest increment of 10 minutes.

Step 1: Start The Round

The GM should track the passage of time, and announce whenever an hour has passed. Players should track the durations of their spells, torches, and similar, decreasing the amount of time left by 10 minutes. Parties that struggle with tracking these things might want to consider a dungeon tracking sheet.

The GM will also check for complications by attempting a flat check. On a success, the PCs are safe. On a failure, a complication occurs. The DC depends on how much danger the PCs are in, though the details depend on what kind of complications you are using.

• Nearby Creatures. If running an existing dungeon that does not have a random encounter table, encounters will usually represent nearby dungeon denizens leaving their listed areas and stumbling into either the PCs or signs of their passage (e.g. bodies or other mess left behind). In this case, the flat check is DC 2 if the PCs are in rooms adjacent to such creatures, DC 3 if they are 2-3 rooms away, and DC 4 if they are either far away or  in a secluded area.
• Random Encounter Table. If using an encounter table, the flat check DC depends on how dangerous the area is: DC 4 for quite dangerous, DC 3 for moderately dangerous, and DC 2 for slightly dangerous. Random encounters will often be creatures, but might also be environmental effects or shifts in the dungeon.
• Narrative Encounters. The GM might instead tie complications to some sort of narrative events. For example, perhaps a rival adventuring party shows up or a timed sequence progresses. The DCs will usually increase over time, starting at DC 2, moving to DC 3, and then to DC 4 as the climax approaches.

The GM might decide to add other situational modifiers to this complication check. For example, if the PCs left a mess in an area where dungeon inhabitants are likely to go, the GM might increase the DC by 1, or by 2 if they instead left bodies lying around in such a location.

Note: by default, this means that there will be an encounter on a 1, a 1-2, or a 1-3 depending on the danger level. I personally prefer when high numbers are good for the PCs, and low numbers are bad for the PCs, but you can certainly run this as DC 20, DC 19, and DC 18 respectively if you want to match more closely to the encounter checks in Hexploration.

Step 2: Describe the Scene

The GM will give a description of important details, items, or creatures. Players should ask questions about anything that is unclear. The GM will either answer those questions if the answer is obvious, attempt a secret Recall Knowledge or Perception check on the PCs behalf if there's a chance they might know the answer, or suggest that the player take action to find the answer if the information is hidden.

Step 3: Declare Exploration Activities

Players describe what they are attempting to do, and the GM either decides what exploration activities fits best or determines how else to resolve them. Each player must declare their intent before any activities are resolved.

If the PCs are moving between areas, they should declare a direction and their relative positions as they do. In addition, they should each select travel activities. If players tend to select the same activities each time, the GM may want to write them down to speed up play.

Players might want to take shorter actions or activities, such as casting spells or using abilities. PCs can usually perform tasks measured in actions freely, and can do so at any point before or after everyone declares their exploration activities for this round. They may be able to perform a single activity of around 1 minute in addition to their 10 minute activity, though taking multiple such activities will likely count as their activity for the round. The GM has the final say.

Step 4: Resolve Exploration Activities

The GM resolves the exploration activities the players chose, asking for checks or attempting secret checks on their behalf as necessary. The GM narrates the results of each activity and answers any follow up questions or shorter activities.

Generally speaking, exploration activities that interact with the world such as Investigate and Search can target an area no more than 30 feet by 30 feet in 10 minutes, or 10 feet by 10 feet if doing particularly fine investigation (e.g. flipping through books). Detect Magic can cover a 10 feet by 10 feet area at lower levels, or 30 feet by 30 feet once heightened to 4th rank. In some cases, a GM might rule that PCs must get more specific about where they are focusing their attention (e.g. if there are both several bookshelves and papers scattered across desks, a PC must pick one of those to Search).

Step 5: End The Round

If exploration triggered an encounter, then the GM shifts the game out of exploration mode and into encounter mode. In this case, the GM should make a note of how much time has passed so they don't lose their place once the encounter ends. If the exploration of the location is completed, then the GM shifts the game to another mode of play, either downtime mode or a different form of exploration mode.

Otherwise, once this round is over, the next one begins. Repeat this cycle until one of the conditions above occurs.

Example of Play

A Fighter, a Rogue, and a Wizard enter a room. After checking for encounters, the GM describes a wide open chamber with an empty throne, several banners behind it, and a mound of corpses scattered throughout the room. The Fighter asks what the banners look like. The GM says they are purple with some sort of heraldic crest on them, and, rolling a secret Society check for the Fighter, says that the Fighter knows this is the crest of the local baron. The Wizard says he is going to scan the room with detect magic,* and the Rogue begins to go through the bodies for treasure. The GM asks the Fighter what their main activity will be, since small details like banner color are free - the Fighter's shield is still damaged from the last fight, so she will start repairing that. The GM rules those as Detect Magic, Search, and Repair activities respectively. She tells the Wizard of any magical items among the bodies, not requiring a roll since Detect Magic cannot miss their presence. She then makes a secret Perception check for the Rogue to see what she finds, and asks the Fighter to roll Crafting to see how much she repairs her shield. Once all that resolves, the cycle repeats!

*In this case, the GM and the Wizard have already established a routine where the Wizard starts from a location that does not ping with magic (ignoring all the party's gear) and then carefully moves through the room step by step, pinging whenever he senses magic. Whenever he finds magic, he guesses which item is the most likely to be magic, pulls into a magic-less area, and tries again, repeating the process with other items if it doesn't work. If he had 4th rank spells, this could be much faster, as he located the highest rank effect, ignored it, and found the next highest. GMs might want to strongly consider simply allowing casters to use this method even if their players don't think of it, as it would likely be standard practice in all places of magical learning. The first option might take more time, especially in rooms that are particularly large or filled with items.

Defining Activities

You can broadly speaking fit activities into four categories. This section is mostly here for added GM clarity, it's not written like proper rules.

Travel Activities

These are the exploration activities listed in the exploration mode chapter of the player core: Avoid Notice, Defend, Detect Magic, Hustle, Investigate, Repeat a Spell, Scout, Search, Sustain an Effect, or Follow the Expert to mirror someone else's choice.

Hustle is mostly going to be for rushing back through a whole dungeon or running away from foes. Track travel pace as normal.

Some of these are more or less unrelated to time. Avoid Notice, Defend, Repeat a Spell, Scout, and Sustain an Effect can be done for 1 minute, 10 minutes, or any other amount of time. Generally speaking, these are preparatory, giving you a benefit when you need it but otherwise fading into the background.

The more active ones - Detect Magic, Investigate, and Search - can also theoretically happen on different time scales, but they'll work a little differently.

• Detect Magic. If you cast the spell once, it just gives you a yes or no. If you want to use it to find all the magical items or effects in an area, that will take 10 minutes in the same room. From spell ranks 1-3, you can cover a dense 10x10 foot area or a sparse 30x30 foot area. For spell ranks 4+, you can comfortably cover a 30x30 foot area. While traveling, you will notice when you get within range of something magical, but unless you have higher ranked versions of the spell, you won't know where or what it is.
• Investigate. If you spend 10 minutes examining aspects of a room, you can make a check regarding about 30x30 feet of wall carvings or similar, or briefly flip through a couple books. More intensive study might warrant the research subsystem instead. While traveling, you will recognize anything that might warrant closer inspection, but won't make a check yet.
• Search. If you spend 10 minutes searching through a portion of a room, you can make a check regarding about 30x30 feet if sparse (e.g. flipping through drawers or cabinets) or a smaller area if denser (e.g. 10x10 for a dense battlefield full of bodies). While traveling, you might notice a trap or hazard, but would have to examine it closer to get more details.

Stationary Activities

You can do any of the travel activities above while stationary as well! In addition, the following activities all take 10 minutes by default: Affix a Talisman, Compose Missive, Decipher Writing, Identify Alchemy, Identify Magic, Impersonate, Repair, or Treat Wounds. 

Minor Actions

Anything that is measured in rounds instead of minutes is effectively free. The GM may allow a PC to take a single activity taking up to 1 minute (like Call Companion or Pursue a Lead) in addition to your main Exploration Activity, or might let you do several such shorter activities instead of taking a longer one.

Analysis

These changes give several benefits to PF2 dungeon crawling:

1. Time. While the existing Exploration Rules gesture towards 10 minute rounds, they don't make this explicit, which can make tracking things much harder. Note that GMs can still skip over longer periods of time when necessary!
• I also give at least brief advice on how large an area the PCs can Search, Investigate, or Detect Magic in each round. It's possible that more advice here would be productive, but I think this is a good start!
2. Structure. It's a lot easier to track durations when the GM is marking down every 10 minutes. It's a lot harder for one player to dominate play when they've already taken their main activity for the round. Likewise, the explicit reminder for the GM to re-set the scene and for the players to ask questions can help avoid confusion.
3. Tension. Now, choosing to rest (or even where to rest) is a meaningful choice. This gives more value to feats like ward medic and revelation's focus, but more importantly gets PCs thinking about how they want to approach the dungeon. Is there a safe place they can rest to avoid wandering monsters? Is there a way for them to secure the doors so that anyone coming by will have to break through them, likely making noise and alerting them in the process?
• I went with flat checks for complication checks because that fits the core PF2 rules better, but you could just as easily roll a d6 every 1-3 rounds instead like many old school games do.
4. Dynamism. Instead of a static series of rooms for the PCs to explore at their leisure, this system, without any additional work, already creates at least some feeling that the dungeon is alive! Especially if you use "Nearby Creatures" encounters, even just the idea that creatures might not stay static can drastically change how your players approach things!

Something I see a lot on reddit that I want to get ahead of is the idea that "Pathfinder 2e expects PCs to be at full health." That is sort of true, in the sense that the encounter guidelines do indeed assume that the PCs won't be particularly low on resources. Indeed, this section says "a lack of resources can easily turn a severe-threat encounter against the characters!" However, there's a difference between "these labels are the most accurate when the PCs have sufficient resources" and "the game assumes the PCs will be at full health." Remember, even Severe encounters are still tilted in the players' favor, and are only likely to turn against them if things go extremely wrong or the PCs are particularly hurt. Most encounters are much less scary! They are certainly likely to be harder if the PCs are injured, as this post goes into, but I have often seen my players go through several encounters in a row before needing to rest.

All this to say, if your knee-jerk reaction is "let the players rest! The game requires they be at full health!" then my recommendation is 1. Trust the system - it won't break just because the PCs are a little banged up! And 2. Trust your players - ultimately, I have faith that every group of players can learn to manage these sorts of situations and avoid getting in over their head! And if they fail, that will be an interesting story and a good learning experience.

Extra Credit

I firmly believe that adopting the above procedure with no other changes can immediately and positively impact your group's dungeon crawling experience, especially if you don't have another procedure that you tend to use. Hopefully, since by default complications can easily just be nearby creatures wandering by and 

However, there is still more that we can do to improve dungeon crawling in PF2 beyond just that procedure if we are willing to put in the work!

Run Intelligent Baddies

As I've talked about before, dungeons are much more dynamic and engaging when bad guys react intelligently! At a minimum, I highly recommend having overmatched bad guys (or wandering complications / scouts) retreat and/or run to get help. This does two primary things:

1. It creates a sense of realism, as creatures have a realistic sense of self-preservation. They stop being static baddies to fight and turn into people with desires and agendas.
2. It opens up the realm of possibilities. I love running dungeons as a Theater of Operations, with fights sprawling between rooms and into corridors, and starting the PCs in initiative chasing down fleeing creatures in close quarters automatically leads to things spilling beyond single encounter areas.

GMs will want to be extremely careful about how much to combine encounters - remember, 2 Trivial encounters is a Moderate encounter, and 2 Low encounters are a Severe encounter, but 2 Moderate encounters are an Extreme encounter. For best results, GMs will want to do some balancing on the fly, potentially finding reasons for some of the creatures to stay behind - maybe they're guarding something? Nervous? Dismissive of the guard who alerted them? Napping or otherwise slow to react? Or maybe they just need to spend a round or two grabbing their gear!

The way that I usually manage this is to only have foes flee towards the end of the fight. That way, even if they get away, encounters aren't usually being combined so much as strung together. That being said, I think it is 100% okay for PCs to get in over their heads if they are reckless, as long as you set the expectation appropriately! For choices in exploration to be meaningful, there has to be a chance of failure, even if the worst case scenario is pretty rare.

If you want to go beyond merely grabbing reinforcements, you can start designing dungeons with mobile patrols, have enemies set ambushes or barricades, or take other intelligent actions - in this case, I recommend using an adversity roster to track movement (or just move the tokens invisibly on the VTT if you're playing online). See my Abomination Vaults Retrospective for some examples of what that can look like in practice!

Matryoshka Searches

We've talked a lot so far about dynamic dungeons, but so far we've kept that mostly constrained to how we can make combat encounters less repetitive and more mobile. What about things other than combat?

One of the simplest ways that I've found to really bring out the feeling of exploration is the Matryoshka Search Technique. My personal preference is also to minimize Perception checks when exploring dungeons - I much prefer things like "if you Search this rubble, you will find [x]," though you may need to make a check to identify it. I love putting hints in my initial descriptions of what things the PCs might find, so that PCs feel rewarded for paying attention and asking followup questions.

Applying this technique to APs takes a little bit of work. If you're really good at improvising details like this, then go right ahead! Otherwise, it might not be a bad idea to add little sticky notes to your book as you read through to call out where secrets like this live and what signs the PCs might imagine.

Adventure Design: Building In Secrets

Most dungeons in APs are fairly linear, and don't have a ton in the way of cool secrets to discover. There are a few exceptions, but when there are, they tend to feel a bit random in terms of whether you find them or not. I'm not going to get too far into this topic now, because "how to design dungeons" is a pretty big topic! For right now, all I'll say is the following:

If you are designing your own dungeon, try to create secrets that the players can feel really cool for uncovering! Then, try to think through 1-3 signs that might guide the PCs towards discovering them. Things they'd notice, patterns that repeat, or details that seem off but point to the truth. Examples include:

• Secret doors that follow a specific pattern (ideally with clues like one being left open or being old and easier to find), so that once the PCs clock the pattern they'll feel like they understand the dungeon better.
• Secret elements or potential changes to the dungeon, like a big underground lake being held back by several flood walls, with several signs like water-based plants cropping up in the parts of the dungeon nearby, even if there's no water visible.
• Elements that the PCs can use to their advantage, like hazards, teleportation circles, secret hidden rooms that are 100% safe, or powerful artifacts that give a temporary advantage
• Factions and creatures that the PCs can make alliances with or turn against each other (the creatures likely won't be secrets, but their agenda's will be! make it worth the PCs while to try to figure them otu)

If you are running existing APs, it might not be worth your time to add in such elements yourself, though obviously feel free if you're excited about it! However, if there are secret elements - for example, the myriad secret doors in Abomination Vaults - try to think through what signs might point to them ahead of time, just like we were talking about with the Matryoshka Technique.

What we don't want is for PCs to check every corridor for traps (at least not more than choosing the "Search" activity), check every wall for secret doors, and slow down play to a crawl to avoid missing anything. What we do want is for PCs to be able to solve the dungeon like a puzzle and use clues and signs to make discoveries!

Designing Complications

If you're using an encounter table, what goes on it? I recommend a mix of combat threats, guards/patrols that will go get help, potentially friendly NPCs, and environmental effects or dungeon shifts.

If you're doing a more narrative style where each Complication advances the baddies goals or progresses the timeline, what is going on narratively? And how can the PCs know that?

I think that's all for now? Not bad for a first draft! Hopefully, the nest post in this series will be #1, but who knows maybe I'll get really excited about Point Crawls or something. I'll definitely update the framing around this piece once I have some of the other posts up. Basically, #1 is the intro, #2 looks at subsystems like chases, #3 looks at Infiltration specifically, and #4 looks at the other existing PF2 Scenario Structures. Part 5 could honestly probably be a series of its own, basically focused on creating new Scenario Structures for PF2 that are easy to run and integrated into the core rules. I don't have an exact order that I'm likely to go in, it'll honestly probably just be whatever I'm feeling excited about!

More than likely, my next post before that will be on the epic finale of Strength of Thousands - regardless, see you soon!