gfishfunk
2016-08-18, 04:59 PM
I am mostly fleshing this out for myself, but I thought I would share it to get feedback, etc. Either way, if you like what you read or are critical of it or whatever, please comment as it does encourage me to keep hashing this stuff out. This is my prior post on individual enemy design (http://www.giantitp.com/forums/showthread.php?497787-Enemy-Design-Philosophy).
Version 1.1: Rewording. Completing Incomplete Sections.
1. Social Encounters.
Social Encounters are tricky to define. Many interactions are not actually encounters because there are no adversarial elements. As a DM I don't want players just to receive information or money or whatever they want just because they ask, so there is a desire to withhold information or resources but that ends up just drawing out an uninteresting scene. I think this is best fixed by ascertaining goals: if the NPCs' goals are aligned with the PCs, then I can just give them a few pieces of info and do a little bit of in-character chatter and move on - its not an encounter.
The situation becomes an encounter when the NPC has a different goal than the PCs. The NPC wants something.
Generally, when I plan out social encounters, I create a list of NPCs, the NPC's goal, what is actually happening upon the PCs arrival, and what will happen if the PCs decide not to do anything.
My NPCs look like this:
Name, Race
Physical Description
-Goal:
-Principles:
- Information Points
I rarely stat NPCs. If the PCs try to use deception, insight, stealth, or any other PC led skill check, I will generally use a simple DC level rather than provide for an opposed check. I want the focus to be on how good the PCs are at what they do and not on how good the NPCs are at stopping the PCs, so a skill check vs. DC is usually preferable to an opposed check. NPCs are different than enemies in that I did not create them to fight. If the NPC might foreseeably engage in combat, I will stat them out.
Physical Descriptions are usually over-the-top to make the NPCs memorable, but no more than 3 quick things. Players rarely remember names or characteristics unless they are super invested - and then they will remember maybe 2 NPCs and a villain. They will then refer to other NPCs as 'beardless dwarf guy' or 'that bald dude.' That is fine, as long as the Players are able to keep the story straight, I am happy.
Example:
Fists McGee, Dwarf
Bowler hat, thick red mutton chops, uses fists instead of weapons.
- Goal: First McGee wants to get out of his mercenary contract, but is worried about the consequences. Otherwise, he is looking for good drink and maybe a brawl.
- Principles: don't cross someone stronger, all is forgiven the day after a tavern brawl, petty crimes aren't too bad, chaotic good
- He has been hired as a guide for a group of paladins that he really doesn't want to work with.
- The paladins are not 'good' per se, but lawful to the point of tyranny.
- The paladins are hoping to go to the same place that the party is going to, and will invoke a legal charter prohibiting the party's interference if and when they encounter the party.
This is an NPC that I actually used in a game about half a year back. The PCs did not end up asking anything about him what-so-ever, but when they got trapped on a boat together (due to PCs decisions and plans, nothing to do with anything I planned at all), I was able to play him according to these motivations. His past, goals, and principles did not come up at the table but fully informed me of what he would do. In short, it was useful.
1a. NPCs Have Goals. NPCs should want something for some reason; either the want or the reason is related to the story. If the NPC is just there to dump exposition, there is no reason for the NPC. Just tell the players that they heard the information in town or whatever.
Goals inform the DM how the NPC will relate and help or hinder the party, and how the PCs can overcome the NPC through social interactions, skill checks, or otherwise. If the NPC has no goal, then the DM is just left with obstructionist interactions that do nothing but draw out the scene OR the DM creates an exposition dump. Neither are good. Why have a tavern keeper that refuses to give the PCs some vital piece of information unless the PCs hit a certain amount of skill checks? Is it just to draw out time? However, if the NPC has a goal (and principles), the PCs do not necessarily need to hit those skill checks - they can figure out what is bugging the NPC and figure out a way around it. This is the difference between dice rolling and role playing.
The NPC goal should relate to the PCs, the plot, or important NPCs. If the NPC's goal is to open her own shop on the corner one day, selling vegetables, then that does not provide anything interesting for the DM to use and does not inform the DM how the NPC will interact with the PCs. However, if the NPC cannot open the shop due to the thieves guild...... etc.... etc..., there is some good motivations there. Similarly, having a simple goal of staying out of trouble is fine, too, though a bit simplistic.
1b. NPCs Have Principles. D&D provides an eight point axis (and a center) to align PC and NPC principles: lawful evil, chaotic neutral, etc. I am find using that, but I prefer to assemble some fleshed out specific principles that are likely to occur during an encounter, and then summarize it with a point on the access. I think going the other direction creates an atmosphere of black and white behavior. Consider these two questions: 'What will the chaotic evil person do?' versus 'What type of behavior is this?' The first questions leads to stereotypical behavior while the second leads to nuance. By finishing off with the general principle (lawful good, for example), I also have some insight into situations likely to reoccur. A lawful paladin with the principle 'Evil Must Be Annihilated' will respond very differently than a lawful paladin with the principle 'Our Democratic Processes Must Be Upheld,' although both could conceivably have the lawful neutral alignment.
The principles should relate explicitly to goals and likely occurrences. An NPC that won't stab a priest is only informative in so far that a priest is present. If the NPC has a goal of 'amass a fortune', one of those principles should directly relate to what methods are acceptable and which methods are unacceptable.
1c. Information Points are Simple. Most players cannot handle a complex cast of people or a complex plot in a role-playing game. In a novel, tv series, or video game -sure. Keep it simple. When players have it, add more. I say this and I have massively complex plots and factions, but I keep information points simple and build off understanding. For the purpose of building a single encounter, only two or three points should be offered, and in bite-sized chunks. This amount is digestible. Anything more should be introduced later through a subsequent encounter or info drop.
Gaps in information are fine, best used as plot twists. Do not provide all the information bits and assume that the PCs will eventually put it all together. Hidden motivations, hidden agendas, and unseen forces work really well in serialized stories and complete works of fiction because one person has total narrative control. There can be Gaps in information, but you need to point at that Gap over and over to get the players to realize that there is a Gap. The fewer Gaps the better. One good gap at a time provides a decent hook. Too many Gaps and you have an incoherent story.
Consider this simple plot: The evil baron wants the neighboring mine because he is greedy. He is using the mercenaries so that it cannot be traced back to him directly.
Consider this complex plot: the evil baron is actually a devil in disguise. He wants to marry the princess of a neighboring kingdom in order to become the reigning monarch by secretly killing off all the other heirs. The ultimate goal is to create a reigning bloodline of half-devils, and slowly marry the subsequent heirs to other kingdoms so that all the monarchs are of fiendish decent and owe allegiance to hell. In the meantime, he needs to amass wealth to curry favor and has thus decided to attack the mine of a neighboring kingdom. This is something that can be build up to, but takes a lot of subsequent social encounters. It really doesn't work for a sole information dump whatsoever. Start with Mercenaries are attacking the mine, but we do not know who hired them....
Bad: Mercenaries are attacking the mine. Can you stop them?
Good: Mercenaries are attacking the mine, but we do not know who hired them. Can you stop them and find out who is ordering it?
Too Many Gaps: The princess sure is acting strange on her way to visit the Baron - and look! Mercenaries attacking some of the villagers near the mine!
1d. There Is Something Happening If the PCs Do Not Intervene. Inevitably, I will set up a scene between two NPCs, or an NPC doing something on its own, present it to the players, and they will say something like "we watch and hide, eavesdropping in." Then I have to scratch my head because I only had this amount of dialogue in mind and I don't know where the scene was going to go without the PC's intervention.
If there is a social encounter where the PCs interrupt an argument, the PCs might very well just sit back and listen. Be prepared. This is equally true for pre-planned combat encounters: the PCs see a robbery in progress and then inexplicably say 'we chose not to get involved'. Well, there goes your McGuffin, right? Perhaps the PCs will follow, or perhaps the pre-planned combat encounter turns social: the person who lost the item can now convince the PCs to track it down.
Boring, non-engaging scenes rarely pull PCs in. You enter the tavern and there is some guy there; 'we talk to the guy.' They are effective, but non-engaging. In terms of narrative structure, there is a descriptor called 'the inciting event' that causes the hero(es) to go outside of their comfort and engage in a story. Inciting events should NOT be sit-down social encounters where nothing is going on. They should include something else happening that the PCs stumble into.
1e. DMPCs. Don't freak out! I really just mean NPCs that travel with the party, even for a single encounter. Maybe they are being escorted, maybe they are lending on honest hand, maybe they are additional hired muscle. My rule for creating and running DMPCs is actually really simple. They are NPCs with a couple of Levers (special abilities, usually but not always a specific class feature) -- and the PCs are the ones that get to pull the levers. I'm going to get a little into combat encounters as well, here. Another way to look at it: if the PCs have someone walking around with them, be it a princess, a mercenary, or a donkey, that person has stats and is combat ready.
Good DMPCs:
First, they are necessarily weaker and simplier than the PCs. They have maybe 2 skills, designed to give the PCs access to more skills or designed to allow them to work with the PCs, not show them up. For example, Knowledge: Religion and Stealth. Also, they may have one single class feature, and that is it. They are not supposed to overshadow the party, but supplement them.
Second, the DMPC does not do anything without a request to do so by the PCs. They are present, they have HP and an armor class, and they will generally move on their initiative slot to stay with the pack. When the DMPC's turn comes up on initiative, the DM says something like, "The Phoenix Commander steps up behind the PC, saying 'I have your back.'" When the PCs ask for him or her to do something, s/he will to the degree that the request is reasonable. The DMPC does not offer the solution to any problem, but can provide hints when asked. The DMPC is not the DM's avatar, its the PC's resource. The spotlight stays on the PCs.
Third, the DMPC is involved in the plot. The DMPC needs to see something accomplished or needs to get somewhere. They are not just allied and are along for the ride. They are not an extra wheel, but they have some relevance. Once a DMPC is done with the plot, all the loyalty in the world will not make him/her stick around to continue providing whatever stagnant benefits were initially provided.
Some fun things to do with DMPCs:
Mechanical buff: they can provide simple area of affect buffs (10' radius) that make them advantageous for the PCs to position, but make them targets for enemies.
Targets: Have some of your enemies hunt the DMPCs in encounters, honestly and sincerely trying to kill them. That provides for some interesting combat.
1f. Avoid Overused Twists. The friendly NPC....is really a villain! The twist is obvious, and your PCs will see it a mile away. Plot twists are only as effective as they are unforeseen. Good twists make perfect sense upon reflection, everything was pointing to it the entire time. Bad twists are just exposition drops that reinforce played-out tropes. Furthermore, before you throw in one twist, you have to have nine other straight lines. If everything is a twist, nothing is, and you train your PCs to distrust everything you say.
1g. Role Playing, Dice Rolls, and Overcoming Social Interactions. My approach goes like this: 1. Set up in-person dialogue, 2. skill check, 3. act out the results. I let my players engage with the NPCs in character or out of character, whatever they are comfortable with. If the goals and principles are sufficient to keep the NPC from offering the PCs what they are after, they will need to make a roll of the dice no matter how well they perform in character. I then ask them to roleplay it out based on the result of the roll.
A lot of DMs will skip the dice roll if the roleplaying is sufficient to convince. I think that is backwards: lets roleplay it out, but only after we know whether the attempt will be successful. That sounds odd, but its actually REALLY fun for everyone to roleplay out a social failure.
For example - The PCs want to convince the town mayor to post additional guards, because they are worried about an attack.
Player: I hit on the mayor. I flirt with him and then ask him, batting my eyelashes, to pretty please post some additional guards.
DM: alright, roll for deception if your character doesn't find him attractive or persuasion if your character does.
Player: Deception. /roll.
DM: You fail. What do you say that makes your attempt to flirt with him fail?
Player: You probably don't see a lot of pretty girls behind this desk all day, just that hag that mans the desk out front....
DM (picking up on it): Miss, please do not talk about my wife that way.
It seems backwards, but my players enjoy failing social encounters. As a general principle: when a PC fails at something, ask your players, "what did you do to screw it up?" It takes the sting out of the failure and makes the whole thing more enjoyable and immersive while maintaining player agency.
2. Non-Combat Non-Social Encounters.
Whether the encounters are traps, wilderness exploration, or puzzles, encounters need more than skill checks. They need player interaction. This is different from PC interactions. When a PC climbs a cliff face, its a PC interaction but its a simple Player dice roll. The Player is not engaged. If the encounter can be solved by simple dice rolls, its boring. Skill rolls can be used, but the Players must be engaged on some level above selecting the skill to use. These should not be games of Yahtzee.
Social Encounters provide roleplay interactivity and Combat Encounters provide map and maneuver interactivity - both can be dice heavy, but both also provide meaningful ways for the players to engage, as well as the PCs that they are playing. Non-Combat / Non-Social Encounters require something to make them similarly interactive.
When the scene is set, there needs to be meaningful props, details, and some kind of challenge. Without the challenge, there is no encounter, just scenery. An empty room with some papers on a desk that contain exposition is not an encounter at all, but an exposition dump. If the papers are hidden in a locked door, it becomes a boring encounter.
2a. Interesting Nonfunctional Things are Also Present. This is the opposite of Chekhov's Gun. The best way (imo) to create a sense of the world is to add an element to each encounter that has no functionality other than to describe the world or provide ambiance and provides some limited interaction. Enterprising characters will still tend to use the element, so it also provides some tools for out-of-the-box thinking.
Without nonfunctional things, you ARE stuck with Chekhov's Gun: everything present has a reason to be present. The corridor ends unexpectedly and there is a gargoyle statue set into the wall. Everyone knows to bee-line to the Gargoyle. If every detail has a 1:1 correlation to functionality, you end up with a point+click game where the protagonists gather every item because there is a known use for every item, and they interact with every mentioned detail. This also has a tendency to discourage out of the box thinking. Players will only interact with what ever is present, knowing that each feature is a key of some sort to move the story forwards.
Nonfunctional details are different than providing additional backstory. The goal is to create ambiance and give a sense of the world to the players. Backstory is fine, but it will cause the Players to think that they need to remember these details, whereas ambiance creates a mood. There are crude gnomish glyphs and paintings that are nothing more than ancient graffiti. There are a bed of roses that quietly bark at the players and sniff them. That sort of thing. There is always the danger that players will latch onto some ambient element as being important, which a DM can make important in a flash.
The Fallout series is both good and bad at this. A lot of times they just provide ancient skeletons in compromising or hilarious positions. That is more amusing than what I go for. At other times, there are posters and computer data entries in factories that give a sense of the world and provide no additional plot - just ambiance. That is great, and gives the series a distinct feel.
This is also more than just describing a room. For Example:
Bad: the inn has several tables and with stains, a grumpy looking man working the bar, and several patrons drinking and talking loudly.
Good: the inn has several tables and with stains, a grumpy looking man working the bar, and several patrons drinking and talking loudly. You notice someone carved a depiction of a dwarf in the wall next to a table. The dwarf carving is crude, and the words "I eat roks" is carved into the wall next to him.
2b. Important Intractable Things Have Sign Posts. The PCs should be informed what is in the room that grabs their attention without the PC asking 'what do I see.' I am never at a loss here because I create my own maps on roll20, but I have interacted with other DMs that purchase beautiful maps that are highly detailed. When PCs enter a room they start asking about everything present on the map and the DM starts making things up because there was nothing planned. The players do not know what is there intentionally, what is important, and what they should be looking for. I repeatedly will inform players what is present in the room just to remind them after they finish up interacting with one thing.
Even hidden things have sign-posts, but those hidden things have clever sign-posts. If the players walk down an empty hallway getting from one room to another room and miss a hidden door because the hallway was described by the GM as "a hallway", then that secret door was poorly designed. There is a reason that I put this part right after section 2a, because Interesting Nonfunctional Things provide the perfect cover for hidden things. Whenever my players want to search for something, I require a skill check, even if nothing is there. A success will often conclude with 'you are certain there are no traps' or 'you are certain that there are no doors.' I don't want to train my players into thinking that a skill check means that there is something to discover by not requiring a skill check when there is nothing to discover. That being said, I will not require a check when I have not put a sign-post down. That empty hallway that I describe as "a hallway" warrants no check - you automatically know that there is nothing there. If there is no sign post, then there is nothing there. I will train my players by never requiring a check to confirm that there is nothing there.
Good Secret Door Signpost: "You enter a large rectangular room with several circular tables surrounded by tables. A long, narrow carpet runs the length of the room. A large tapestry covers the west wall, depicting dwarves mining deep underground, with a large gem in the very center or the picture. There is a cupboard along the east wall, with a countertop."
Bad Secret Door Signpost: "You enter a large rectangular drinking hall, decorated for an upscale audience."
2c. Traps Are Icing. Put Them On a Cake. Traps are annoying on their own for both the DM and the players. When Players are aware of traps, they search EVERY room ALL THE TIME. Traps are better combined with other game elements, like hidden doors or combat. Really cool traps are still effectively boring when located by a perception check and then disabled, no matter how elaborate the trap. Elaborate traps that go off only delight the DM. No one else cares, and just get frustrated that they failed a save-or-suck effect out of nowhere.
I like traps that are not necessarily traps. Water chutes, avalanches, and giant venus fly traps are great and readily identifiable as potential dangers. Once the PCs are aware of the danger, interesting things can happen.
2d. There Should Be a Right Answer - And Other Right Answers. The game should not be 'guess what the DM is thinking' and require players to do something specific to pass through it. This is one of the reasons I hate puzzles: if you do not solve the puzzle, you are stopped. On the other hand, the DM should have thought of at least one way for the party to reasonably get through the obstacle. Ideally, there will be several different ways to overcome an obstacle and at least one other feature present that an enterprising party could use.
3. Combat Encounters.
On the adventure planning side, combat features and interactions should get more complex as the dungeoneers get further into the dungeon, but early encounters should be simplier and build on player understanding. There will always be a 'starting location' in any strings of encounters, and that starting encounter should be fairly simple.
Some of this should be written out, but most of it doesn't need to be:
Calculated Weighted XP Total
Enemy Unit + Number + XP
Goal:
Special: <if applicable: including principles, weird things happening, specific set ups, whatever>
Terrain:
I think writing and planning this out is appropriate prior to drawing out the map, but sometimes I am not quite inspired until I have the general idea of the terrain, so I will draw out the map and then get to the type of enemies that would be interesting there. Calculated Weighted XP Total includes the modifier found in the DMG regarding number of enemies, so if I have, say 2 enemies I will apply the 1.5x modifier. I don't list the exact difficulty per the party composition because I typically create multiple places for my players to visit. If they go some place that I created a while ago, I don't change it or upgrade it or anything. If the PCs outgrown the location significantly without ever going over there, I will eliminate that location as a possible quest area.
Special is useful if there is a specific set-up, like the enemies are leading a group of captives away, specifically targeting a DMPC, or if the encounter is meant to be an ambush. 'Principles' are only useful if the enemy is particularly principled or will not act in very certain ways. The vast majority of the time, its not going to matter.
Example:
2,000 xp
-Necromancer x2 (450 xp each)
-Skeleton x2 (50 xp)
Goal: The Necromancers are killing townsfolk and raising them back to life as skeleton warriors to create a powerful undead army. They will view the PCs are potentially powerful undead.
Terrain: a Farmstead with several layers of waste-high fences, and only a couple of narrow gates.
3a. Enemies Should Have Goals. Goals keep the encounter from being a superficial pretext for murderhoboism. If you are running a straight dungeon crawl, I could see this being less important.
Why are thugs in the back allies of a large town targeting a group of extremely well armed people? The enemies should have a reason to be engaging in violence beyond the idea that the players are getting bored of exploring the marketplace. Beasts might be hunting. Armies might be attacking. Some enemy goals provide a way to avoid combat, and some goals will provoke specific in-combat behavior. Either way, the goal should address what the enemies are doing absent PC interaction as well as how they will interact against the PCs specifically. If the enemies are specifically targeting the PCs, that provides both aspects of the goals: they are hunting the PCs AND once they find the PCs they will attack the PCs.
Bandits on the road that are willing to fight the PCs to the death provide a poor and non-immersive experience. Bandits that attack, realize their mistake and try to flee can provide an interesting experience - especially if they withdrawal immediately, provoking the players to pursue. Orcs that are attacking a town and taking captives that immediately withdraw upon seeing an armed response make for an interesting encounter, as now the PCs have to engage and prevent the Orcs from fleeing or otherwise pursue.
Enemy goals should usually relate to the plot, but well-placed combat that includes enemies whose goals have nothing to do with the PCs can give a sense of the larger world. For example, suppose that the PCs need to sneak into a fortress and choose to go through an underground crypt filled with Oozes. The Oozes want to feed. Throw in some patrols from the fortress above and you have some interesting encounters - some of which might be best avoided. Maybe the PCs can lure an Ooze to attack the guards. Maybe the PCs will get blocked in by PCs and be forced to fight on two fronts because Oozes begin coming up from behind. Either way, the Oozes' behavior is goal oriented: the Ooze wants to feed and it doesn't much care whether it attacks a PC or the guards.
3b. Combat Encounters Should Be Avoidable - But Not Always. The PCs should not be forced to engage in combat every time you plan a combat. Some encounters should have the possibility of an alternative solution at the outset. Some Combat Encounters should start as social encounters, or could be planned as combat encounters until clever PCs turn it into a social encounter. Or maybe the PCs work to avoid the encounter through sneaking and trickery. If enemy goals are fleshed out, the ability of the party to bypass an encounter socially becomes easier to determine.
Other encounters should never be able to be avoided no matter how socially adept the PCs are. At best, the PCs can flee. If the PCs can always sneak past the guards using the Warlock's invocation 'Mask of Many Faces', things get stale quickly. It should work sometimes. It should even work reliably. The moment one approach works all the time, the game gets stale.
Some enemies will not listen to reason, no matter how charming, ferocious, or lying the PCs are. This goes for more than just beasts and abominations. On a gameflow note, any encounters that still expend resources do not feel like a waste to me. If my players want to expend spells to bypass combat in clever ways, everyone wins.
3c. Terrain Features Should Be Everywhere. Terrain features should be included in every combat in a meaningful way. The biggest lack that I see in games that I participate in as a player is a lack of meaningful terrain. 99% of all maps should have 'difficult terrain' to start with. 100% should have interesting terrain features on top of it. Interesting can be nothing more than several tables that a guard can kick over for cover, or can be extreme as a spinning buzzsaw of death that circles through the room. Traps make great terrain features.
It is okay to have features favoring the enemies, but not every time. Some encounters should favor the PCs, some should favor the enemies, some should be equally disfavoring both. If all terrain features favor the enemies, it gets boring because terrain becomes something to overcome rather than utilize. It is understandable if a group of enemies takes up an advantageous position if they are expecting an attack, bivouacking and what-not. But not all enemies will think it through, and not all enemies will have the opportunity to do so. Encounters located behind enemies lines can be a great example of enemies taking disadvantageous positions - they weren't expecting trouble!
A List of Interesting Terrain Features to Include:
-Obvious Traps, Traps of which the enemies are aware, Traps of which the enemies are not aware, Traps the PCs can activate prior to an imminent combat
-Knee Deep Water, Very Deep Water, Flowing Water, etc.; in order to be interesting, it needs to be obstructing, not just along the side or nearby
-Multiple Accessible Elevations, linked by stairs, ramps, ladders, springboards, scaleable walls
-Foliage: covering, thorny, magical, view-obstructing
-Fences, half walls, overturned tables, fallen pillars
-Furniture, throw-able furniture, on-fire throw-able furniture
-Fires, camp fires, torches, oil lamps, unlit knocked over oil lamps
-Stacks of paper, stacks of paper reaching up to a person's chin, torrential whirlwinds of air spinning around loose pages of paper
-Teleporting mirrors whose other side is within eye range
-Quicksand, mud, wet cement (the Romans used an early version!)
-Portcullises (Protculli? Portculles?) and Doors in the middle of the combat space, not just outlining it as dividers between rooms
-Snowstorms, Heavy Rains, Large Awnings, and Ice
-Etc.
3d. Map Sizes and Room Layout Should Vary. Usually maps will be on the 40' to 80' scale. On the largest end of the spectrum, say 350' across or more. While that is far more room than needed, it provides room enough for the encounter to move as it progresses (such as when one faction tries to flee, or when the players try to bate the enemy into going into a trap). Also, there are some spells with that kind of massive range. It is very impractical in a physical map - but I do my maps on roll20 and that is no problem.
3e. Fudging It. Just a Little. Prior to the start of the combat, I will generally reduce the HP proportionally for the enemies that the party is facing if fewer than the expected total of players showed up. If I design for 4-5 players and I have 3 show up, the enemies have 3/4 hp, and effects keyed off of hp totals are likewise reduced.
I borrow [steal?] heavily from the Angry GM, but I wanted to customize the approach further for my own design method. His approach has been very abstract, which I appreciate, but I wanted more specifics in my own game. This is my approach, but I am open to further suggestions or approaches and I hope this likewise gives others insight if they are still trying to figure out things.
Version 1.1: Rewording. Completing Incomplete Sections.
1. Social Encounters.
Social Encounters are tricky to define. Many interactions are not actually encounters because there are no adversarial elements. As a DM I don't want players just to receive information or money or whatever they want just because they ask, so there is a desire to withhold information or resources but that ends up just drawing out an uninteresting scene. I think this is best fixed by ascertaining goals: if the NPCs' goals are aligned with the PCs, then I can just give them a few pieces of info and do a little bit of in-character chatter and move on - its not an encounter.
The situation becomes an encounter when the NPC has a different goal than the PCs. The NPC wants something.
Generally, when I plan out social encounters, I create a list of NPCs, the NPC's goal, what is actually happening upon the PCs arrival, and what will happen if the PCs decide not to do anything.
My NPCs look like this:
Name, Race
Physical Description
-Goal:
-Principles:
- Information Points
I rarely stat NPCs. If the PCs try to use deception, insight, stealth, or any other PC led skill check, I will generally use a simple DC level rather than provide for an opposed check. I want the focus to be on how good the PCs are at what they do and not on how good the NPCs are at stopping the PCs, so a skill check vs. DC is usually preferable to an opposed check. NPCs are different than enemies in that I did not create them to fight. If the NPC might foreseeably engage in combat, I will stat them out.
Physical Descriptions are usually over-the-top to make the NPCs memorable, but no more than 3 quick things. Players rarely remember names or characteristics unless they are super invested - and then they will remember maybe 2 NPCs and a villain. They will then refer to other NPCs as 'beardless dwarf guy' or 'that bald dude.' That is fine, as long as the Players are able to keep the story straight, I am happy.
Example:
Fists McGee, Dwarf
Bowler hat, thick red mutton chops, uses fists instead of weapons.
- Goal: First McGee wants to get out of his mercenary contract, but is worried about the consequences. Otherwise, he is looking for good drink and maybe a brawl.
- Principles: don't cross someone stronger, all is forgiven the day after a tavern brawl, petty crimes aren't too bad, chaotic good
- He has been hired as a guide for a group of paladins that he really doesn't want to work with.
- The paladins are not 'good' per se, but lawful to the point of tyranny.
- The paladins are hoping to go to the same place that the party is going to, and will invoke a legal charter prohibiting the party's interference if and when they encounter the party.
This is an NPC that I actually used in a game about half a year back. The PCs did not end up asking anything about him what-so-ever, but when they got trapped on a boat together (due to PCs decisions and plans, nothing to do with anything I planned at all), I was able to play him according to these motivations. His past, goals, and principles did not come up at the table but fully informed me of what he would do. In short, it was useful.
1a. NPCs Have Goals. NPCs should want something for some reason; either the want or the reason is related to the story. If the NPC is just there to dump exposition, there is no reason for the NPC. Just tell the players that they heard the information in town or whatever.
Goals inform the DM how the NPC will relate and help or hinder the party, and how the PCs can overcome the NPC through social interactions, skill checks, or otherwise. If the NPC has no goal, then the DM is just left with obstructionist interactions that do nothing but draw out the scene OR the DM creates an exposition dump. Neither are good. Why have a tavern keeper that refuses to give the PCs some vital piece of information unless the PCs hit a certain amount of skill checks? Is it just to draw out time? However, if the NPC has a goal (and principles), the PCs do not necessarily need to hit those skill checks - they can figure out what is bugging the NPC and figure out a way around it. This is the difference between dice rolling and role playing.
The NPC goal should relate to the PCs, the plot, or important NPCs. If the NPC's goal is to open her own shop on the corner one day, selling vegetables, then that does not provide anything interesting for the DM to use and does not inform the DM how the NPC will interact with the PCs. However, if the NPC cannot open the shop due to the thieves guild...... etc.... etc..., there is some good motivations there. Similarly, having a simple goal of staying out of trouble is fine, too, though a bit simplistic.
1b. NPCs Have Principles. D&D provides an eight point axis (and a center) to align PC and NPC principles: lawful evil, chaotic neutral, etc. I am find using that, but I prefer to assemble some fleshed out specific principles that are likely to occur during an encounter, and then summarize it with a point on the access. I think going the other direction creates an atmosphere of black and white behavior. Consider these two questions: 'What will the chaotic evil person do?' versus 'What type of behavior is this?' The first questions leads to stereotypical behavior while the second leads to nuance. By finishing off with the general principle (lawful good, for example), I also have some insight into situations likely to reoccur. A lawful paladin with the principle 'Evil Must Be Annihilated' will respond very differently than a lawful paladin with the principle 'Our Democratic Processes Must Be Upheld,' although both could conceivably have the lawful neutral alignment.
The principles should relate explicitly to goals and likely occurrences. An NPC that won't stab a priest is only informative in so far that a priest is present. If the NPC has a goal of 'amass a fortune', one of those principles should directly relate to what methods are acceptable and which methods are unacceptable.
1c. Information Points are Simple. Most players cannot handle a complex cast of people or a complex plot in a role-playing game. In a novel, tv series, or video game -sure. Keep it simple. When players have it, add more. I say this and I have massively complex plots and factions, but I keep information points simple and build off understanding. For the purpose of building a single encounter, only two or three points should be offered, and in bite-sized chunks. This amount is digestible. Anything more should be introduced later through a subsequent encounter or info drop.
Gaps in information are fine, best used as plot twists. Do not provide all the information bits and assume that the PCs will eventually put it all together. Hidden motivations, hidden agendas, and unseen forces work really well in serialized stories and complete works of fiction because one person has total narrative control. There can be Gaps in information, but you need to point at that Gap over and over to get the players to realize that there is a Gap. The fewer Gaps the better. One good gap at a time provides a decent hook. Too many Gaps and you have an incoherent story.
Consider this simple plot: The evil baron wants the neighboring mine because he is greedy. He is using the mercenaries so that it cannot be traced back to him directly.
Consider this complex plot: the evil baron is actually a devil in disguise. He wants to marry the princess of a neighboring kingdom in order to become the reigning monarch by secretly killing off all the other heirs. The ultimate goal is to create a reigning bloodline of half-devils, and slowly marry the subsequent heirs to other kingdoms so that all the monarchs are of fiendish decent and owe allegiance to hell. In the meantime, he needs to amass wealth to curry favor and has thus decided to attack the mine of a neighboring kingdom. This is something that can be build up to, but takes a lot of subsequent social encounters. It really doesn't work for a sole information dump whatsoever. Start with Mercenaries are attacking the mine, but we do not know who hired them....
Bad: Mercenaries are attacking the mine. Can you stop them?
Good: Mercenaries are attacking the mine, but we do not know who hired them. Can you stop them and find out who is ordering it?
Too Many Gaps: The princess sure is acting strange on her way to visit the Baron - and look! Mercenaries attacking some of the villagers near the mine!
1d. There Is Something Happening If the PCs Do Not Intervene. Inevitably, I will set up a scene between two NPCs, or an NPC doing something on its own, present it to the players, and they will say something like "we watch and hide, eavesdropping in." Then I have to scratch my head because I only had this amount of dialogue in mind and I don't know where the scene was going to go without the PC's intervention.
If there is a social encounter where the PCs interrupt an argument, the PCs might very well just sit back and listen. Be prepared. This is equally true for pre-planned combat encounters: the PCs see a robbery in progress and then inexplicably say 'we chose not to get involved'. Well, there goes your McGuffin, right? Perhaps the PCs will follow, or perhaps the pre-planned combat encounter turns social: the person who lost the item can now convince the PCs to track it down.
Boring, non-engaging scenes rarely pull PCs in. You enter the tavern and there is some guy there; 'we talk to the guy.' They are effective, but non-engaging. In terms of narrative structure, there is a descriptor called 'the inciting event' that causes the hero(es) to go outside of their comfort and engage in a story. Inciting events should NOT be sit-down social encounters where nothing is going on. They should include something else happening that the PCs stumble into.
1e. DMPCs. Don't freak out! I really just mean NPCs that travel with the party, even for a single encounter. Maybe they are being escorted, maybe they are lending on honest hand, maybe they are additional hired muscle. My rule for creating and running DMPCs is actually really simple. They are NPCs with a couple of Levers (special abilities, usually but not always a specific class feature) -- and the PCs are the ones that get to pull the levers. I'm going to get a little into combat encounters as well, here. Another way to look at it: if the PCs have someone walking around with them, be it a princess, a mercenary, or a donkey, that person has stats and is combat ready.
Good DMPCs:
First, they are necessarily weaker and simplier than the PCs. They have maybe 2 skills, designed to give the PCs access to more skills or designed to allow them to work with the PCs, not show them up. For example, Knowledge: Religion and Stealth. Also, they may have one single class feature, and that is it. They are not supposed to overshadow the party, but supplement them.
Second, the DMPC does not do anything without a request to do so by the PCs. They are present, they have HP and an armor class, and they will generally move on their initiative slot to stay with the pack. When the DMPC's turn comes up on initiative, the DM says something like, "The Phoenix Commander steps up behind the PC, saying 'I have your back.'" When the PCs ask for him or her to do something, s/he will to the degree that the request is reasonable. The DMPC does not offer the solution to any problem, but can provide hints when asked. The DMPC is not the DM's avatar, its the PC's resource. The spotlight stays on the PCs.
Third, the DMPC is involved in the plot. The DMPC needs to see something accomplished or needs to get somewhere. They are not just allied and are along for the ride. They are not an extra wheel, but they have some relevance. Once a DMPC is done with the plot, all the loyalty in the world will not make him/her stick around to continue providing whatever stagnant benefits were initially provided.
Some fun things to do with DMPCs:
Mechanical buff: they can provide simple area of affect buffs (10' radius) that make them advantageous for the PCs to position, but make them targets for enemies.
Targets: Have some of your enemies hunt the DMPCs in encounters, honestly and sincerely trying to kill them. That provides for some interesting combat.
1f. Avoid Overused Twists. The friendly NPC....is really a villain! The twist is obvious, and your PCs will see it a mile away. Plot twists are only as effective as they are unforeseen. Good twists make perfect sense upon reflection, everything was pointing to it the entire time. Bad twists are just exposition drops that reinforce played-out tropes. Furthermore, before you throw in one twist, you have to have nine other straight lines. If everything is a twist, nothing is, and you train your PCs to distrust everything you say.
1g. Role Playing, Dice Rolls, and Overcoming Social Interactions. My approach goes like this: 1. Set up in-person dialogue, 2. skill check, 3. act out the results. I let my players engage with the NPCs in character or out of character, whatever they are comfortable with. If the goals and principles are sufficient to keep the NPC from offering the PCs what they are after, they will need to make a roll of the dice no matter how well they perform in character. I then ask them to roleplay it out based on the result of the roll.
A lot of DMs will skip the dice roll if the roleplaying is sufficient to convince. I think that is backwards: lets roleplay it out, but only after we know whether the attempt will be successful. That sounds odd, but its actually REALLY fun for everyone to roleplay out a social failure.
For example - The PCs want to convince the town mayor to post additional guards, because they are worried about an attack.
Player: I hit on the mayor. I flirt with him and then ask him, batting my eyelashes, to pretty please post some additional guards.
DM: alright, roll for deception if your character doesn't find him attractive or persuasion if your character does.
Player: Deception. /roll.
DM: You fail. What do you say that makes your attempt to flirt with him fail?
Player: You probably don't see a lot of pretty girls behind this desk all day, just that hag that mans the desk out front....
DM (picking up on it): Miss, please do not talk about my wife that way.
It seems backwards, but my players enjoy failing social encounters. As a general principle: when a PC fails at something, ask your players, "what did you do to screw it up?" It takes the sting out of the failure and makes the whole thing more enjoyable and immersive while maintaining player agency.
2. Non-Combat Non-Social Encounters.
Whether the encounters are traps, wilderness exploration, or puzzles, encounters need more than skill checks. They need player interaction. This is different from PC interactions. When a PC climbs a cliff face, its a PC interaction but its a simple Player dice roll. The Player is not engaged. If the encounter can be solved by simple dice rolls, its boring. Skill rolls can be used, but the Players must be engaged on some level above selecting the skill to use. These should not be games of Yahtzee.
Social Encounters provide roleplay interactivity and Combat Encounters provide map and maneuver interactivity - both can be dice heavy, but both also provide meaningful ways for the players to engage, as well as the PCs that they are playing. Non-Combat / Non-Social Encounters require something to make them similarly interactive.
When the scene is set, there needs to be meaningful props, details, and some kind of challenge. Without the challenge, there is no encounter, just scenery. An empty room with some papers on a desk that contain exposition is not an encounter at all, but an exposition dump. If the papers are hidden in a locked door, it becomes a boring encounter.
2a. Interesting Nonfunctional Things are Also Present. This is the opposite of Chekhov's Gun. The best way (imo) to create a sense of the world is to add an element to each encounter that has no functionality other than to describe the world or provide ambiance and provides some limited interaction. Enterprising characters will still tend to use the element, so it also provides some tools for out-of-the-box thinking.
Without nonfunctional things, you ARE stuck with Chekhov's Gun: everything present has a reason to be present. The corridor ends unexpectedly and there is a gargoyle statue set into the wall. Everyone knows to bee-line to the Gargoyle. If every detail has a 1:1 correlation to functionality, you end up with a point+click game where the protagonists gather every item because there is a known use for every item, and they interact with every mentioned detail. This also has a tendency to discourage out of the box thinking. Players will only interact with what ever is present, knowing that each feature is a key of some sort to move the story forwards.
Nonfunctional details are different than providing additional backstory. The goal is to create ambiance and give a sense of the world to the players. Backstory is fine, but it will cause the Players to think that they need to remember these details, whereas ambiance creates a mood. There are crude gnomish glyphs and paintings that are nothing more than ancient graffiti. There are a bed of roses that quietly bark at the players and sniff them. That sort of thing. There is always the danger that players will latch onto some ambient element as being important, which a DM can make important in a flash.
The Fallout series is both good and bad at this. A lot of times they just provide ancient skeletons in compromising or hilarious positions. That is more amusing than what I go for. At other times, there are posters and computer data entries in factories that give a sense of the world and provide no additional plot - just ambiance. That is great, and gives the series a distinct feel.
This is also more than just describing a room. For Example:
Bad: the inn has several tables and with stains, a grumpy looking man working the bar, and several patrons drinking and talking loudly.
Good: the inn has several tables and with stains, a grumpy looking man working the bar, and several patrons drinking and talking loudly. You notice someone carved a depiction of a dwarf in the wall next to a table. The dwarf carving is crude, and the words "I eat roks" is carved into the wall next to him.
2b. Important Intractable Things Have Sign Posts. The PCs should be informed what is in the room that grabs their attention without the PC asking 'what do I see.' I am never at a loss here because I create my own maps on roll20, but I have interacted with other DMs that purchase beautiful maps that are highly detailed. When PCs enter a room they start asking about everything present on the map and the DM starts making things up because there was nothing planned. The players do not know what is there intentionally, what is important, and what they should be looking for. I repeatedly will inform players what is present in the room just to remind them after they finish up interacting with one thing.
Even hidden things have sign-posts, but those hidden things have clever sign-posts. If the players walk down an empty hallway getting from one room to another room and miss a hidden door because the hallway was described by the GM as "a hallway", then that secret door was poorly designed. There is a reason that I put this part right after section 2a, because Interesting Nonfunctional Things provide the perfect cover for hidden things. Whenever my players want to search for something, I require a skill check, even if nothing is there. A success will often conclude with 'you are certain there are no traps' or 'you are certain that there are no doors.' I don't want to train my players into thinking that a skill check means that there is something to discover by not requiring a skill check when there is nothing to discover. That being said, I will not require a check when I have not put a sign-post down. That empty hallway that I describe as "a hallway" warrants no check - you automatically know that there is nothing there. If there is no sign post, then there is nothing there. I will train my players by never requiring a check to confirm that there is nothing there.
Good Secret Door Signpost: "You enter a large rectangular room with several circular tables surrounded by tables. A long, narrow carpet runs the length of the room. A large tapestry covers the west wall, depicting dwarves mining deep underground, with a large gem in the very center or the picture. There is a cupboard along the east wall, with a countertop."
Bad Secret Door Signpost: "You enter a large rectangular drinking hall, decorated for an upscale audience."
2c. Traps Are Icing. Put Them On a Cake. Traps are annoying on their own for both the DM and the players. When Players are aware of traps, they search EVERY room ALL THE TIME. Traps are better combined with other game elements, like hidden doors or combat. Really cool traps are still effectively boring when located by a perception check and then disabled, no matter how elaborate the trap. Elaborate traps that go off only delight the DM. No one else cares, and just get frustrated that they failed a save-or-suck effect out of nowhere.
I like traps that are not necessarily traps. Water chutes, avalanches, and giant venus fly traps are great and readily identifiable as potential dangers. Once the PCs are aware of the danger, interesting things can happen.
2d. There Should Be a Right Answer - And Other Right Answers. The game should not be 'guess what the DM is thinking' and require players to do something specific to pass through it. This is one of the reasons I hate puzzles: if you do not solve the puzzle, you are stopped. On the other hand, the DM should have thought of at least one way for the party to reasonably get through the obstacle. Ideally, there will be several different ways to overcome an obstacle and at least one other feature present that an enterprising party could use.
3. Combat Encounters.
On the adventure planning side, combat features and interactions should get more complex as the dungeoneers get further into the dungeon, but early encounters should be simplier and build on player understanding. There will always be a 'starting location' in any strings of encounters, and that starting encounter should be fairly simple.
Some of this should be written out, but most of it doesn't need to be:
Calculated Weighted XP Total
Enemy Unit + Number + XP
Goal:
Special: <if applicable: including principles, weird things happening, specific set ups, whatever>
Terrain:
I think writing and planning this out is appropriate prior to drawing out the map, but sometimes I am not quite inspired until I have the general idea of the terrain, so I will draw out the map and then get to the type of enemies that would be interesting there. Calculated Weighted XP Total includes the modifier found in the DMG regarding number of enemies, so if I have, say 2 enemies I will apply the 1.5x modifier. I don't list the exact difficulty per the party composition because I typically create multiple places for my players to visit. If they go some place that I created a while ago, I don't change it or upgrade it or anything. If the PCs outgrown the location significantly without ever going over there, I will eliminate that location as a possible quest area.
Special is useful if there is a specific set-up, like the enemies are leading a group of captives away, specifically targeting a DMPC, or if the encounter is meant to be an ambush. 'Principles' are only useful if the enemy is particularly principled or will not act in very certain ways. The vast majority of the time, its not going to matter.
Example:
2,000 xp
-Necromancer x2 (450 xp each)
-Skeleton x2 (50 xp)
Goal: The Necromancers are killing townsfolk and raising them back to life as skeleton warriors to create a powerful undead army. They will view the PCs are potentially powerful undead.
Terrain: a Farmstead with several layers of waste-high fences, and only a couple of narrow gates.
3a. Enemies Should Have Goals. Goals keep the encounter from being a superficial pretext for murderhoboism. If you are running a straight dungeon crawl, I could see this being less important.
Why are thugs in the back allies of a large town targeting a group of extremely well armed people? The enemies should have a reason to be engaging in violence beyond the idea that the players are getting bored of exploring the marketplace. Beasts might be hunting. Armies might be attacking. Some enemy goals provide a way to avoid combat, and some goals will provoke specific in-combat behavior. Either way, the goal should address what the enemies are doing absent PC interaction as well as how they will interact against the PCs specifically. If the enemies are specifically targeting the PCs, that provides both aspects of the goals: they are hunting the PCs AND once they find the PCs they will attack the PCs.
Bandits on the road that are willing to fight the PCs to the death provide a poor and non-immersive experience. Bandits that attack, realize their mistake and try to flee can provide an interesting experience - especially if they withdrawal immediately, provoking the players to pursue. Orcs that are attacking a town and taking captives that immediately withdraw upon seeing an armed response make for an interesting encounter, as now the PCs have to engage and prevent the Orcs from fleeing or otherwise pursue.
Enemy goals should usually relate to the plot, but well-placed combat that includes enemies whose goals have nothing to do with the PCs can give a sense of the larger world. For example, suppose that the PCs need to sneak into a fortress and choose to go through an underground crypt filled with Oozes. The Oozes want to feed. Throw in some patrols from the fortress above and you have some interesting encounters - some of which might be best avoided. Maybe the PCs can lure an Ooze to attack the guards. Maybe the PCs will get blocked in by PCs and be forced to fight on two fronts because Oozes begin coming up from behind. Either way, the Oozes' behavior is goal oriented: the Ooze wants to feed and it doesn't much care whether it attacks a PC or the guards.
3b. Combat Encounters Should Be Avoidable - But Not Always. The PCs should not be forced to engage in combat every time you plan a combat. Some encounters should have the possibility of an alternative solution at the outset. Some Combat Encounters should start as social encounters, or could be planned as combat encounters until clever PCs turn it into a social encounter. Or maybe the PCs work to avoid the encounter through sneaking and trickery. If enemy goals are fleshed out, the ability of the party to bypass an encounter socially becomes easier to determine.
Other encounters should never be able to be avoided no matter how socially adept the PCs are. At best, the PCs can flee. If the PCs can always sneak past the guards using the Warlock's invocation 'Mask of Many Faces', things get stale quickly. It should work sometimes. It should even work reliably. The moment one approach works all the time, the game gets stale.
Some enemies will not listen to reason, no matter how charming, ferocious, or lying the PCs are. This goes for more than just beasts and abominations. On a gameflow note, any encounters that still expend resources do not feel like a waste to me. If my players want to expend spells to bypass combat in clever ways, everyone wins.
3c. Terrain Features Should Be Everywhere. Terrain features should be included in every combat in a meaningful way. The biggest lack that I see in games that I participate in as a player is a lack of meaningful terrain. 99% of all maps should have 'difficult terrain' to start with. 100% should have interesting terrain features on top of it. Interesting can be nothing more than several tables that a guard can kick over for cover, or can be extreme as a spinning buzzsaw of death that circles through the room. Traps make great terrain features.
It is okay to have features favoring the enemies, but not every time. Some encounters should favor the PCs, some should favor the enemies, some should be equally disfavoring both. If all terrain features favor the enemies, it gets boring because terrain becomes something to overcome rather than utilize. It is understandable if a group of enemies takes up an advantageous position if they are expecting an attack, bivouacking and what-not. But not all enemies will think it through, and not all enemies will have the opportunity to do so. Encounters located behind enemies lines can be a great example of enemies taking disadvantageous positions - they weren't expecting trouble!
A List of Interesting Terrain Features to Include:
-Obvious Traps, Traps of which the enemies are aware, Traps of which the enemies are not aware, Traps the PCs can activate prior to an imminent combat
-Knee Deep Water, Very Deep Water, Flowing Water, etc.; in order to be interesting, it needs to be obstructing, not just along the side or nearby
-Multiple Accessible Elevations, linked by stairs, ramps, ladders, springboards, scaleable walls
-Foliage: covering, thorny, magical, view-obstructing
-Fences, half walls, overturned tables, fallen pillars
-Furniture, throw-able furniture, on-fire throw-able furniture
-Fires, camp fires, torches, oil lamps, unlit knocked over oil lamps
-Stacks of paper, stacks of paper reaching up to a person's chin, torrential whirlwinds of air spinning around loose pages of paper
-Teleporting mirrors whose other side is within eye range
-Quicksand, mud, wet cement (the Romans used an early version!)
-Portcullises (Protculli? Portculles?) and Doors in the middle of the combat space, not just outlining it as dividers between rooms
-Snowstorms, Heavy Rains, Large Awnings, and Ice
-Etc.
3d. Map Sizes and Room Layout Should Vary. Usually maps will be on the 40' to 80' scale. On the largest end of the spectrum, say 350' across or more. While that is far more room than needed, it provides room enough for the encounter to move as it progresses (such as when one faction tries to flee, or when the players try to bate the enemy into going into a trap). Also, there are some spells with that kind of massive range. It is very impractical in a physical map - but I do my maps on roll20 and that is no problem.
3e. Fudging It. Just a Little. Prior to the start of the combat, I will generally reduce the HP proportionally for the enemies that the party is facing if fewer than the expected total of players showed up. If I design for 4-5 players and I have 3 show up, the enemies have 3/4 hp, and effects keyed off of hp totals are likewise reduced.
I borrow [steal?] heavily from the Angry GM, but I wanted to customize the approach further for my own design method. His approach has been very abstract, which I appreciate, but I wanted more specifics in my own game. This is my approach, but I am open to further suggestions or approaches and I hope this likewise gives others insight if they are still trying to figure out things.