An excellen introduction to the Sandbox at Raven Crowking’s Nest. I has a number of links, starting with what is great about the Sandbox and discussing the rules that enable it, and the rules that thwart it.
I’ve blogged quite a bit about running a Sandbox, and I’ve added my Swiss Referee Style Manual to my house rule document, Halberds and Helmets, which also has some points on how I run my sandbox. And yet, perhaps the author of the Sandboxes and Quagmires blogpost is right: we should also talk about failure modes and how to prevent them. +Ed Ortiz mentions the following problems:
What works at my table:
Clearly establish which plot elements belong to which character. This is how we make sure that plot time is distributed fairly even though many players have a thing going. It sounds weird, but saying it at the table makes it easier for people to make fair decisions. Resurrecting Arden is Johannes’ plot element. Building the ivory tower is Claudia’s plot element. Going after bandits is Flavio’s plot element. Sometimes it isn’t easy to say. Samuel is easy going and he seems mostly interested in spreading poisonous giant frogs wherever he goes, for Tsathoggua. Michael is mostly interested in getting treasure and better armor and avoid all dangers. (Chicken!) Lilly is new and hasn’t found her thing, yet. Stefan is interested in things, but I haven’t felt a particular push in any direction. But, knowing that we’ve done a number of sessions pursuing Johannes’ plot, it makes it easier to say that the next few sessions will be about Claudia’s plot, out of the game. This is not an in-game decision.
Explicitly list open plots and ask for preferences concerning the next session. Even if players cannot decide, or no majority can be found, at least you can prepare for one of them and tell people that you’ve decided that they were going to go after X. Narrate the transition and off you go. It’s not “pure” sandbox—the players can see the man behind the curtain when they read their emails, but I don’t think that’s a problem. They couldn’t make up their mind and the referee picked Limbo and Slaads for the next adventure. If you didn’t like it, why didn’t you say so when you got the email? Sometimes this will fail and the referee will have to improvise. It happens. It’s OK. But this is important to me: This, too, is not an in-game decision.
Provide enough information. When I recently listed the open plots, I provided more information than the characters actually had available at the time. It went something like this: You could a) go look for the Formian city mentioned by the slaad spies and try and prevent the spread of the iron shadow, or b) visit Limbo, the home turf of the slaad, looking for a grey elf wizard who supposedly researched the iron shadow, or c) learn more about said grey elf wizard by visiting his home town in the astral sea, or d) continue exploring the mirror labyrinth (and stumbling into the Red and Pleasant Land, which I didn’t tell them). Provide more information than is strictly available in-game.
Make sure there are consequences and announce them. You don’t have to be super explicit, but if you take the golem armor made of old brass magic off the dead dwarven hero and envoys ask you to give it back, and you don’t, and instead you write a letter to the dwarf clan saying that you’ll wear it and use it wisely—then there will be consequences. The enjoys will fume. The scribe will shake his head. And the campaign news page will describe the dwarfs raising an expeditionary force of about two hundred dwarves and there will be interesting sessions ahead. Make sure that interesting actions have interesting consequences and make sure your players know.
Recently, Rob Monroe asked about tools to use when preparing a sandbox.
I listed the following:
Alternatively, I could also imagine running my next game as a point crawl based on the structured input of my players as suggested by +Jason Lutes in The Perilous Wilds.
Jason Lutes then linked to two examples of how he did it in 2013, which are also based on blog posts by the Welsh Piper, random generators by Chaotic Shiny, and all that, in order to create the world and a village.
It looks like a lot of work but it looks super beautiful!
I still think that starting out with a random Text Mapper Map and adding notes and settlements to the text describing the area and using the tool to render a picture of the area still is the best value for my time. See How To Get Started With Text Mapper for more.
Recently, William Nichols argued on Google+ that some games avoid the dichotomy of Sandbox vs. Railroad one often sees discussed. This was in reply to me sharing a hilarious video, about 15min, about two campaigns: the sandbox that turned into The Hobbit and the railroad that turned into The Lord of the Rings. It comes with many asides that I remember myself thinking when I was younger, e.g. the idea that players had a social obligation to go along with what I had prepared.
William basically argued that some rules designed the problem away by using improvisation and he listed Dungeon World (which I have run) as well as Fiasco and Apocalypse Now (both of which I have played).
I think these examples are definitely role-playing game designers trying to design their way out of the problem space of “wasted prep” – either because it’s a lot like work for the GM or because it affords railroading, which is not fun for players.
But then again, if you manage to set expectations such that people know that some parts of your game are not improvised, then these locations on the map will be “more real” than things you all just thought up. That’s how I work, at least.
So that’s the counterweight I see: we can design away the option of a railroad, but we must be careful not to design away an important source of immersion, the suspension of disbelief that there is an actual, imagined, shared, pre-existing world out there. For me, that idea is powerful. In games that afford a lot of improvisation, this is often lost, I feel.
Dungeon World navigates this by suggesting the creation of a map beforehand and Perilous Wilds even offers a procedure to create a shared map at the table.
To make a long story short: I think it’s important to remember that adding more improvisation also means that you loose something. Being aware of that trade-off is important.
Ramanan S. recently asked on Google+: “So what exactly do people do to track what the hell is going on in their game? What stuff do you have on hand when running a game?”
I replied the usual stuff. Stage fright never goes away. Keep notes on a Campaign Wiki.
And I mentioned how prep using fronts has been creeping into my game.
So here’s the evolution of how things had been going, on a campaign level. First, I had a passive world, waiting for the players to mess with it. My motto was and still is: “The harder you look, the more there is to see.“¹ Then I started using An Echo Resounding and thought that the domain game would provide for the kind of slow movements in the world around them. As it turned out, the domain game didn’t get my players excited. It felt a bit like accounting and it was too much effort to simply introduce some random setting changes. I then turned to using a random table to introduce setting changes. But we kept forgetting to roll on the table. There was simply no incentive. So finally I have arrived at Fronts.
Fronts are easy to write up. Here’s what I have been using:
¹ The longer form of my motto is this introduction I recently elaborated:
“We’re playing in a sandbox. Dangers are not adapted to the strength of the party. Generally speaking it’s safer near civilized settlements. The further you move into the wilderness, the more dangerous it is. That’s how players control the risks they want to take.
You learn of rumors from travelers in taverns, merchants at markets, sailors at harbors, books in libraries or sages in their ivory towers. This information is not always accurate or complete. Use these rumors to add new locations, goals and quests to your map. The actions of your characters determines the direction the campaign will take. There is no planned ending for the campaign. As long as you keep investigating rumors, exploring locations and following quests, I will keep developing the game world in that direction. The harder you look, the more there is to see.”
On Google+, Brendan S asked for an article about sandbox play for someone with no sandbox experience. I thought of the intro page I wrote for my own campaigns, back in 2012. Sadly it doesn’t talk too much about sandbox play. It’s also the first page of my Halberds and Helmets house rules.
The rules offer very little mechanics: there aren’t many classes to choose from, no feats, no skills, no prestige classes and hardly any special abilities. Furthermore, elves, dwarves and other demihumans are simply separate classes. There are no elven thieves of dwarven mages. On the other hand, missing rules also leave a lot of freedom for players. The characters are as diplomatic, friendly or intimidating as the players want them to be. There are no rules governing it.
We’ll add rules as time passes. Discovering and befriending intelligent humanoids, for example, will allow you to hire them and eventually to play them. Certain magic-users can teach player characters new spells, too.
We’re playing in a sandbox. There is no planned ending for the campaign. The actions of player characters determines the directions the campaign grows in.
You learn of rumors from travelers in taverns, merchants at markets, sailors at harbors, books in libraries or sages in their ivory towers. This information gained is not always accurate or complete. Use these rumors to add new locations to your map and determine your goals in-game.
Players determine where the campaign will head. If player characters investigate rumors and locations, I will develop the game world in that direction. The harder you look, the more there is to see.
Dangers are not adapted to the strength of the party. Generally speaking it’s safer near civilized settlements. The further you move into the wilderness, the more dangerous it is. That’s how players control the risks they want to take.
Preparation and experience should help you avoid situations where your character’s survival depends on a single die roll. If you’re rolling dice, it’s already too late. A saving throw is your last chance to survive due to luck and experience. Ideally you would never have to roll dice because you’re well informed and equipment. Perseus didn’t have to save against the medusa’s petrifying gaze because he was well prepared.
Retainers are another safeguard against character death: torch bearers, porters, men-at-arms and mercenaries all cost money, but they will also keep your character alive. Should player characters die, the next character is most probably going to be one of the retainers.
Experience points (XP) is gained by spending the gold you gained in adventures. If you manage to obtain the gold without combat, good for you. The best strategy is to pick your battles and stake the odds in your favor as far as possible. Remember, if you’re rolling dice, it’s already too late.
Reading the Dungeon World chapter on fronts makes me want to rewrite the list of open plots and the todo lists for a quest or two, and the list of random upcoming campaign changes as fronts. Perhaps that would make all these things clearer to me. Now that I think about it, my campaign threats are a confusing mess of half baked ideas. They work – I think – but perhaps they’d work better if written up as fronts.
See the picture on the right for what I have for my campaign fronts. I probably have one or two more which I don’t consider to be a urgent. One thing I noticed is that the old structure of my notes was this: if you want to resurrect Arden, you need to do the following… and what followed was a list of quests, each of which I felt would make a nice adventure, should the players decide to follow up on it. The write-up as front changes the setup: if players don’t resurrect Arden, his insanity will spread, somebody else will take the throne of light and so on. I’m not sure I like this shift from “this is a sandbox and whatever you want to achieve will be full of adventure” to “the world will go from bad to worse if you don’t take matters into your own hands”. I suddenly feel like might be preparing two or three campaign arcs or adventure paths… a kind of campaign setup I tried to avoid because players end up feeling like they have less choice. Everything is falling to pieces and there is pressure everywhere and time is running out and go, go, go!
This seems to be the biggest difference in terms of how fronts work compared to my traditional preparations. In my sandbox, players get interested in things, they learn more about it, they formulate goals and then they discover all the difficulties that need to be overcome. The world is essentially static.
Sure, we like to talk about “living” sandboxes and all that but my campaign events are random intrusions where I think to myself, “an invasion of mind-flayers sounds great” and then the setting starts to change.
This process is less structured than the fronts of Dungeon World. Fronts are also tied into moves, so a failed roll by a player can advance a front.
No such thing happened in my sandboxes. People felt free to calmly consider the missions they care about and do some horse trading: “You’ll help me bring down Susrael and I’ll help you bring back the fire giant’s wife, OK?” Fronts put pressure on players and I don’t think they’ll feel as free to pick and choose because there will be consequences, always.
Anyway, I recently bought Freebooters on the Frontier, A Book of Beasts, Perilous Almanacs, The Perilous Wilds and The Perilous Wilds Survival Kit by Jason Lutes as well as Dungeon World by Sage LaTorra and Adam Koebel.
Comments here or on G+.
OK, so we talked about setting up a game of Hexcrawling and how the game will eventually reach its limit if the known region keeps growing and more and more factions are being introduced, more lairs, more assets, more domain turns; the game starts to collapse under its own weight. We also talked about my Domain Game Goals. The things I like. The things my players like. We have come to the point where we need to talk about the kind of procedures that will offer us an interesting domain game without growing as the domain expands.
I think this is key: The procedure must always take the same amount of time. Think about random encounters. No matter how big your party, you always roll once for random encounters. The monsters might be stronger. The trek might be longer. But the number of rolls is constant. But think also about its failure modes. If the party travels for eight weeks, do you roll for over 100 random encounters? I don’t. That’s why random encounters only work at a certain scale. Our domain game procedure will also work at a certain scale. We’ll postpone thinking about attaining immortality and godhood, for now.
The simplest solution would be a random domain roll. The results on the table are all either adventure hooks or role-playing opportunities where we get to see what kind of people the player characters are.
Several things are still missing. In order to track the “mood” of the current campaign arc, you could run with Chris Kutalik’s idea of a chaos index as explained in his blog post The Weird is Rising, Thanks World Engine.
I think I’d like more of a multi-dimensional framework that takes the gods into account. You could use something like the fronts on the MC sheet for Sagas of the Icelanders. Have a list of gods or other influences, list some keywords (“Hel: breathe disease, consume, hoard with greed”) that will color current events. This forces you to vary the description of the results depending on what front is in ascendancy. Use the result of the random domain roll to build a little four step countdown. If the party does not engage, step one happens. If they leave it to fester, step two happens. If they are busy elsewhere, step three happens. If they don’t take care of it now, step four happens. As time keeps passing and more rolls are made, issues are piling up. This is good.
If your players have “traits” that influence the domain game such as Sticky Fingers which I mentioned in previous post on the same topic, some of the results on the domain roll table should reflect that. In a Dispute situation, for example, Sticky Fingers might allow you to ignore the first two steps of the countdown as your thieves infiltrate your neighbor’s domain. You will have to handle the issue eventually or just move to War.
The important thing is this: I’m looking for a solution that limits the number of dice rolls and that doesn’t require any sort of computation before rolling. I don’t want to roll for every unconquered monster lair. I don’t want to add a bunch of numbers on the wiki for every roll I make. I don’t even want to look at what the last roll four sessions ago was before making a roll.
I recently wrote about my current setup for a campaign wilderness map and the associated hexcrawling that goes along with it. The greater context is the promise of ever changing gameplay. This is true for characters with saving throws replacing armor class as your most important defense, this is true for spells that change how the game is run, and I want it to be true for the campaign itself where dungeon looting yields to wilderness exploration, and eventually to kingdom building.
Kingdom building is what the domain game is all about. Wilderness exploration is about travelling from here to there and the creatures you encounter. It’s about learning who your allies and enemies are, new towns with new leaders and their own economic goals, monster lairs, humanoid tribes, instigating war, brokering peace. Eventually, the players are going to lay claim on a lair or a town. Now what?
Let us consider existing options for the domain game. The simplest rules I know are the ones in the Expert set by Cook and Marsh. Fighters get a land grant, build a castle, clear the surrounding area of monsters, organize patrols, attract settlers, raise taxes. Any mercenaries hired cost money. Clerics do the same thing, but their castle is only half as expensive and they get fanatically loyal troops for free (5d6×10). A magic-user gets to build a tower and attracts apprentices (1d6). A thief gets to build a hideout and attracts more thieves (2d6). Demihumans are like fighters. They build a stronghold and attract settlers of their own kind. Elves are automatically friends with the local animals. As for the attraction of settlers, all it says is that spending money on improvements (“inns, mills, boatyards, etc.”) or advising will do it. The details are up to the referee.
If you want a bit more detail you can use An Echo Resounding. It’s what I have been using for a while. A while back, I wrote a summary of the rules. Apparently you can add a lot more details by using Adventure Conqueror King System. There is an interesting comparison of An Echo Resounding and Adventure Conqueror King a forum I read a few years ago.
Unfortunately it’s turning out to be too much work for me. When I look at the monthly campaign summaries—something I write every four sessions—I notice that there is some free form stuff in the Sages and Spies inspired by recent events, my players’ interests and adventure hooks, and there is some stuff generated by the rules of An Echo Resounding. For every lair I need to find out whether it spawns units. If it does, these units need to attack a nearby location. I need to resolve these fights and if the units win, they plunder the location they attacked. For every non-player domain I need to figure out what sort of move they make during their domain turn. This involves looking at the numbers and rolling a d20, but often it has been so long that I feel I need to double check those numbers or I find little mistakes. In the end, a lot of time gets spend for very little gain. Or, to look at it from another perspective, I spend some time looking at numbers and rolling dice to produce text that is boring compared to the free form stuff I write up for the Sages and Spies section.
The stuff players like about the system don’t involve that much maintenance. They like knowing about their units and they like going to war every now and then. They like to build things in their domain. In my game, gold spent yields experience points. Since I have a list suggested prices for buildings, this encourages them to build temples, hospitals, towers, bath houses, and so on.
|a small statue for a well or a garden||50gp|
|a small, public altar made of stone with spirit gate und a small well (5ft.×5ft.)||250gp|
|a small shop made of wood with a place to sleep in the back room (15ft.×15ft.)||300gp|
|a simple wooden building with one floor such as a tavern, a gallery or a gambling den (50ft.×50ft.)||700gp|
|a wooden building with two floors in a village (50ft.×50ft.)||1500gp|
|a stone building with two floors in a village (50ft.×50ft.)||3000gp|
|a manor house with two floors, marble columns and statues in a city (50ft.×50ft.)||10,000gp|
|a provincial castle with six floors (60ft.×60ft.) and an inner courtyard (30ft.×60ft.) surrounded by a wall||75,000gp|
This leads to a strange effect: Build a large wooden Freya temple for 1500 gold and you’ve got a temple and 1500 experience points (gold spent = xp gained). Spend a few domain turns building a temple, however, and you will have a temple, it will give you Wealth -1 and Social +4, and a powerful 9th level cleric will come and settle here (using An Echo Resounding).
Having two very different ways of building a temple complicates things. It seems to me that paying for the temple using their own gold is a more visceral experience for players. They built it. This is what it cost. It’s easy to embellish it. It’s easy to list it on the campaign wiki. It doesn’t require anything on my part except determining a suitable price when they ask for a quote.
I also think they don’t mind getting a 9th level cleric, but there are still questions: why haven’t we met them before? Why aren’t they coming on adventures? In fact, why isn’t this a player character?
My game allows players to run multiple characters. In a particular session, players can bring up to three characters. The character with the highest level is the main character, the others act as secondary characters. Experience point gained for killing monsters is split on a per head basis. Treasure—and therefore experience points for gold—is split by shares. Every main character gets a full share, every secondary character gets half a share.
Sometimes, players will grow tired of characters. Sometimes, characters will break bones or loose limbs. These characters are perfect fits for these roles. Majordomos of castles, priests in temples, heads of guilds, captains of ships, regents of towns.
This is how I hope to achieve a greater identification with the setting. Over time, more and more important folks will be former player characters. It’s also ideal for a new campaign. At first, no high level priests exist. As soon as the first player character cleric reaches 9th level, however, raise dead is an option for all the player characters in the region—even if they’re playing in a different group! And raise dead will remain an option even if the player running the character abandons them or if the player leaves my table. The character has been established, backstory included.
My players also love their units. This is not a problem. We can keep the champion levels introduced by An Echo Resounding. The chapter introducing champion levels is Open Game Content. I’d go further than that, though. We could get rid of all the resource points and simply say that all other need to be equipped and hired.
The party could build an armory, buy equipment for four hundred heavy infantry (swords, chain and shield is 60 gold per person based on prices in Moldvay’s Basic D&D or 24000 gold total + 3000 gold for the armory itself based on my list of buildings above). Then, if the town is big enough to supply enough able bodied fighters, four units of heavy infantry militia will automatically be available whenever the town is attacked.
Hiring mercenaries will require less money. Human heavy foot guards in peace time will cost three gold per month (1200 gold per month for four units), twice as much in war time (2400 gold per month for four units).
I don’t think I need to use the War Machine rules introduced in the Rules Cyclopedia. I can keep using the unit combat rules in An Echo Resounding, the B/X Companion by Jonathan Becker, or the M20 Mass Combat Rules by Greywulf. I’m not sure what my favorite mass combat rules are, for the moment. I’m tending towards keeping the rules from An Echo Resounding because rolling for attack and damage is easy to remember. There is no scale factor and there is no /Unit Attack Matrix/. That makes it easier to understand.
What about the abilities your champion gets that aren’t tied to units? Sticky Fingers gives you +4 Wealth value. I don’t want to think about domain income, upkeep, taxes or tolls. When Chris Kutalik started rethinking domain-level play in his campaign, he suggested the use of domain skills and a skill check to go along with it. I don’t want to introduce skill checks and I don’t want minor and major skills in my game, however. Sticky Fingers does sound like a skill, though.
So, that’s where I’m at right now. What about abilities, or aspects?
Based on a recommendation on Google+ I took a look at Houses of the Blooded. There, you have domains consisting of provinces and each province consisting of ten regions. Each region produces something, and based on that you can have armies, goods, trade, and so on. I think it interesting, but I don’t think I’d want my D&D to be about it. Too much detail, it’s not really part of player characters, we wouldn’t want to spend time on it at the table, and so on.
I was also looking at the King Arthur’s Pendragon and The Great Pendragon Campaign. My campaign fell apart because of many reasons, but the lousy winter season where you’re supposed to look after your family, your manor house, your lands, build fortifications and all that—this part of the game just was not exciting enough at the table. And that is a problem. As Chris says in one of his blog posts, there’s always the danger of these systems turning “boardgamey” or “beancounterly.” Or that all the decisions have no consequence after all.
I’m still chewing on this.
Johnn Four asked on Google+: “I haven’t run a treasure-based XP game in a long, long time. I have also never designed one. What’s the general process for planning out a treasure-based adventure?”
I said that I find the key to be the following:
Thus, if players are level 5 and want to keep fighting goblins, they will practically not gain a level anymore. If they want to level up, it’s time to face those trolls or rob that castle…
It’s harder to design the minimum number of rooms per levels for a megadungeon and the mean treasure parcels you want per level. You can do that, sure. I hate this kind of work. That’s why I just design open ended dungeon (if at all): if you explore further, I’ll add further! (Between sessions)