Sandboxing Part 2 - Sandboxing for New DMs
I was asked about how this all works for beginners, it's great to recommend sandboxing to an experienced DM, but what about a newer one. I mentioned in my first post that in my own sandbox game:
1. The DM was very familiar with the rule set
2. The DM was very familiar with the setting
So this was challenging to new DM’s.
In actuality, this isn’t the case, but it will take a bit of explanation to show why.
Familiarity with the rule set
This is the least important of the two conditions, familiarity with the rule set makes it easier, but it is entirely possible to “learn as you go” and run a improvisational sandbox game.
You just need players who are on board with the DM taking breaks and pauses to address rules situations that are new to them. This is in principle no different than a new DM would have to do with a bog-standard narrative heavy, more “structured” game.
Rules familiarity can contribute to the illusion of continuity and the speed that allows a DM to maintain immersion in the game. But it is not a necessity to run a sandbox game.
So condition 1), though important, isn’t a necessity, and a “new” DM can run a sandbox game of this kind perfectly well.
Familiarity with the Setting
This condition is absolutely necessary, even if you have mad improv skills and extreme rules familiarity, if you don’t know the setting you can’t run a sandbox game. The setting provides both the backdrop and the environment for the players to interact with.
The basic premise of a sandbox game is that the players drive what happens through the setting or environment reacting to their actions, no setting, no reaction, no sandbox game.
I think it’s fair to demand that any DM, “new” or old, does SOME prep with the setting before running a sandbox game. So “new” in this case just means “new” to DMing, not “new” in the sense of knowing nothing about the setting. A “new” DM can become familiar with the setting before play, we are not considering some abstract case where the DM sits down with no prep as our case of a “new” DM.
So familiarity with the setting is not a barrier to a “new” DM any more than a seasoned one, in both cases, the DM would have to get to know the setting before starting, and that’s a reasonable request for even a novice DM.
If you believe a fully immersive open-ended sandbox world should not require any advanced preparation or familiarity to run then this is not a method for you.
So for me it is possible for a new DM to run a fully sandbox setting with some
In summary, being new to DMing doesn’t prohibit you from running a sandbox setting of this kind. If you have read through the setting once, or at least most of the setting once, that's enough to get started. By all means get more familiar with it, the point is that a new DM doesn't have to spend years or months learning a setting, just have a degree of familiarity that supports improvisation.
However, I would like to propose another condition that has a direct impact on the
ability to run a sandbox game, 3) The right kind of setting. The wrong kind of setting can make it impossible for a new DM to run a sandbox game, so I want to delve into this a bit deeper.
Building the Perfect Sandbox
About 6 years ago before setting up a home game for my son and his friends,
I read up on settings for about 8 months. I grabbed every one I could get my hands on, and read them straight through. Campaign settings, city settings, old school, new school, independent, etc. I cast my net wide and tried to determine what made these settings compatible with sandbox, open ended gaming.
They varied in many ways, but all had a few features that were necessary for a sandbox game:
- Mechanisms that motivate the PCs
- Mechanisms that impact the PCs
These two mechanisms are necessary for a sandbox game as a key element of a sandbox game is improvisation. Although the setting will provide a lot of what you use in the game, ultimately the role of the setting is to inspire the DM to be able to improvise. In a sandbox
game the players can (attempt to) go anywhere and do anything. You can’t have all possibilities mapped out ahead of time. So the main role of the setting is to have a degree of direct content and a degree of inspirational material for improvisation.
Settings that lack these features don’t work well with sandbox games.
There are many mechanisms for generating improvisational content, here are a few that I incorporated into my setting to achieve these goals.
1. Mechanisms that Motivate the PCs
In my setting these are achieved through factions, maps and environments, rumor tables, and special event (holiday) calendars, etc.
Sandbox gaming works because anywhere the players choose to explore you can provide content, on the spot or by the next session. Each of these mechanisms generates content
when the players interact with the world outside of their own immediate experience.
Special event or holiday calendars create events that periodically appear in the PC’s world.
So for example, one of our games started out with the party witnessing the ceremonies of the marsh festival. During the Marsh Festival the city gives tribute to the lizard men who live in the marsh around the city by burning huge piles of yin-root, a plant poisonous to lizard men. City dwellers gather yin-root in the days before the festival and then it is burned in a huge pyre outside the temple of Kali. The lizard men fly around the city in a circle on their black dragons to signal their gratitude, and denizens of the city take to the canals in boats to watch them.
There are multiple possible angles for encounters and adventures with this sort of event, so it is a direct inspiration for adventure hooks. In addition, external events like this give the illusion of a “real” world to interact with beyond the players. A sandbox setting needs this to create its feeling of immersion and to motivate the players to explore.
Rumor tables are another mechanism for this. In many cases, rumors are not immediately acted upon, in some cases they never are. In my setting I don’t specify if the rumors are true or not, I decide that on the spot if the players decide to carry out actions in response
to the rumor. Chasing down a false rumor can lead you to another adventure hook as you discover it is false but something else interesting is happening. Rumor tables can inspire gaming goals for the party and inspire the DM to new avenues of adventure.
Maps and environments are another mechanism for this. Most areas on the map will be outside of the player’s experience, e.g. they won’t go there. Even with my city setting, which is limited in size, there are huge swaths of the map the players have not explored.
However, even an unexplored portion of a map is evocative to a player, it *suggests*
a greater whole and inspires players to explore that whole, even if in small chunks
at a time. Maps are great for sandbox gaming as the players can point and say,
“We want to go here”, and that drives a series of decisions and events.
Environments or locations (with or without maps) are another driver for a sandbox game. In my city setting there is a shantytown located outside the city walls. Unlike the city where there are laws about carrying weapons and such, the shantytowns are essentially a lawless place. Once the players found out about the shantytown they knew this was a place to go to do “illegal” things, or to find dodgy types who have fled the city to hide.
Just telling the players that something exists can inspire actions on their part. Once the party discovered there was a huge open air market in the city, they immediately wanted to go there and poke around. So environments or locations, mapped or otherwise, are a great mechanism for sandbox games.
But the single most important mechanism for driving events outside of the player’s
immediate experience is factionalization. Give the setting factions, groups with
sometimes conflicting, sometimes harmonious goals.
Factions give the players groups to either form alliances with or form enmities with.
In my setting the city is ruled by 53 noble Houses, each one of them in tension with
the others, competing for prestige, power and resources. Most players enter the city
and quickly become members of a House and begin to work their way up in power
within that House.
This gives them immediate goals to work towards, but goals that are open ended
and long term. So for example, the group might take on a mission for a powerful
House Lord, and in completing the mission gain prestige in that Lord’s eyes. Or, they might run afoul of a group representing another house who have a similar goal.
Or, they decide they don’t want to be beholden to a House, and they try to operate outside of the existing system, which creates it’s own challenges and adventure hooks, as the established city structure is built around the power and authority of the Houses.
Factionalized play produces endless plot hooks and justifications for adventuring.
Factions inspire players to action, through awareness of those factions and how they can either ally or oppose them, or through their actions triggering a response on the part of an existing faction. I’ve used factions in modules as well to good effect.
All of these mechanisms, factions, maps and environments, rumor tables and
holiday events in the setting create goals and events outside of the PC’s immediate
experience, and become things to work towards or to work to avoid. They are
primarily for the players, not the DM, to create an independent environment for
them to interact with.
2. Mechanisms that Impact the PCs
The second feature of a sandbox setting is primarily for the DM, mechanisms
to have the world react to the PCs presence and actions. There are of course
many ways to do this, but the most common methods I found were creating
NPCs and Monsters, and random tables.
NPCs and Monsters
In RPGS NPCs and monsters are the most visceral, immediate manifestations
of the game world. PC’s interact with NPCs, sometimes violently, sometimes not.
When the DM needs something to engage their players, role-play or combat with
NPCs is a straightforward way to achieve this. In a sandbox game the PCs can
go anywhere and do what they want, encountering NPCs and monsters is one
of the consequences of those choices, and they provide the DM with potential
adventure hooks for the game.
Essentially, the PCs poke the world with the stick of their actions, and NPCs and Monsters are the most direct response to that poking.
So my setting has homebrew monsters and a list of prominent NPCs. That list cannot be exhaustive, it would be impossible to stat and detail every NPC in a setting. But a selection of NPCs representing the various groups in the setting is sufficient to the task. When unsure of how to react to the player’s actions, an NPC description or monster ecology might provide the requisite inspiration.
So for example, my players once retrieved a powerful ancient necromantic grimoire from a treasure haul, they had no idea what it did, and were afraid to read it for fear of curses and such, so they took it back to a sage in the city who identified it. I had a NPC warlock in one of the city Houses who was obsessed with ancient necromantic grimoires and knowledge for various nefarious reasons. That warlock was tipped by the sage about the grimoire
and contacted the PCs unexpectedly to obtain it from them. The only reason
this happened was that I remembered that one of the NPC warlocks I detailed
had a thing for necromantic lore.
This is how you bring your world to life.
If I was to point to one of the most brilliant design features of my favored edition of D&D (1st edition AD&D) it would be the proliferation of tables.
Tables take advantage of one of the most important mechanisms of RPG’s, bounded randomness. “Bounded” means you don’t have an infinity of options, instead you have a curated set of options, a closed, limited set of possibilities. “Random” means you don’t just pick what you want,
you roll on the table to determine what happens.
Both of these features are, IMO, crucial to sandbox gaming. It is important to have bounded results otherwise you get results that are not thematically cohesive (ruining immersion). Also bound results can allow you to encode whatever metastructure you want in your game mechanics, e.g. if you want a “balanced” game you can ensure that all the options on the table
are similar in power level, if you want an unbalanced game you can make sure there are some options on the table that are very different in power level.
Randomization is IMO, one of the greatest improvisational tools ever made, and absolutely crucial to sandbox gaming.
To help you improvise, your setting has to do two things with randomization, one, help you find new ideas, two, it has to keep you from always using old ideas. These are two sides of the same coin but worth thinking about separately.
Sometimes you are stuck, you just can’t think of what to do next. So a randomized table allows you to roll and produce a result. You could just invent whatever you want on the spot, and I do a lot of that for sure, but a randomized table has curated options for you, and they can help you
when you are stuck.
The other benefit, from the player’s side, is that randomized tables keep you from doing the same thing too often (sometimes is good, too often is bad). Repetition kills sandbox gaming, if every monster, every trap, every NPC, is the same, players get bored, and immersion disappears as you see what’s behind the curtain.
So randomization is key to making a sustained sandbox game, as sandboxing requires improvisation, and randomization inspires you and keeps you from repeating yourself too often.
I use two kinds of tables in my game that aid with a sandbox setting, location tables and random encounter tables.
Location tables are part of the city setting. My city is IMMENSE, it is that way for a campaign specific reason, but it gives me a massive palette to work with. However, I had to detail entries for
thousands of buildings if I were to do this without tables. Some old Judge’s Guild cities give a listing for every building on the map, but these cities have at most around 100 or so buildings.
Instead, I created ward tables. The city was broken down into 53 wards, one for each city House, and I created tables of the kind of building you would find in that ward. Here’s an example:
8. Scholar’s Ward– Crest: Black elephant on a yellow background
House Pentas - Magus Warlock Kurnadale* (M) - Garudin - Il 10 Karus Lord Nirona the Kair (M) – F13
Pick Pockets: 1 in 6, Foliage Cover: 4 in 6
1-2: Sage: Monstrous Creatures
9-10: Sage: Flora
11-12: Sage: Supernatural & Unusual
15-16: Paper and Vellum
17-18: Sage: Flora
21-22: Ink and Writing tools
23-24: Private Residence
25-26: Sage: Physical Universe
29-30: Sage: Flora
31-32: Outdoor stable
37-38: Sage: Physical Universe
39-40: Sage: Supernatural & Unusual
41-42: Outdoor stable
47-48: Sage: Monstrous Creatures
49-50: Sage: Demi-Humankind Culture and History
54-55: Sage: Humanoids & Giant-kind Culture and History
56-57: Sage: Humankind Culture and History
60-61: Sage: Flora
62-63: Private Residence
64-65: Sage: Fauna
66-67: Sage: Monstrous Creatures
68-69: Sage: Physical Universe
70-71: Private Residence
74-75: Sage: Humankind Culture and History
78-79: Outdoor stable
82-83: Sage: Monstrous Creatures
86-87: Sage: Demi-Humankind Culture and History
90-91: Document preservation
92-93: Sage: Humanoids & Giant-kind Culture and History
94-95: Private Residence
96-98: Sage: Humankind Culture and History
100: Temple of Kroghari, goddess of Knowledge.
There is not meant to be a 1:1 relationship between the number of buildings seen on the map and the number of buildings listed in the table. There are unique buildings on the table (the Ward garrison and temple for each ward), once they are located they stay put, you cannot roll that result twice. I have not put the towers on the table, as they are obvious on the map.
The table can be used in several ways.
If the characters want to visit a particular kind of establishment, say a blacksmith, you can see which ward has them, and a glance at the table suggests their relative
frequency. So you could then go to the map and pick a place that is your blacksmith in a particular ward, and direct the characters there through NPCS who answer their inquiries.
Alternatively, you can let the characters wander through the ward, and roll for each building they encounter to see what it is.
In neither case was it necessary to detail the building ahead of time. Instead, the work was put into the tables. So in this ward (like most) you have some private dwellings, a garrison, a temple, stables, inns and such, and you have buildings oriented to the ward’s theme. There are certain unique buildings that are specified on the map, in this case warlock towers, otherwise it’s all open.
As the characters adventure in the setting, their interaction with the building
specifies its content, like opening the box on Schrodinger's Cat. In most games that go on for a time there will be a small number of specified buildings created by adventuring, but the vast bulk will be unspecified.
There is also a table to help generate the occupants of a particular building:
Establishment Residents Table
When a PC enters a business establishment, you can roll on this table to determine it’s occupants if you do not want to improvise.
Roll one of each of the primary dice: d4,d6,d8,d10,d12 and d20
A. Primary Occupant - (d4 – 1-3: non-classed, 4: classed)*
B. Level of Primary Occupant – (d6 – 2)**
C. Number of Occupants – (d8x2)
D. Alignment – (d10, 1-CE, 2-LE, 3-NE, 4-CN, 5-LN, 6- N, 7-CG, 8-LG, 9-10-NG)
E. Class of Primary Occupant if Classed – (d12 – 1: cleric, 2: druid, 3-4: fighter,
5: Paladin, 6: Ranger, 7: Magic-User, 8: Illusionist, 9-10: Thief, 11: Assassin, 12: Bard)
F. Race – (d20 –1-15: human, 16-18: halfling, 19: Saan, 20: Dwarf(voidnik))
* Non-classed characters will have 3 HP, are 0-level and will have AC 10, classed characters will have HP appropriate to their class and level
** Characters will have a number of magic items equal to their level/2, as follows:
a) a +1 weapon appropriate to their class
b) a +1 protective item appropriate to their class OR a roll on the potion table
It is assumed you will “skin” the NPC to the establishment, e.g. if it’s a blacksmith, then the primary occupant is a blacksmith no matter what class that NPC is.
So my players have been told to “Go to the Hanging Vine inn”, and by looking at the relative frequency of inns in a Ward, I’ve placed that inn. Otherwise, they have explored areas in the city and I’ve rolled for the identity of a building on the spot.
All of this is available to help inspire, you don’t need any of it. You can make any building whatever you want it to be, by fiat. The point of the randomization is to avoid having to detail everything, to give you some inspiration when you have to improvise on the spot, and to keep it from being repetitive.
Random Encounter Tables
Genius game mechanic ahead. Random encounters are genius. In D&D, sitting
around doing nothing means you get bumped around by the game world. It is
meant, no matter what the setting, to be a dangerous place, and random encounter
represent that. Even if you do nothing, something will eventually happen. That’s good game design, as players in open ended games can get stuck.
Random encounters also help inspire a resource management mindset. Knowing
sitting around memorizing spells and healing might provoke an encounter means that activity has a cost associated with it that must be managed or counterbalanced.
They additionally allow you to encode preferred game calibration. For example,
I like an ‘unbalanced’ game where the challenge level is somewhat opaque to the
players. So random tables in my game are unbalanced, including entries that are out of the party’s power range.
Bounded randomness means that you get variety that thematically fits the setting.
Here are my tables:
When in any Ward there is a chance for a random encounter with NPCs that may or
may not expand into a longer term event. Of course, hundreds of people are met
every day, but usually passed by in haste, a roll on the table indicates an extended
interaction of some kind.
Unless otherwise listed in the specific ward listing, the odds of an encounter are 1 in 10 during the day, 1 in 20 at night, check each hour in the Ward. So if you park yourself on a busy thoroughfare in a Ward and don’t move for say 8 hours during the day, there’s about a 56% chance you will have an encounter with someone or something.
Encounters are managed with an encounter reaction roll, with all charisma modifiers in place and a flat -10% for outsiders WRT citizens. Note that even hostile encounters will likely lead to the hostile party leaving if there is a chance of the city watch happening by. Note also that openly carrying prohibited weapons will lead to a -20% modifier. When an encounter is determined consult the following chart:
City Random Encounter Table1-3 individuals appear for any result except where noted.
1- 2: Ward/City Work Crew (12 workers+ 2 city guards + 3-5th lv monk)
3-4: Displaced/poor ( chance of pick-pocket attempt noted in Ward listing)
11: Ward Soldier
12: House Staff
17-18: Ward/City Patrols*
19: NPC - Special
20: Monster - Special
Special NPCs tend to be from the specified Ward, so the DM will customize this result:
1-Significant Non-Classed NPC (e.g. wealthy merchant, important government official)
2-Significant Classed NPC (e.g. noble from a house, high priest).
- Class [d20]: 1-3: Fighter, 4-5: Cleric, 6-9: Magic User, 10-11: Paladin,
- 12: Ranger, 13-15: Monk, 16-18 Bard, 19: Multi-Class, 20: Dual-Class Level [d20/2] – as a rough guide, any classed NPC will have one minor magic item
- per 3 levels of experience, and one miscellaneous magic item per 7 levels of experience.
Monsters Roll on the Swamp Wandering Monster Table below. Note that the city
is fairly good at keeping out monsters, but the swamp tends to impose itself. Number appearing in the city is 1-2 unless otherwise noted, outside the city number appearing is 2-5 for large creatures,
2-16 for small.
Swamp Wandering Monster Table
1-3: Centipede, Giant
4: Crocodile, Normal
5-7: Dog, Wild (2-8)
11-14: Rat, Giant
15-18: Raven, Normal
19-21: Spider, Huge
22-24: Spider, Large
25-26: Toad, Giant
27-30: Vulture, Normal
31-33: Falcon, Small
34-36: Frog, Giant
37: Snake, Constrictor
38-40: Snake, Poisonous
41-43: Toad, Poisonous
44-46: Beetle, Giant Fire
47-49: Boar, Warthog
50-52: Boar, Wild
53-58: Flightless Bird
59-62: Saan (1-10)
63-65: Lycanthrope. Wereboar
66-68: Lycanthrope, Weretiger
69-71: Lycanthrope. Werewolf (2-4)
72-74: Snake, Giant Spitting
75-77: Ape, Carnivorous
78-82: Bat, Giant
83-85: Boar, Giant
86-88: Crocodile, Giant
91-94: Black dragon
100: Marsh Dragon
So whenever the PC’s go anywhere in the city, depending on their method of travel,
we roll at least once per trip, sometimes more than once. And remember that given the table structure and the encounter reaction rules, the vast majority of these events will be non-violent, and involve some RP.
The important part is how this contributes to sandbox DMing. What this means is twofold, first, when you are stuck as to what to do, a random encounter can bridge you to the end of the session so you can walk away and muse it over.
Second, when the players are stuck on what to do, and are spinning their wheels, random encounters can spur them to action or reaction, and keep them engaged in the game. They can inspire diversions in activity and can also be spots for me to drop information (through NPCs).
A combination of items to entice your PCs (factions, maps and environments, rumor tables, and special event (holiday calendars), ideas to inspire you (tables,
bounded randomness, location tables, random encounter tables), and the give and take of PC actions in the gameworld sustain a sandbox game. These are irrespective of a DM’s experience running games. Of course someone extremely familiar with the ruleset and setting will be able to do this more effortlessly, but in principle if you become familiar with the setting and have a basic grasp on the rules you should be able to
make this work.