Sandboxing - story before, story now, story after

[Callan S.]

Unrelatedly, our experience with the Keep on the Borderlands suggests that sometimes, all of the choices on the menu are unappetizing.  "Ugh, not an endless menu of poverty-stricken humanoids and bickering with the very uptight denizens of the Keep.  Let's get out of here."  So it's sometimes helpful to include an escape hatch, where you can jump one layer out and find some other sandbox to muck around in.

Doesn't this come across as an incredibly uninspiring basis on which to write material/create content? Just in case someone goes "Ugh", you have to write out a whole new sandbox? And maybe they'll go Ugh to that as well, so you need another one or if you didn't, somehow you didn't provide?

I read that and think about the linear plots I've been run through by GMs, and for me, the point of a sandbox is to create an adventure that's a lot harder to "break". I'm sure a lot of us have been there - - we prep something, we sit down with some players, and then at some point during play, we all can sense the ... limits of pre-planned adventure.

What strikes me is that say, no one considers being a sorcerer, or a young zealot with a gun and a funny blanket on their back as some constricting limit. Presumably because it's written on the tin. While, although ostensibly the GM is supposed to go with his imagination and be creatively free as the wind, really he is incredibly constricted by any individual players sense of "this game isn't about that", which comes up as "Oh, we can feel a constricting limit here!" in the player. I mean, this even goes for D&D - why are dungeons not considered a constricting limit? Well, it's written on the tin. To me it points to the importance of indicating clearly, in advance, what things were doing now, with these people, in exclusion to anything else.

[James_Nostack]

Unrelatedly, our experience with the Keep on the Borderlands suggests that sometimes, all of the choices on the menu are unappetizing.  "Ugh, not an endless menu of poverty-stricken humanoids and bickering with the very uptight denizens of the Keep.  Let's get out of here."  So it's sometimes helpful to include an escape hatch, where you can jump one layer out and find some other sandbox to muck around in.

Doesn't this come across as an incredibly uninspiring basis on which to write material/create content? Just in case someone goes "Ugh", you have to write out a whole new sandbox? And maybe they'll go Ugh to that as well, so you need another one or if you didn't, somehow you didn't provide?

In my experience this has never been a problem.  Early-edition Dungeons & Dragons was designed in such a way that creating new content--dungeons, monsters, spells, NPC's, whatever--is dirt-simple and extremely low-investment.  I'm typically noodling away on some side project in development.  If the players say, "Eh, let's try our luck elsewhere," it gives me an excuse to finish it and parade it out.  Alternately, at this point there are literally hundreds of one-page adventures not to mention dozens of classic modules.  And all of that assumes that the players are totally passive, instead of actively scheming to do new things in the setting which lead to new hijinks.  Coming up with new menu items has been trivially easy. 

[contracycle]

Well yeah, but that has the penalty of being disjointed and thematically contradictory.  I mean, "sandbox" surely doesn't have to mean a randomised collection of junk you happen to have laying about.  It would be a bit funny to be playing GTA and to find a mission to raid a kobold lair. So although there has been a history of unrelated content (vis. Barrier Peaks
) and genre-busting (gunpowder, dinosaurs), even getting slightly more coherent means that poetneital content has to be sifted and selected.  So I don't think  that it has been "trivially easy" at all; there may be a lot of stuff out there, but it may be that only a miniscule proportion of it is usable, and all too often the anything goes attitude has produced a lot of trash.

[James_Nostack]

Gareth, all I can tell you is about my own experience.  If you look at those modules, and at the one-pagers I mentioned, a whole big heaping helping of them assume a vanilla fantasy landscape.  Some of the ones that don't--X1: Isle of Dread and X8: Drums on Fire Mountain plus one of the 2010 One-Page Entries Den of Villainy plus D3: Vault of the Drow plus B4: The Lost City plus C3: The Hidden Shrine of Tamoachan all fit very well together if you wanted to do, say, a tropical island pulp adventure with ziggarauts built by a vanished, decadent civilization.  But you can harmonize these with a vanilla fantasy world if, like Tavis's White Sandbox game, your setting explicitly includes everything that's ever existed in D&D. 

I agree that you've got to sort through stuff, and occasionally rename some god or nation-state.  But this is very low-impact work.

[Callan S.]

In my experience this has never been a problem.  Early-edition Dungeons & Dragons was designed in such a way that creating new content--dungeons, monsters, spells, NPC's, whatever--is dirt-simple and extremely low-investment.

I have a strong instinct. And being an instinct, it could simply be firing in the wrong direction. But my instinct is that if it's dirt simple, then what you get is dirt simple? I mean, I can grab an old crossword and mark off where monsters are in the 'tunnels', sure. But it seems like alot of dice rolling and your done? And not necessarily with any sense of forfilling a win condition (I can, for example, play snakes and ladders just fine (which is alot of dice rolling), I feel some fun there, even though it's simple). I guess in terms of story after, ie picking the highlights of a whole session, it could fit with that?

[James_Nostack]

Callan,
To be polite: the complexity in a sandbox game doesn't come from GM prep. You are killing monsters in tunnels.  It comes from players' hijinks, both in the tunnels and in-between delves as the world evolves.  This is a story of people killing things in tunnels and what happens in between.

I would remind you that, as stated in my first post, I have no interest in discussing counter-factuals based on instincts and analogies where they directly contradict my stated personal experience.  If others wish to do so, they are welcome. 

I would, however, love for you to discuss your own actual experience of sandbox play, or your own history of non-sandbox play with "early RPG" systems (i.e., D&D, Traveller, something of that vintage).  Or doing this with your actual play of indie games--I'm thinking that there's nothing exactly analogous on the indie scene, but Trollbabe and Apocalypse World come fairly close with some significant differences.  Alternately, given your experience with D&D 3.5 and D&D 4e, whether those games facilitate this style of play, and if not, why not.  The 4e "Points of Light" model, which many people were raving about in late 2008, has at least superficial elements of sandbox play: a large, hostile, perhaps unexplored wilderness, and players with hometown to which they have their own attachments.  The Nentir Vale, for example, strikes me as a perfectly adequate starting point for a sandbox, though I'd be curious if anyone plays it in an entirely player-directed way.  (The design of H2: Thunderspire Labyrinth seems to have some elements of this as well, but again, heavily shepherded through the GM's plot, at least as written.) 

I think there's a lot to discuss here that involves actual play, rather than what must be true a priori, and I'm eager to have that discussion.

[contracycle]

Fair enough, but I just want to say this: that sample game you;ve got there epitomises everything I grew to hate about D&D.  You mention above that inability to go off the railroad prevented you from entering Exploration, in another game, but to me, introducing characters called Clamato and Chowder would have kicked me the hell out of exploration.

Now the point is this: none of this is inherently necessary to the sandbox style of play.  As the Gamasutra article you linked pointed out, sandboxing is often used as an excuse to skimp on high level design, while in fact it benefits greatly from strong high level design.  I think that was a big part of the development of the Ravenloft setting; it was still sandboxy, but it had a much stronger aesthetic running through it, established expectations more firmly.

I don't think "sandbox" should be conflated with "anything goes".  That isn't necessarily true.  I don't think it is true to say that complexity does not, in an absolute sense, arise from GM prep, and is all down to the "player's hijinks".  That stuff is functional in that yes, you can get people to play it.  If that floats your boat, cool.  But it also produced the sub-Renfair culture of Python jokes and fratboy socialisation which many people found alienating and dull.

I'm wary of presenting this sort of thing as a good representation of sandboxy play.  B2:KOTB is actually thematically very tight; a semi-random selection of modules and jokey monsters is not.

[James_Nostack]

Gareth wrote:

I just want to say this: that sample game you;ve got there epitomises everything I grew to hate about D&D.  You mention above that inability to go off the railroad prevented you from entering Exploration, in another game, but to me, introducing characters called Clamato and Chowder would have kicked me the hell out of exploration.

Gareth, as you'd see, those characters left the Glantri game almost immediately.  Eric, the DM, has told people, "You cannot name your character ________."  I have some reservations about this approach: early edition D&D makes it extremely perilous to take your character too seriously, because they can die in an instant. 

Now the point is this: none of this is inherently necessary to the sandbox style of play.  As the Gamasutra article you linked pointed out, sandboxing is often used as an excuse to skimp on high level design, while in fact it benefits greatly from strong high level design.  I think that was a big part of the development of the Ravenloft setting; it was still sandboxy, but it had a much stronger aesthetic running through it, established expectations more firmly.

If I've given that impression, it wasn't intentional.  I'm not saying design isn't important!  I was explicitly responding to Callan's assertion that it would be difficult or aggravating to generate new material when the players get bored with their current options.  It's super-duper-easy to generate stopgap stuff on your own in keeping with the aesthetic you established, or failing that to grab something off the shelf.  You asserted that stuff from off the shelf isn't easily adapted.  Well: so far it hasn't been too hard to adapt to aesthetics.  And then you asserted that things need to either be designed in keeping with the aesthetic or adapted.  I'm not sure what we're arguing about. 

None of this is to say that sandboxes or the dungeons within them must be, or should be, simple affairs.  There's a huge role for design if you want to design.  I've got, like, 45 pages of hardcore design notes on a new project.  High-level dominion play (operating your own kingdom, etc.) is largely hand-waved in most early versions of D&D.  What I'm mucking away on is an attempt to do that--starting out at 300,000 XP.  So typical things involve getting a land grant to explore and conquer unexplored territory, building a castle (subcontracting with Dwarves, finding them a supply of adamantite, dealing with a territorial dragon).  Managing a Thieves Guild war.  Researching magical items.  Fielding armies.  All that high-level stuff that is, at our present rate of advancement, decades away.  Thinking about these types of adventures requires back-designing a world where that kind of thing is meaningful and fun--so thinking about cultures, and how Detect Evil might get used in an Inquisition, a Lawful Evil Elven aristocrat who uses good-natured but blindly obedient Blink Dogs as his assassins, that kind of stuff. 

What you want to think about also is: geographical proximity,.  Do you want to keep some stuff further away until later?  What if you want to put a level in a nearby dungeon that you'll very likely only learn about from some adventurer's notebook squirreled away in some dungeon many miles away?  (This is a bad example of "connective tissue" for several reasons, but hopefully gets the point across.) 

So designing a sandbox can be a huge pain, but doesn't have to be.  (I'm doing it because I find it an entertaining set of design puzzles to solve in its own right, not because it's necessary to do it this elaborately.)  Remember: we're only talking about plucking (not-so-)random things off the internet because of a chain of nested hypotheticals about what do you do if players get bored of one part of the campaign.

I don't think it is true to say that complexity does not, in an absolute sense, arise from GM prep, and is all down to the "player's hijinks".  That stuff is functional in that yes, you can get people to play it.  If that floats your boat, cool.  But it also produced the sub-Renfair culture of Python jokes and fratboy socialisation which many people found alienating and dull.

I should have been more clear about "players hijinks." 

In Tavis's White Sandbox game, our favorite Magic-User PC, Maldoor, got killed.  Specifically: while leaving a dungeon we had a wandering encounter with a group of NPC's.  One of them polymorphed into a gargantuan Purple Worm, swallowed Maldoor, and digested him utterly.  (This is almost the very best thing you can do with a polymorph spell in 0e.  Maldoor became worm-poop--"super-killed" because you can't get resurrected from that, all due to the simple act of rolling a bunch of dice while bumping into monsters in a crossword puzzle. 

Maldoor's fate became a rallying cry for the rest of the party, because our next-best Magic-User is an inept doofus (played by me), and we really needed the guy back.  So we spent a lot of time in-game and on-line researching where Maldoor's soul had ended up, and what in the world we might do to retrieve it.  One of our Elves scored some fantasy-LSD to take us into the Astral Plane, and another researched an ancient goddess of dreams, introducing a new religion into the campaign world.  We'd rescued an extra-dimensional demi-urge several sessions ago, who helped us construct a return-portal once we found Maldoor's soul.  Along the way through the various planes, we defeated a couple of demons (which we'd never done before!) and I got humiliated in a riddle contest by a pair of Shedu (you have made a powerless enemy today, Shedu!) 

But more importantly, in the process of retrieving Maldoor, my inept Magic-User gradually assumed increasing responsibility in the party, and as a role-playing experience I ended up showing how he grew into the role of the primary Wizard--putting aside his usual scrambling to make "bad" spells useful, and going with raw power.  Meanwhile when we ran into some Astral Dragon Riders or something guarding Maldoor's soul, our Paladin leader was the only one who would joust them without cheating--and lost.  The conditions of the bet were that he would lose all of his magical items. 

So in the process of overcoming a bad roll of the dice, we (a) traveled to dimensions we'd never been before, (b) revived an ancient religion, (c) overcame extremely dangerous foes, (d) I made some lasting enemes, (e) I got to role-play a significant shift in my character's personality, (f) our "good guy" suffered serious consequences along the way. So the campaign was vastly enriched by players' response to a single unlucky die roll.  If the Purple Worm had rolled a 13 instead of a 14, Maldoor could have been raised by conventional means and none of this would have happened.

[James_Nostack]

And then in the very next session Maldoor got backstabbed by an invisible Giantess and died.

[Caldis]

My experience with those old D&D modules is markedly different.  I dont think it would count as sandboxy more like episodic, we have a bunch of characters they're all ready to go let's go hit this adventure, I think it was pretty much up to the DM which adventure we'd be going through but I sometimes remember a list of options and a kind of vote taking place.

A bit of background first.  We were pretty young when D&D came out, most of our groups born in 1969 or 1970.  I was first introduced at a friends place when a slightly older group of guys were running a game (one of the Against the Giants series of modules in 1980 or so?) handed a premade dwarven multiclass character and pretty much followed along.  Never played regularly with those guys again but we had enough fun that we started doing our own thing and picking up the books as they showed up on local store shelves, wasnt much at first but it became a lot soon enough.

So episodic adventures with whoever we could get to play usually about 4-6 players including the DM.  I dont think anybody really loved they idea of DMing and saw it more of a necessary evil so we tended to swap out the dm fairly frequently.  There was nothing in the way of an ongoing world ar at least nothing that had any relevance to play.   We played straight through an adventure until we completed it.  We rarely left a dungeon to resupply or anything, we'd sleep to heal setup camp and watches and the dm would roll for wandering monsters, which would provoke a fight and possibly more rest afterwards but was rarely deadly.   The C series of modules were pretty big for us and didnt really give you an option of running away and coming back later, Ghost tower of Inverness and Hidden Shrine of Tamoachan were a couple early ones that we liked.  

We definitely didnt play the opposition as very smart.  The wandering monsters tended to stumble on to the party and just start fighting, no going back and getting reinforcements or anything like that (not sure that any modules made suggestions to do so).

Rather than learning the game from any specific module (I'm sure at one point I read most of that stuff in Keep on the Borderlands) we had the DM's Guide and the Players handbook, various Dragon Magazines, and a hodge podge of modules many of them for higher level characters.  I think this clutter had a big influence on our play style.  For one the DM's guide had the various methods of character generation.  Roll 4d6 drop the lowest was popular, when unearthed Arcana came out and the gonzo dice rolling method it brought came out things went mad.  Dragon magazines had a lot of nifty ideas and a lot of them added power to characters, making them better at surviving.  Magic users were always a bit wonky with their spell book restrictions and that table in the dm's guide for what spells were in their book to start.

We definitely didnt play as adversarial a game as the new OSR is advocating.  I dont think we were missing anything rather I think a mix of influences was swirling about that made the D&D scene at the time rather confusing.   The two modules that really stand out as examples of the sandbox play are B2 Keep on the Borderlands and T1 the Village of Hommlet but they were just a couple in a swarm of different influences.  

[C. Edwards]

James wrote:

...while leaving a dungeon we had a wandering encounter with a group of NPC's.

Encounter tables are important in so many ways. For one, and this goes to some of Callan's concerns about prep time and players deciding to do something else, they can buy the GM a lot of time. They can increase the play time it takes for the PCs to reach their next destination, giving the GM more time for at least minimal prep during play or delaying the journey long enough so that the next destination can be readied between sessions. They're also random injections of complexity and can end up generating whole swaths of play all on their own. James' example is a great showcase of that. You never know just how much complexity a random encounter will spawn until after it happens.

Callan wrote:

But my instinct is that if it's dirt simple, then what you get is dirt simple? I mean, I can grab an old crossword and mark off where monsters are in the 'tunnels', sure. But it seems like alot of dice rolling and your done?

While things may look simple on paper, the real complexity is injected when you add in people. The GM's contributions and the players' contributions turn two dimensions into three. This is pretty much how most roleplaying games I'm aware of function in play. If the players approach play in a non-ironic manner and the GM is doing the whole "make the world seem real" thing that GMs do, then "a few monsters in a tunnel" becomes pregnant with possibilities.

James wrote:

And then in the very next session Maldoor got backstabbed by an invisible Giantess and died.

Sounds like Maldoor was the weakest link.

[contracycle]

If I've given that impression, it wasn't intentional.  I'm not saying design isn't important!  I was explicitly responding to Callan's assertion that it would be difficult or aggravating to generate new material when the players get bored with their current options.  It's super-duper-easy to generate stopgap stuff on your own in keeping with the aesthetic you established, or failing that to grab something off the shelf.  You asserted that stuff from off the shelf isn't easily adapted.  Well: so far it hasn't been too hard to adapt to aesthetics.  And then you asserted that things need to either be designed in keeping with the aesthetic or adapted.  I'm not sure what we're arguing about. 

Well, I'm disagreeing that it is super duper easy, except where the kitchen sink aesthetic applies.  Now of course there's no accounting for taste, but there is very little useful to my taste (social realism) out there.  The form to date has not IMO produced a wide range of aesthetics, it's just had a fairly consistent junkyard one; it looks easy only becuase its not being seriously attempted.

Now as you say, we're only talking about this because of a hypothetical that was raised, but I came in on that to make a slightly different specific point.  The basis of the thread was that the sandbox hasn't been sufficiently defined.  I'm suggesting is that part of the reason for that lack of definition is because of the junkyard aesthetic, which has operated in much the same way as the "golden rule" to muddy the waters.  I mean the very fact that someone could write and print and sell a module like Barrier Peaks, with its spaceship elements, shooting the hell out of any kind of pseudo-historical consistency, speaks volumes - and yet this is a bona fide Gygax module, introduced at a convention.

So this junkyard thing has always been about, has some authoritative basis, and so on.  I'm suggesting think that to define sandbox more precisely the junkyard aesthetic has to be separated out from the method of sandbox play, and not assumed to be folded into it, as it often is.

[Tav_Behemoth]

Contracycle, my experience as the GM of the White Sandbox is that the gonzo Acid Fantasy kitchen-sink approach James describes is indeed a necessary part of the method of sandbox play. When a player says "I want to be 23, a robot cleric who gets his spells from worshipping the Server, a vast computer at the center of the earth", for me to say yes to this - the aesthetic decision that anything can potentially fit into the setting - is part and parcel of me saying yes to the idea that if you go far enough down into a dungeon, you might be able to douse the Server's memory banks with flaming oil - the sandbox method that leaving the edges of everything undefined and having robust tools for creating what's there procedurally provides infinite scope to handle player-driven exploration.

As a player in the Glantri campaign, when I've been allowed to name a cleric the Boss and have him inspire his own religion, the sandbox sings for me - even though the result is essentially silly, I enjoy taking it seriously, and the fact that it runs contrary to the tone of the game is part of the point: having GM Eric provide seriousness and drama is a necessary foil for our comic relief. When other players have been told that they can't name a character Sosexia, this to me feels like my agency and scope of action in the sandbox is limited; there aren't rails exactly, but now there are walls on the sandbox that will keep us from ever traveling to wherever Sosexia comes from and learning about a land where that's a normal name.

Acid fantasy is to my tastes, for sure, but I take issue with the idea that the only alternative aesthetic to social realism or pseudo-historical consistency is a dog's barf lowest-common denominator. A lot of work in the OSR and the Forge alike has gone into establishing how the original swords and sorcery canon makes a crashed spaceship a valid and exciting aesthetic vehicle for exploration; count the number of works in the reading guides of AD&D's Appendix N or Edward's Sword and Sorcerer where this would fit in just fine. Personally neither Dernyi Rising/Harn realism nor Shannara/3E D&D vanilla fantasy are to my taste, but factually both these movements come later in genre fiction and gaming than the boundaryless mix of the fantastic that preceded them.

But I totally agree with you about encounter tables and players adding the third dimension. I just disagree that it's a problem if the players start out with an ironic distance; I think a sandbox works best when both the goofiest thing the players come up with and the weirdest flumph in the canon of the game you're playing are part of what the GM is mandated to make seem real. As James says, characters that survive will inevitably be taken seriously; the trick is to encourage player input so they can make characters whose survival they start out caring for.

[Tav_Behemoth]

Oops, that was C. Edwards I was agreeing with about encounter tables and disagreeing with about irony.

On strong high level design: I'm writing at the Autarch blog about all the work that goes into designing random tables for wandering monsters and random treasures and procedural dungeon stocking, and our lead designer is are posting about designing the economic frameworks within which these make sense, and yeah it's a lot of work to create a new game designed to make it easy to do sandboxing because this work is done for the GM. (OD&D makes it easy for me to run the White Sandbox because it represents three+ years of playtesting such a framework by Arneson's group, plus Gygax adding a layer of systematization of the game mechanics that makes it easier to run and a layer of prismatic obfuscation of its intent and vision that makes it easier to customize.)

A sandbox certainly benefits from additional work by the GM, even if it doesn't require it. The random encounter table you make for your dungeon is automatically going to be better than mine because it's meaningful to your campaign; both are going to look like crosswords to an outsider, just like a musical score looks like gibberish to me because I don't read music.

[James_Nostack]

Hey Tavis, thanks for dropping by.  If you feel like it, would you care to explain how you set up the White Sandbox in its early days, and how you've been "refreshing" that environment between play?  I know one input is our carousing results.  I suspect you've also got a bunch of random charts describing what dungeon-denizens do next.  I know that as a player, sometimes I got frustrated because it seemed like the Caverns of Thracia changed so much between delves that essentially our previous recon missions had been for nothing.  Was that just bad luck of the dice?  Did you trim back major changes later?  Were you deliberately offering some resistance at that point so that we would have a delicious feeling of frustration?  Could loudly voiced, and widely shared, player dissatisfaction cause you to override the dice? 

I guess I'm wondering about how you update the sandbox from a procedural standpoint, even if that standpoint is simply, "I squint at this NPC and ask myself, 'What would he like to do, and how much of it can he reasonably accomplish by next session?'"