Over the last 9 years or so, I've been playing - with roughly the same core group of players - in three sandbox campaigns I really enjoyed. Based on those experiences, here are a couple of thoughts:

- Don't feel beholden to dungeoncrawling.

There's a strong current of groupthinking in old school D&D circles which states that dungeons should be the main thing in a campaign. From this follows that the main thing in a sandbox campaign is that you're crawling dungeons, and it's "sandbox" because you get to choose which dungeon to crawl. Well, that's wrong.
In our 9 years and three campaigns, we've explored exotic islands full of dangers, mysteries and riches, saved imperilled communities (once retroactivelly, centuries after their actual demise), upset the balance of power between the potentates and master schemers of bustling cities, helped a coup inside a powerful dark cult, sought out an artifact that could save or doom a world, went from vagrants to fugitives to grand heroes, and are presently trying to save 17th century Switzerland from a Templar vampire and the army of the Damned.
And all through this, dungeon delving was almost always on the sidelines, at best. Sure, there were various enclosed locations like mansions, temples, fortresses, caves and the like, and these saw quite a bit of action. But actual, traditional dungeons - large, topographically complex, enclosed adventuring areas with a self-serving reason of existence (i.e., they exist so they can be explored, and the party explores them because they are there) -, while not completely absent, were certainly limited in the presence, and when we went there, it was almost always in support of our wider goals and activities: we didn't explore them because they were there, but because finding something or someone or someplace was a tool we wanted to achieve something else.


- Sometimes it's okay to narrow down the sandbox.

Several times our campaigns were losing focus. Instead of having some sort of reasonably clear-cut goal or motivation, we would end up going into a dungeon - or sailing to a new island - just because it was there. At these points, the DM always introduced a new event, factor or plot that somewhat reduced our freedom (to putter about aimlessly, as it were), but which also gave us a new, clear focus - which we could then tackle in a free, sandboxy manner.

- Adventure hooks.

Since you specifically asked, I think it's totally fine to start a sandbox campaign with a single preset adventure which established the party (they got into this situation together), and which can be followed up later if they choose to do so. (Ok, the evil wizard is dead, but his plan might have already been set into motion. Or maybe this enemy of his you've learned about might want to reward you. Or maybe that artifact you took from him coild be investigated, it might come in handy.) Also have maybe 3-4 hooks for small, relatively simple things which the players might or might not bite on, and which might or might not lead into greater things. The sort of thing the party might learn via rumours: two rival alchemists are having a spat over some matter. A group of religious fanatics is looking for mercenaries for some undisclosed work. A noble has a magical ship in harbour that travels without sails, oars or crew.

In general, a good sandbox isn't one where you can ostensibly do anything but in reality will end up just going to the dungeon because you can't think of anything else to do; it's one where there are always things - possibly simple, but still handcrafted, DM-designed things - to do, which can be approached in numerous ways and freely taken up or dropped.

I have been running my Ogre Gate games mainly as sandbox for the past two years. I would be reluctant to label them pure sandbox. So take my advice with that in mind.

I've benefited a lot from Estar's posts on the subject. I have never run into problems employing his ideas.

What I would emphasize is the importance of tracking NPCs. Failing to track information about NPCs was the one thing that led to issues for me. At first I had individual entries on each NPC. I still do but I found having a master table with every NPC on it where I track ongoing developments to be the best solution. I don't keep stats here. I just add a sentence or two any time something new happens with them. You'd be surprised how easily you can forget very basic things (like is this guy alive or dead) if you don't track it consistently.

Quote from: BedrockBrendan;897189I have been running my Ogre Gate games mainly as sandbox for the past two years. I would be reluctant to label them pure sandbox. So take my advice with that in mind.

I've benefited a lot from Estar's posts on the subject. I have never run into problems employing his ideas.

What I would emphasize is the importance of tracking NPCs. Failing to track information about NPCs was the one thing that led to issues for me. At first I had individual entries on each NPC. I still do but I found having a master table with every NPC on it where I track ongoing developments to be the best solution. I don't keep stats here. I just add a sentence or two any time something new happens with them. You'd be surprised how easily you can forget very basic things (like is this guy alive or dead) if you don't track it consistently.

I heartily second this. I keep a google spreadsheet that has a master list of NPCs, with columns for their disposition toward the PCs, if the PCs have encountered them yet, where first met, etc. and of course a comments column pulling up the rear.  

I also recommend pulling the ideas of Fronts from Dungeon World for sandboxing.  It's specifically helpful when you want your NPCs/organizations to be active and progressing, whether the players are involved or not, but don't want to try to play it out in minutiae on the backstage.

I've got a few blog links of my own bearing on the subject:

Opening Gambit: Your town and its NPCs
Faith Manages: Designing religions
Setting The Table: Party composition and equipment
The Appetizer Round: Tips on portraying NPCs
The Main Course: Your First Adventure
The Dessert Round: Random tips and suggestions

Medieval Demographics Done RIGHT (Pt II)
Need a quick scenario? Stuff You Can Use

... a few extra thoughts.

* First off, you know your players.  We don't.  We have no idea whether they're used to being led around by the nose, and how proactive they're likely to be in chasing leads.  Our advice on the subject is going to be dependent on what problems crop up.  But in the end, the bottom line is whether everyone's having fun or not.  There's no Holy Ordinance of Sandboxing which decrees that campaigns have to produce X number of plot hooks, for instance; the "proper" number is the number your players find interesting.

* Secondly, you know you; we don't.  Now me, I game in a relatively cramped area where all the players use laptops, and my scope for electronics of my own is nil, even if I didn't prefer printed material on dead trees.  I've always disliked GM "shields" as creating an artificial barrier between me and the players, besides which I game from my living room, and am in my comfortable armchair; it wouldn't be ergonomic anyway.  So where I stick my random and not-so-random tables is in a three-ring binder, close to hand.

But that's me.  You've got a much better handle on your comfort zone than we do.

* NPCs: following up on Opaopajr's comment, something that bugs the hell out of me in many commercial RPG products is their mania for statting out not only movers and shakers, but to give full stat blocks for local businessmen.  While I linked to my blog post on the subject above, I think this bit's worth emphasizing:

QuoteBusinesses?  Well, you've got my previous article on town building.  Write a paragraph or two on each.  Here's an example from a small village in my campaign:

Sign of the Red and Blue Pot: With the death of the previous owner, her last surviving relative by marriage, a foreigner, Kesem kin Swallowflame, has taken over this well-stocked general store, which has a good array of housewares, tools, bulk grain and provisions, and textiles.  While he is a decent enough merchant (-13, various scholarly subjects-14/15), he has been trained to a scholarly life and educated at a great university, and somewhat resents having to take a menial job in the countryside.   Postings for foreign philosophers are not plentiful, but Kesem still pours his meager profits –- he's wont to let customers run up a tab –- into books brought in from the capital, trying to keep up with new teachings and still hopeful of scholarly preference.

And there you have it.  What, no stat block?  No weapons skills?  No magical items?  Of course not.  The PCs aren't going to fight this guy, and we don't care what his Health or Move are, whether he has Climbing skill, or how much damage he can do if he clouts you over the head with that grain flail leaning up against the corner.  What they're going to want from him is to fill up their packs with smoked sausage and biscuit for their adventure into the forest, and if they find out he's a wannabe scholar, whether he can read that weird text they found.  What you're going to need from him is an insight into his personality so that you can play him effectively as a vivid NPC, and we can all see the image that arises: a fellow starting to show grey hairs, somewhat fussy, somewhat distracted, somewhat irritable, possibly dressed grander (if shabbier) than the village standard, always with his nose in a book, and excited only when travelers come through town with books to sell.

And heck ... if he does need to fight, a GM ought to be able to determine, very quickly, the combat stats for an average villager.  Using GURPS, average stats are 10, so if Kesem trains once a month with the village militia, he may well have ST 10, DX 10, HT 10, a Speed of 5.5, a Spear skill of 12, with (say) a leather jerkin for armor (DR 1), a Parry of 9, and does 1d-1 HT of damage with a successful thrust.  Those details, including the time it took me to type them, took me 35 seconds to work out.  So why not establish that as the standard if you need to work up the mook villagers for that large-scale bandit raid?  A strong villager?  ST 12, and that damage is 1d HT instead of 1d-1.  A nimble villager?  DX 12, and that Spear skill becomes -14, her Speed becomes 6, her Parry becomes 10, and suddenly she's a legitimate threat in a fight.  There.  That's all you need.

And that's also the sort of NPC for whom you want to work out a paragraph or two.  The players aren't going to be interacting with the Queen of Warwik (and indeed, it took eight years for anyone to meet the monarch); they're going to get jobs from local lords or the deputy Chancellor's spymaster.  The players aren't going to be interacting with the Commander of the capital city's militia; they're going to want to know about the lieutenant commanding the night watch of the district in which their inn is located.  They might never meet with the Maid In Black, the matriarch of the faith of the Night Goddess, over there in the grand basilica.  They will meet Mother Ginevra, the parish priestess at St. Taria's, kitty-cornered across from the aforementioned inn.

I have a Fantasy Demographics which is my take on the data that S John Ross used for Medieval Demographics.

Quote from: estar;897251I have a Fantasy Demographics which is my take on the data that S John Ross used for Medieval Demographics.

It's a slightly better list than Ross', but not by much, I'm afraid. My full-scale rebuttal of Ross is here, but the TL;DR version is that (a) the "Paris tax roll" list was quite incomplete, and (b) edited down by the 19th century compiler, who (c) never intended it as an actual business list, and which (d) relied on one of the most atypical cities extant for the numbers.

Your own list shows the danger of the plug-and-play method.  Seaports were dominated by maritime trades, and generally speaking about two-thirds of the businesses in any port town were related.  Using a river port a hundred river miles from the sea to gauge the numbers of sailors, shipwrights, chandlers and harbor pilots, even if you presume these were all guilded tax payers, which they of course weren't?

Quote from: Ravenswing;897316It's a slightly better list than Ross', but not by much, I'm afraid. My full-scale rebuttal of Ross is here, but the TL;DR version is that (a) the "Paris tax roll" list was quite incomplete, and (b) edited down by the 19th century compiler, who (c) never intended it as an actual business list, and which (d) relied on one of the most atypical cities extant for the numbers.

I agree the Paris tax roll stuff isn't much use. The population density stuff is good though.

What I've learned from my last three sandbox campaigns:

(1) When returning to the sandbox, start with a smaller one (e.g. duchy, regional, or town and environs).

(2) Make it dense with adventuring opportunities, big and small. Players should stumble upon them everywhere.

(3) Use a lot of randomisers, utilise weather, make the sandbox breathe.

ADDENDUM: estar's Majestic Wilderlands and Blackmarsh are a great resource for specific hex-encounters, so are his Points of Light game supplements (via Goodman Games). Old Judges Guild stuff of course (Wilderlands of High Fantasy, Fantastic Wilderlands Beyonde). I would highly recommend the Midkemia city supplement if you wanna run ad hoc town encounters, otherwise Vornheim has served me well.

Quote from: Nerzenjäger;897335What I've learned from my last three sandbox campaigns:

(1) When returning to the sandbox, start with a smaller one (e.g. duchy, regional, or town and environs).

(2) Make it dense with adventuring opportunities, big and small. Players should stumble upon them everywhere.

(3) Use a lot of randomisers, utilise weather, make the sandbox breathe.

ADDENDUM: estar's Majestic Wilderlands and Blackmarsh are a great resource for specific hex-encounters, so are his Points of Light game supplements (via Goodman Games). Old Judges Guild stuff of course (Wilderlands of High Fantasy, Fantastic Wilderlands Beyonde). I would highly recommend the Midkemia city supplement if you wanna run ad hoc town encounters, otherwise Vornheim has served me well.

That's all very true. You need a keyed map with adventure sites (at least two to start with), plus tables for procedural content generation - wandering monster especially. For weather I just roll a d6, 1=wetest, 6=dryest, the specifics depending on the season. The d6 is your friend.

Quote from: Nerzenjäger;897335I would highly recommend the Midkemia city supplement if you wanna run ad hoc town encounters

I would too, if you can get your hands on it.  It's not only keyed per wealth and type of district -- you'll get different selections in a poor industrial area than in the high rent district -- but they're not the goofy type of 70s urban encounters where all of a sudden a horde of monsters are loose on a quiet city street.

Quote from: Ravenswing;897316It's a slightly better list than Ross', but not by much, I'm afraid. My full-scale rebuttal of Ross is here, but the TL;DR version is that (a) the "Paris tax roll" list was quite incomplete, and (b) edited down by the 19th century compiler, who (c) never intended it as an actual business list, and which (d) relied on one of the most atypical cities extant for the numbers.

Your own list shows the danger of the plug-and-play method.  Seaports were dominated by maritime trades, and generally speaking about two-thirds of the businesses in any port town were related.  Using a river port a hundred river miles from the sea to gauge the numbers of sailors, shipwrights, chandlers and harbor pilots, even if you presume these were all guilded tax payers, which they of course weren't?

I am well aware of the issues involved. Hence the title Fantasy Demographics. Within the document I pointed out that my categories were quite arbitrary and designed to be gamable within a fantasy campaign. The goal to make something that within the ballpark and usable for the average referee. Not that it is historically accurate, otherwise I would have done was S John Ross had done is present it a reflection of medieval life.

Also I read your rebuttal post when you pointed it out the last time and obtained some of the books on the list. While interesting to read, I like reading history books, it not straightforward as to how you get the numbers you need from it. You have to do some extensive scholarship and even then make more than a few judgement call to make it usable for what is a leisure activity.

Even then, all the historical numbers point to a higher population at a higher density than your typical fantasy campaign. While accurate, the result in my opinion is not very gamable for a typical D&Dish fantasy campaign. It needs to be ratcheted down by a least an order of magnitude. This way you get area of dense civilization but there is plenty of wilderness and the actual number of high status individual (feudal nobles, merchant magnates, etc) is more managable.

This is why I wrote Fantasy Demographics, the catagories that S John Ross used in my opinion were not well suited for a typical fantasy campaign. So I used the categories, I developed for my Majestic Wilderlands and recalculated the numbers.

In my rough draft of "How to make a Fantasy Sandbox." I have the same list of professions  but now there are multiple lists to account for specific circumstances. The numbers have been arbitrarily shuffled to reflect my sense of how a seaport differs from a rive town which differs from a town on the Silk Road, etc, etc. This is based on what I read in my history books including the one I read from your lists.

If I was writing for a game whose appeal rest on historical accuracy then your criticism is on point. However what being done is to be make thing that is consistent, in the ballpark, and above usable by the average referee.

Quote from: Ravenswing;897359I would too, if you can get your hands on it.  It's not only keyed per wealth and type of district -- you'll get different selections in a poor industrial area than in the high rent district -- but they're not the goofy type of 70s urban encounters where all of a sudden a horde of monsters are loose on a quiet city street.

For who want the Midkemia city supplement you can get it here in PDF form.

However I STRONGLY recommend getting the app (android, mac, windows). There are a ton of tables and not well suited for use during a tabletop session. If you don't want the app, then just roll up 50 encounters for each categories before.  However it because of all the tables that it is so good. It really adds something to trying to referee the life of a fantasy city.

It also has a village generation system and a excellent set of rules for managing down-time between adventures.

My advice is a GM should talk with their players about a sandbox campaign beforehand and get the players buy-in to the concept. I believe that would alleviate much of the problems with sandbox campaigns upfront.

The biggest issue I have found in sandbox campaigns is that players are required to be proactive for the campaign to work, and I find that most players are not very proactive.

Instead of a pure sandbox, I prefer to run "theme" sandboxes to create motivation for the PCs.

My Mazes & Minotaurs campaign was a sandbox in an archipelago, AKA islands instead of hexes. The premise was Athens was suddenly frozen in time, every living thing was frozen in stasis, as was anything entering the city. Even the gods were dumbstruck until the Oracle of Delphi prophesied the cure for Athens would be found in vanished archipelago which Neptune confirmed had suddenly appeared. The heroes were tasked by the gods to search these mystery islands and free Athens from its terrible curse.

Thus, I gave the players a week-to-week motivation (save Athens) while giving them full autonomy to cruise around the islands (hexcrawl) picking up clues and going on various side adventures. Replacement PCs were easy as the heroes came along with several support ships where new fresh PCs could be plucked. Other heroes from across Greece also heeded the gods call as potential allies and of course, there were evil Greeks who wanted Athens to stay in stasis so their cities would rise in prominence who sent their own assassins to stop our heroes.
 
In the 90s, I ran a CoC  campaign where a bunch of PCs suffered nightmares of an Old One destroying New York on the dawn of the Winter Equinox. The PCs meet each other in the Arkham Sanitarium because of their nightmares drove them temporarily insane, and now they have sworn to sacrifice everything to save New York from destruction. I didn't provide any direction for them, just the concept and let them do whatever and go wherever. But that was a proactive player group.

Quote from: Spinachcat;897417The biggest issue I have found in sandbox campaigns is that players are required to be proactive for the campaign to work, and I find that most players are not very proactive.

That's why I run a pure sandbox with an overabundance of adventuring opportunities, where they almost can't escape the open threads laid out everywhere. When running it for players without sandbox experience, this makes it seem like some of them matter more than others, just like in a more linear approach. In reality, they really don't. What I then often find is, that the players usually want to go back to pursue other tasks they missed out on because they went after something of more immediate interest. Though of course that's not to say that the pure sandbox can't have NPCs with enough power and an agenda that dwarves other happenings in terms of impact.

Quote from: estarHowever I STRONGLY recommend getting the app (android, mac, windows). There are a ton of tables and not well suited for use during a tabletop session. If you don't want the app, then just roll up 50 encounters for each categories before. However it because of all the tables that it is so good. It really adds something to trying to referee the life of a fantasy city.

Again, estar to the rescue! I had no idea there's an app for that. Thank you. I bought the PDF years ago and it's good enough, but the app makes using that thing way easier.