Quote from: Baulderstone;982124Here is my opinion on the floating tavern issue.

If you have a list of thinly detailed taverns that have a name, some staff and maybe a distinctive menu item or two, I am fine with the GM just putting them down in the players' path as needed. They aren't imposing a story on the players. They just help the GM give the world a little flavor that might emerge into something interesting if the players actively interact with the tavern. It saves the GM the indignity of telling the players that a tavern has no name because it's not important. I don't consider this a serious violation of the rules of sandbox.

On the other hand, let's say GM just watched Texas Chainsaw Massacre and has detailed an tavern run by a family of murderous cannibals with an abattoir/dungeon underneath. If the GM decides that no matter what direction the players take, the next tavern along the road will be this one, then the GM is flat out railroading. He is pushing a threat/story on the players that is unavoidable. This is not sandbox play.

One thing I've done, and this might violate the spirit of sandbox for some, is treat these sorts of things as random encounters with a little bit of depth. I actually have a table of random inn encounters I use for when players go to inns. These are usually just situations that might encounter in any sort of inn, but occasionally they get specific (like a texas chain saw type situation). I find the tables particularly useful for inns I am running on the fly more than established inns in the campaign. So for situations where the players seek out inns I haven't really thought of yet in a city (maybe they are tired of the 3-5 that I have fleshed out and seek out another option).

This is an example of one I used in one campaign: http://thebedrockblog.blogspot.com/2017/05/wuxia-inspiration-inn-encounters.html

That looks like a fun table.

In a sandbox you have to be ready for your players to pounce on a bit of tavern gossip and immediately drop everything to outfit a sea expedition halfway around the world. Despite having nothing prepped for this, and intending it to be merely a bit of background color.  I had this happen to me once, and did it to a GM once myself.  When faced with a choice of going right or left, be prepared for the players to answer "we dig a hole and go down".

I prep my game worlds in broad strokes, with detail resolution becoming finer the closer it gets to where the PCs are. I like to have a good detailing of everything within a campaign map movement turn or two. That way I can hope to keep ahead of them no matter which way they run. If playing Traveller that would be all the star systems their ship could reach in a single jump. In D&D at least a day's march, and a sketch of a week.

Then I improvise like crazy, and write down what happened. Once it has been brought into the game it needs to remain. If I don't write down an NPC's name it is guaranteed that one of my players will, and will pull that name out when I have forgotten it.

There has some good advice so far. My approach is basically,

1. set up a bunch of NPCs and situations with conflicts between them
2. roll everything else randomly, then change bits of it to make it fit in with the rest more coherently, and
3. be one textbook chapter ahead of the students - as Gronan said, one or two dungeon levels, or just over the next hill

Sorry, this is a bit long, but this sort of thing is always hard in the abstract, when you have some play examples it makes more sense.

I'll repost here my response to a similar discussion.

   If you play computer games, think of it this way: don't try to design a main questline, just design sidequests.

As others have said, create a little world where there are a bunch of people with real personalities - simple ones, but real nonetheless - and real histories, all living in a land with real geography. Conflicts are always happening, and the PCs find out about them and decide where they'll step in. For example, "rescue the hostages from the orcs" can have a bit more to it.

Count Barnacle rules a small county on the seacoast, with its capital his keep and market town, known as Seaview. A rare kind of fish is found here, the Babelfish, it tastes awful but its liver is used in a Potion of Speak In Tongues which lets the two people drinking it understand one another despite no common language. It gets mixed in with wine when two merchants sit down to bargain, there's a bit of ritual around it. Taxes on this bring in a steady revenue. Count Barnacle has two children, a 16yo son Johann and a 20yo daughter Marie. They don't like each-other much. As daughter, Marie would not inherit the county unless there were no male heir.

Further up the river is a much smaller county, the March, ruled by William the Ugly. The Babelfish actually spawn in his part of the river, but the King granted the right to harvest them to Barnacle, so what can he do. William is unmarried, and has his eye on Marie. A woman can rule alone, but if she marries, her husband will rule.

Seaview sits beside a river that comes down the valley from the mountains, with its main source on Mount Finicky, so-called because it's rocky and hard to traverse, impossible on horseback. It's also bleak as fuck and nothing grows there. Of course this is where the orcs live in a series of underground caverns and dungeons. In ancient times, dwarves lived in those hills, but they are all gone now. Sometimes the orc tunnel into those old dwarven ruins but the undead and traps take their toll.

The orcs recently had a new chief come to power - he killed and ate his predecessor - Bagolog Stonefist. One day one of Stonefist's raiding parties took a young woman, Sally, who was the daughter of Barnacle's engineer, Harold. From her he got an idea - why not get Harold to help digging into the dwarven ruins? So one day his orcs captured a work party working on a bridge over the river, and took Harold prisoner.

So now the Count wants his people rescued, and especially his engineer. His son Johann wants to prove himself. Barnacle is considering sending 50 of his men-at-arms to deal with the orcs, with Johann leading them. Marie argues that it would be better to send a small group of adventurers, but Johann doesn't want to be associated with rabble. However she strongly urges that Johann go out into facing hundreds of orcs with a small group of armed people. Can you think why?

The party must travel through the March to get to Mt Finicky, by the way, and will need to stop for rest and provisions along the way. The river sometimes flood and there are rockfalls along some of the passes, it would help to have a local guide from the March. Count William would be glad to provide one. Can you think why?

You can also think, will Chief Stonefist mistreat engineer Harold, or will he pay him well and give him access to captured wine and food, or women if Harold is interested? So maybe the guy they're rescuing doesn't want to be rescued? And willing or not, will his advice help the orcs have better defences against assault?

So let's say the PCs go in and kill some or all of the orcs and rescue the prisoners. Did they save them all? Or not? Will this change how the various nobles feel about them? Will the PCs want to go in and explore the deeper dwarven ruins themselves?

And so on and so forth. You create real people with real motivations and real geography, conflicts pop up and the PCs can get into all that. On the other hand they could just charge on into the ruins, kill things and take their stuff. It's up to them.

Think sidequest, not mainquest. In computer games it is what it is. In tabletop games, a mainquest is just a sidequest that went on longer than expected.

As it was, the players using William the Ugly's town Oldenford (I got the name randomly from here) as a base, they went into the dungeon several times. As well as Johann who was Barnacle's son, they had another henchman, Ivyst the paladin. She'd come about when I used the random tavern generator here. She was just one of the patrons and they asked her to join them for justice, loot and glory. Already as a paladin in plate mail there was a risk of her dominating combats, to tone it down I had her always refuse the party's offers of magical arms and armour. I wanted her to be good enough to support the party, but not good enough for them to rely on her to pull their arses out of the fire.

So, in and out of the dungeon of Bagalog Stonefist in Mt Finicky. With breaks to heal or level up, it was actually a period of some months in-game. And the orcs responded by doing a raid on Oldenford, burning the granaries just as winter was setting in. Willian the Ugly demanded a share of their loot to pay for grain to replace that lost in the raid. William then told them, "Because you keep poking the hornet's nest, my people have died, and many will face a hungry winter. You have one more expedition against the orcs, if you have Bagalog's head with you, the gates will be open to you, otherwise, die in the dungeon, or outside in the snow, I don't care."

So they went and slew Bagalog, and came back triumphant. By now it was winter and they had to stay there until spring came. Their man Johann, Barnacle's son, was recovering from illness... and became sicker. On investigation they found his food was poisoned. They blamed William, but said nothing to him. Swimming around the rough seas of court politics could have occupied their winter, but the PCs decided to just divert Johann's poisoned food to William's hounds and left it at that. The hounds died and nothing more was said.

To occupy themselves, they ventured into a smaller dungeon that had been on the way. I'd put a few little dungeons in from Seaview to Oldenford and Mt Finicky, the idea being that they could get some loot and xp along the way; but they basically bypassed them. So now they were going back. This dungeon I had the idea it'd be some kind of ancient temple, but I rolled everything up using this. Among the treasures I rolled up for them was a Talisman of Ultimate Evil (an evil cleric holds it, points his finger at a good cleric, and a crack opens up in the earth and swallows him down to be consumed by lava), and a Necklace of Fire Resistance. Okay, so it's the temple of an evil fire cult. I happened to see an image of Kali in a web advert, so I said "It's the cult of Kolak the four-armed evil fire god."

One of the things I like to do with more potent magical items is make them both useful and cursed - like once I rolled up a Ring of Free Action, you're immune to hold and so on. Who else is immune to these things? Ghouls. Okay, let's make the Ring into a Crown, and call it the Crown of the Ghoul King - immune to sleep, paralysis, hold, etc - but you can only be sustained by eating human flesh.

So I thought, okay, nobody will mess with the Talisman since it's obviously evil, but how about the Necklace of Fire Resistance? Let's make it cursed that if you wear it, you turn into an evil cleric of Kolak.

While they're camping just outside the dungeon they roll up a random encounter. With random encounters what I do is say, if you roll low it's something nasty, if you roll low it's something... interesting. Like one time they rolled high and there came along a local monster, an ogre mage called Bird who owned Gauntlets of Ogre Power and liked to take the form of a human fighter. So this time the chart indicated, "local NPC gone adventuring."

I looked at the NPCs of Oldenford. I couldn't really picture William adventuring, or Pertinax the high level wizard, or anyone like that. How about their sidekicks? Like William's chief housecarl Gylliam, and so on? I glanced down the list, there were 10 of them - so I now rolled a d10, and it came up as Phillipicus, the apprentice to the high level wizard. I hadn't any stats for him, just a name. So I rolled 1d4 for level, and he was 4th. Now I rolled 1d10 for alignment, and it came up as neutral evil. Okay, now this could be an interesting encounter.

So in they go and slay and loot, and get the Talisman and Necklace. They haven't cleared out the dungeon, and do plan to go back. On th way back to Oldenford, another wandering monster check is rolled. This time it's ogres, 2-20 - and I roll up 15 of them. I say, "there's a 1 in 6 chance they're off to the Temple to join the cult of Kolak." And up it comes, 1 on a d6. So 15 ogres in purple robes are striding along. The PCs hide and let them pass.

Back in Oldenford, Phillipicus happens to have the identify spell, and once they're back in town where he can get some pearls (as the magical component to the spell), he tells them what these things are. So the party wants to destroy them, of course. "No," says Phillipicus, "we can use them for... good." They say, "no bloody way", and put them in an iron box which they weld shut, and put in the room of their paladin henchman Ivyst.

Ivyst, you will recall, was initially just a random tavern patron. But they'd got pretty fond of her, as she was courageous and straightforward, and, one played Jon said, "Just her being there stops us doing anything evil."
"But," I said, "she's never vetoed anything."
"Yes, because we knew she might oppose nasty things we wanted to do, we never even brought them up. She's like the governor general or something."

So one morning in the winter they see Ivyst isn't at breakfast, and Phillipicus isn't around. They go into Ivyst's room and find her throat cut and the iron box missing. Checking with Pertinax, they find Phillipicus isn't back with his old master, either. "Bastard has killed Ivyst and taken the stuff to go and be a new evil cleric!" they cry. They get Ivyst raised from the dead and set off to go get him.

Back in the dungeon they find Phillipicus initiating the ogres into the cult. There's a big fight and one by one the cultists and the PCs both go down. With Phillipicus on a few hit points, it's down to Johann and Ivyst. Johann falls and Ivyst strikes the last blow, downing Phillipicus. She takes the necklace and talisman, and walks out of the room, up the stairs to where the lava pit is, and disappears into the darkness there.

And the players say, "Holy shit, it turns out we were the henchmen." I certainly hadn't planned for things to end that way. If I were after a "story", then Ivyst should have died heroically saving the party while they slew the evil guy, or ran away, or buried the evil magic items, or something. But I just let the dice decide, I just moderated the dice's effects a bit. So it wasn't as I expected, but the players loved it.

Note that along the way, the party could have ignored the main quest and just got involved with the smaller dungeons, or the court politics in Seaview or Oldenford, and a couple of places along the way. There was a whole bunch of stuff happening, and it was up to them what they got involved in - or indeed, if they made their own stuff happen. No main quest, just a bunch of side quests.

So like I said, you sketch things out, let the dice decide a lot of things, and just kind of run with it, trying to stay one chapter ahead of the students in the textbook. And players will usually love it.

Quote from: BedrockBrendan;982130One thing I've done, and this might violate the spirit of sandbox for some, is treat these sorts of things as random encounters with a little bit of depth. I actually have a table of random inn encounters I use for when players go to inns. These are usually just situations that might encounter in any sort of inn, but occasionally they get specific (like a texas chain saw type situation). I find the tables particularly useful for inns I am running on the fly more than established inns in the campaign. So for situations where the players seek out inns I haven't really thought of yet in a city (maybe they are tired of the 3-5 that I have fleshed out and seek out another option).

This is an example of one I used in one campaign: http://thebedrockblog.blogspot.com/2017/05/wuxia-inspiration-inn-encounters.html

I think random encounter tables for inns are fine in sandbox play. Just like random encounter tables in a dungeon or wilderness, they help giving a feeling of activity to a sandbox. A common rookie mistake with sandboxes is to think of them as settings where nothing happens if the PCs don't act. Tables like yours make sure that there are ongoing events for the players to be drawn into.

In any case, while I like sandbox play, I don't think sandbox purity is the entire measure of quality GMing. Say you have designed the cannibal murder inn that I mentioned earlier and have yet to use it. You might be running a game session on the week of Halloween. While it might be a "sandbox violation" to drop that inn in whatever direction the PCs go, it might be exactly what the group wants that week.

Quote from: rgrove0172;981899At the risk of raising hackles - what is the difference between a railroading GM arranging for the PCs to head in the direction of the Black Goat and another plopping it down in front of them no matter where they go?

There's no difference, because both of them are railroading. One of them is just trying to hide it.
I have more respect for the guy that does it in the open.

BTW rgrove I've been really enjoying the threads you've started here recently, well done mate. :cool:

Quote from: Justin Alexander;982108Exactly. There's a distinction between sandboxes, non-linear scenario design, and not railroading. In practice, there's a set of personal tastes that tend to associate these ideas with each other. But they're distinct things.

I talk about the fact that treating "sandbox" as the opposite of "railroad" heavily distorts the clear understanding of both terms in The Railroading Manifesto.

It's a good article, like all your stuff - but then you already knew that.

So when we cast "sandbox" and "railroad" as antonyms, we actually end up treating the lightest form of railroading as if it were the most extreme form of railroading. And, in response, the meaning of "sandbox" gets warped towards meaning "anything that isn't linear". Neither distortion is useful, with the former radicalizing our understanding of railroads and the latter eradicating the unique utility of the term "sandbox" by turning it into a synonym of "non-linear design".

Definitely agree - we're seeing that in this thread. Also your point about 'default to yes' (to player idea) being the opposite of railroading. You can certainly have sandboxes where the GM has pretty rigid ideas about the setting, where players don't get to add material, it's hard status quo so many/most paths are unviable - and indeed those GMs will often claim that this is the only true sandbox (I'd say it was a sandbox, but definitely not the only kind).

The person running this website is a racist who publicly advocates genocidal practices.

I am deleting my content.

I recommend you do the same.

Quote from: Gronan of Simmerya;982048Only one person said that, and frankly, they're wrong.  That is the OPPOSITE of a sandbox.  If the Goat's Anus Inn is in the town of Fuckmorton, then that's where it is, period.  If the trolls' nest in the first level has six trolls, it has six trolls.  If you search this site, you will find about 95% of the people advocating sandboxes agree with this.  Moving shit around so the players WILL encounter it is the very definition of railroading.

Would't that require having a fair bit of the world mapped out exactly ahead of time? When I use to improv a lot as a DM I would usually just make up shit as I went along to a degree greater than what you're suggesting.

Quote from: rgrove0172;981899One element I THINK Im hearing but would have trouble with is the notion of making up locations/NPCs/situations etc. and just moving them around depending on where the players go. If I understanding this you conjure up a tavern for example and then just plop it down in whatever town the PCs visit, that way you dont waste it should they go in a different direction.

If thats common practice, I gotta say Im not a fan.

Far as I know its not and is more often viewed as a form of cheating. Theres some older threads on the subject. But the general consensus is that prepping an area is one thing. Moving the area into the PCs path when they went left instead of right is a rather different thing and usually seen as bad DMing.

What some are noting instead is where you prep some things for when caught off guard so you dont bog down. Like you have a tavern prepped with some NPCs in it. Its not tied to any local. Its there to drop in if the PCs walk into a town you havent fleshed out and say "We walk into the nearest tavern. What it called?"

Or for random rolls. Theres a chance of bandits on the road so you prep a group in case it comes up. Same with the road patrol. It could be as simple as how many and are they mounted or on foot. Or more elabourate.

Or say a player says they want to talk to a local beggar. If you have one prepped then there you go. Use or say there arent any in town. And so on.

So overall its meant as a time saver for totally unexpected moves. Not as something to replace establishing the area.

Quote from: S'mon;982198BTW rgrove I've been really enjoying the threads you've started here recently, well done mate. :cool:

Thanks, I mean for them to be informative not controversial.

This is quite a useful thread *subscribed*

In my opinion the best sandbox games are localised not global and the history and current events firmly established prior to take off. Just don't overdo the detail, your not Tolkien and no way are you prepping every town, NPC and bush even in a small area.

Once the game starts your going to move set pieces and alter the outcome of events based on the players actions. I like to have a few adventures happening that the characters can get involved in but generally they should be low to slow impact not apocalyptic, that way if the players aren't interested in beating up on the child stealing cult your not nuking the entire game.

Have a few easy hooks you can drop in for when play stalls but generally your probably going to find tons of leads from the players themselves.

Quote from: rgrove0172;981834Ive been guilty of throwing some bad attitude towards the "sandbox" crowd before as I normally lean heavily on pre-game prep to a crazy degree. The closest thing I can say I typically come to sandbox (except for a couple of exceptions such as when playing in the modern world) is a massive amount of preparation completely surrounding the PCs so that Im ready no matter where they go. (and they still typically manage to pick some corner Im weak in)

First of all, there is no One Way to tun a sandbox.

Everyone has their own ways and techniques.

The way I do it would have sandbox zealots throwing their hands in the air and shouting "That's not a Sandbox".

My number one rule of a sandbox is "Don't over prepare". I once spent a few months writing up a 20 storey tower full of vampires (Tower of Lead in Dorastor, Glorantha) and the PCs came close and said "We don't like the look of that" and turned around. Nowadays, I don't do anything like that.

Learn to Improvise - If the players say "We want to go to Hobbiton", let them go and make things up on the spot.

Use plots liberally - (Sandbox Zealots, now is the time to throw up your hands) Prepare some plots and scenarios to drop into the campaign at opportune times. If the players get stuck for something to do, throw in something from a background plot to spark things off.

Don't be afraid to run with things - If the PCs decide to kill the Princess and rescue the ogre who kidnapped her, don't complain about them ruining your game. Instead, follow the concequences. maybe they become friends of ogres, maybe the Princess's family send people after her, maybe the Princess was the head of an evil cult and the ogre was tryign to defeat her.

Quote from: rgrove0172;981834Im swallowing my pride here though and am asking earnestly - as someone who wants to learn - how to you sandbox types do it? When I hear of someone playing a session completely with multiple towns, businesses, NPCs, encounters etc. all from the notes on the back of a napkin I have a hard time ...
1)believing they are telling the truth
2)believing the result doesnt suffer from "pulled it out of my backside, minimal background, itis"

It helps to be familiar with the setting, or to have written your own setting.

For example, if I ran a Robyn Hode campaign, how much do I need to know? The Sheriff of Nottingham is a bad guy, Prince John is a bad guy, Sherwood Forest and the surroundings is the Sandbox, the PCs are outlaws, several key NPCs can be written up. If the PCs go to Nottingham then they might meet guards at the gate, wealthy merchants to rob, peasants to help and so on. It helps to have a one-liner of many towns and villages in the area and of the major NPCs, then you can riff off what happens.

Quote from: rgrove0172;981834I know that published settings, adventure modules and the kinds of games those that preplan everything put out include a great deal of effort to make sure the componants of the game fit into the setting, the regional history, the culture and so on. It seems it would be hard to wing this sort of detail at the gaming table on the fly. Obviously some of you do it regularly.. so Im asking how... lend us a that would like to try it a bit of advise.

Make things up- on the fly. Simple as that.

This has turned into a really great thread. Here's some tips I have posted elsewhere:

QuoteRe: Have you run a dungeon crawl with a system that isnt d20?

I have run a really great short dungeon/sandbox campaign with Earthdawn. My players crawled through the Moathouse and got involved with Hommlet's politics (T1 Hommlet). You will want to reference the classic game for things like time-tracking, random encounter charts, etc. I have been going back to the early rpg materials to re-examine their playstyle, and have found a few techniques that help the GM to manage play.

First, make sure you are not asking for indvidual character intents at the start of each 10 minute exploration turn (or 4 hour overland exploration turn, or 1 day travel turn, etc.... this ain't combat). Ask the party to decide among themselves (in-character, even) what they will do and have one player tell the GM the party's intent. In this way, the party moves through the dungeon (or over the overland map) like a war-gaming unit & you can move groups of monsters around as war-gaming units in response. Then, if necessary, the GM can handle individual character's actions (X searches the south wall, Y goes and finds fresh water, etc) based on the intent given at the start of the turn.

The other technique I created to help was pre-rolling Random Encounter Lists. This is a form of "procedural generation". Take your area's Random Encounter Table and roll a list of 5 items. Now, whenever the player's RE Check roll comes up positive, just use the first thing on the list and cross it off.

Pre-rolling these results allows you to know _what_ happens, and _in what order_, but the RE Checks will ultimately determine the _pacing_ of when each thing actually occurs in play. This allows a bit of prep time to think about how to integrate each random result, or tie it into the adventure because you are preparing this before the session. Only jot down brief notes. For example: "8 additional gnolls will show up at some point, so they're off camped a distance away from the main horde, maybe they are scouts", or "Giant Bats at some point. Cool, now I can work foreshadowing in, like rustling above the players heads as they enter the cave, the smell of guano, etc." 5 Random Encounters for each "dungeon level" (or "area") you expect your players to be exploring next allows you have procedurally generated content ready to go, but not to get lost in over-preparing or over detailing each encounter.

You can also use this procedural generation technique in "sandbox" areas when the players are not around. Roll up your list of events, then make an RE Check for that area each day, week, or month (Depending on campaign needs). If it comes up positive, use the first item on the list to change the situation and evolve the "living world" aspect of that place. For example, the players clear a goblin cave-nest (an adventure site) in a forest (an overland map area). Roll up 5 events and then each week of in-game time check for a Random Encounter for the forest. If it comes up positive, the first item on the list is some-how involved in the old cave-den (a paleo-bear moves in, dwarves move in and expand the tunnels, or whatever is first on the list). As these "events" are derived from the surrounding forest's Random Encounter table (which the GM should feel free to change as the campaign progresses), this allows the area to evolve in ways that will surprise both the GM and the players but stay withing the "theme" of the area as defined by it's Random Encounter Table. Another example would be putting a Named Dragon on the RE Table. Once you've generated that dragon (either in response to the players' exploration, or as procedural context to change the situation elsewhere), you replace that slot on the RE Table with something else.