Monday, March 15, 2010

Conan Knew More About Cimmeria Than Howard Did

Or: Notes, Assumptions, and Conclusions About Travel In A Sandbox

-My preference: If PCs are traveling overland from one place to some other distant place, and I want to have something unexpected happen on the way there, the thing that happens should usually have something to do with the decisions the PCs made about how to get there. (Method of transportation, route, what they brought with them, etc.) Otherwise I feel like it's not as much of a sandbox. i.e. If they take the west route and so I roll on the Random Wilderness Encounter Table or I take the East route and so I roll on the Random Wilderness Encounter table, that's not as fun or interesting as making what happens a consequence of a meaningful strategic decision.

-My particular players must, therefore, be reminded (at least at this stage) that their PCs have a choice as to how exactly they get where they're going.

-The easiest way to do this is to draw a simple map demonstrating what the PCs know about the area between where they are and where they're going every time they take a trip.

-Often, what the PCs know is very little, but there should be enough on the map to make at least two different routes seem viable and, in some way, different in character. i.e. the possible routes should make it seem like different kinds of challenges will be encountered there.

-A ranger should know more than other PCs and thus have more landmarks on this map or know a few more tidbits about what locations on the map are like or are rumored to be like. As should PCs who have some reason to have special background local knowledge.

-After a while, every time the PCs are presented with such a map, they will instantly know that they are in for some kind of trouble in going from A to B, and this may change their decision about how to proceed.

-To avoid this, I should probably provide a map every single time the PCs go from any A to any B, even if nothing actually is going to happen when going from A to B.

-Assuming all these travel maps must, therefore, have at least two different reasonable routes (thus providing choice) then that means every time the PCs take any overland trip, they must pass at--at a minimum--one landmark or change in terrain or geographical feature. i.e., the minimum choice a map could provide is "Do you cut through the Swamp of Many Bones or just take the road?"

-Consequently, the campaign world as a whole will eventually be covered in these "in-between" features that may or may not have any actual content behind them.

-Buildings are more enticing to my PCs than geographical areas. Like if I say "You pass the Tower of 2000 Sorrows on the left and the Forest of 2000 Sorrows on the right," the PCs will probably want to investigate the tower either immediately or later but give the swamp a wide berth.

-I assume an average movement rate of 20 miles per day riding through snow-covered forest (it's mostly snow-covered forest) , half that walking, 15 miles per day in icy swamps or snow-covered mountains, and 30 miles per day in open country if it ever stops snowing. Yeah, I know Napoleon's army marched 30 miles a day, but his army wasn't made up of Str 7 rogues and clerics carrying half their weight in unidentified potions.

-Excluding island nations, European capitals are generally at least 500 miles apart. This is a base number for figuring how long you'd have to travel if moving from the center of some empire to the capital of some other empire. Water, mountain ranges, forests composed entirely of carnivorous trees or other major geographical impediments would generally have to intervene if the distance is any shorter.

-This means getting from a big city to a really different big city takes a long fucking time, in game days, or at least requires crossing some famously dangerous or inconvenient geographic obstacle.

-The distance from Ghent to Antwerp is about 4o miles, from London to Birmingham is over 100 miles, from Rome to Naples is like 140 miles. So going from a big city to the next biggish city where they speak whatever the locals call "common" should take from 2 to 5 days, assuming my alles-ist-frozendoomforest movement rate of 20 mpd.

-Castles are more densely distributed than cities, of course. Depending on what you consider a "castle" and what source you consult, a heavily fortified country like England had a medieval density of like one castle every 6-10 miles, or two a day at an ordinary ride.

-Deciding to travel a long distance should feel to the PCs like a more meaningful decision than deciding to travel a short distance. The method can be direct "Arr, tis a long and dangerous road to Argenvoth!" or indirect, but at some point the girls should start to get it. Aside from making it clear the amount of supplies this'll require, the "rhythm" of the number of encounters between areas (whether planned or random) should begin to feel like it's related to the distance traveled.

-Mystery is nice and good but the more information they have about the gameworld, the more it's actually a sandbox. If, in all directions, the territory is equally unknown, then there's no reason to make any given decision. This may be sort of a Duh type conclusion, but the idea that:

+The gameworld is defined as the players explore it, and the idea that

+The players need information in order to make meaningful decisions about which way to go

are actually two opposing ideas. There needs to be a mix or compromise. If I only include new places when the PCs explore them then they have no basis to decide to go anywhere, if I have the whole world planned in advance, then I'm possibly doing a lot of work on places that never get used when I could be adding depth to areas (geographical or otherwise) that the PCs are actually interested in. So there needs to be a mix.

-That is: the "pulp" method--by which an author and reader learn about an internally consistent world by means of a series of "short stories", is actually somewhat contradicted by the "free, open sandbox" method by which presumably the characters in the pulp stories make decisions, which suggests that the PCs have enough information to decide where they wanna go. They are reconcilable, but it's a balance.

-Or, to put it another way: Conan knew more about Cimmeria than Howard did. Howard knew whatever he needed to write a story where Conan wakes up in jail in Lunaticwizardia--but Conan knew enough facts, rumors, legends, and general cultural bs to make a decision to go toward Lunaticwizardia instead of Safeboringium in the first place. In D&D, for the players to make Conan-like decisions about where to go, they must have more information about destinations they may never go to than I am used to providing.

-Howard was like a reality-TV crew dropping in on Conan when his life got interesting. Conan would've needed to know far more in order to get into the interesting situation in the first place.

-(Other people get around this by relying on published settings--or even real medieval europe. I'm not going that way.)

-Choice should be more detailed and interesting than: "go after the only plot hook dangled or just refuse to and look for rich people to rob instead until the GM improvs something out of it".

-On my personal group, three things to remember:

1) Many players will forget any plot hook or piece of local information a week later if they haven't immediately done anything with it,

2) Some players won't forget this stuff but will miss the next session and so will be limited in their ability to act on information they consider interesting,

3) That's ok.

-In a sandbox, Freedom = Information.

-In order to get and remember information, players must be either hardcore or be constantly reminded of the information.

-The players are not, generally, hardcore. At least not in the information-gathering sense. Except Mandy.

-A mechanism to constantly remind players of the information they have should be adopted if the sandbox style of play is truly desirable.

-A wiki, mass-e-mail, or info-folder will be useless with this group.

-A big-ass map with most macro-level information thus far acquired on the gameworld on the wall is probably the simplest way.

-Integrating player-generated info about where exactly these wood-elf parents she has came from into this map is desirable if this is really going to be a sandbox.

-Freedom depends on constantly reminding people they have freedom.


Jeff Rients said...

I decided to use an actual wargame board for my next campaign map so I can put it on the table. I'd rather give out more information than the PCs could easily access than less.

Snarls-at-Fleas said...

Wow. The last phrase...

Freedom depends on constantly reminding people they have freedom.

Pure win. And not only in RPGs. :)

Anonymous said...

Maybe a little comic/manga style recap of the previous adventure should be provided each session. It only means more work for you. Your players are worth it, right?

Telecanter said...

+The gameworld is defined as the players explore it, and the idea that

+The players need information in order to make meaningful decisions about which way to go

are actually two opposing ideas.

Yes. This stuff is hard, but when I read posts like yours I get excited that using the internet we might be able to assemble some kind of deep, practical DMing canon. Thanks.

Rod said...

I'm trying out a list of "Outstanding Bounties" in a Traveller game. Four or five little situations developing in different locations, with bounties for the PC to go there and accomplish a certain task. Which is not to say that "there" only exists for a mission to be there -- rather, the reward gives players something to zero their compass on, so to speak, when they explore that location. That's the theory anyway.

I guess the D&D equivalent would be, in addition to or instead of bounties to hunt, rumors of intriguing-sounding treasures in intriguing-sounding places.

Kevin said...

You may have seen this already, but Penny Arcade's Gabe is doing his D&D game as a map-driven sandbox. I haven't thought too carefully about the mechanics of the Resolve system he describes, but I love the idea of dedicating a permanent hex-mat to a game world's map.

OtspIII said...

This is something I've been thinking about a lot lately, and it's something I've come to similar conclusions with but somewhat different fixes for.

I really like sandboxes in which the PCs don't really know anything about the setting they've come into. It keeps the important parts of the setting as being based on in-play experiences rather than pre-written setting materials, which I really prefer since I don't want to make reading either my or a publisher's bad setting fiction a prerequisite towards understanding what's going on. This does have the big downside of leaving the players initially without much meaningful information to use to make meaningful choices.

For the first session or two I don't really mind this, since that's the time you're almost more exploring your character/the GM's style/the party dynamic/etc than the game world, but it's pretty bad once you start trying to set out and get things done. My solution is to build a Rumor system.

Basically, whenever the PCs spend some time in a place where people are making small talk I drop a Rumor on them. If they're looking for something to do it's an immediate plot hook, if they're pretty busy it's either (often decided at that very moment) setting information or potential longer-term plot hooks. This is pretty usual so far, though, I figure.

If the player sets out to go learn some rumors, though, the system changes a little. They can either look for rumors of any kind, and then I pretty much just do what I said above, or they can specify a type of rumor they're looking for and I make them make a check of some sort to find useful information on the topic. Basically, I just take what they're looking for and the higher they roll the more accurately and usefully I write it into the game world.

So, if someone is looking for rumors about a magical axe and they roll well they'll find out a rumor about a tomb rumored to contain a magical axe. If they roll okay, it'll be a magical weapon. If they roll mediocre it'll be an unrelated dungeon, and if they roll really bad they'll get some misinformation that still leads them to an adventure of some sort.

I figure it works well since it lets the characters influence what parts of the world get fleshed out, and builds a world they can make generally useful assumptions about while never being quite 100% certain that they know what's coming.

Tom said...

Hey Zak,
Have you checked out 'Hexographer'? It's a free online map-making program. You can find it here:

The biggest problem? The program only runs off the internet, you can't run it while offline unless you've purchased the 'pro edition'.

OTOH the maps look almost exactly like the old school TSR OD&D maps.

While I take nothing away from high-end stuff like Campaign Cartographer 2 (& 3 & perhaps even 4 now) they do have a fairly steep learning curve, particularly if you're not already familiar with CAD software. And they're much more expensive.

Hope it's helpful

Michael McClung said...

At the risk of showing my writing roots, I'd like to mention the philosophy I take when offering the characters choices (be it travel routes, story hooks, or even places to bed down for the night). Essentially, any time the characters have a decision to make, they're basically asking a question: Will this action get me, or get me closer to, what I want? The answer for the DM (who isn't refereeing a completely scripted adventure) will fall into 4 categories:

1) No
2) Yes
3) Yes, but...
4) No, and furthermore...

The most interesting and exciting play (I find) as well as the most interesting fiction comes when the answer is one of the last two. This requires effort on the DM's part, no doubt about it. But if the players at least half-expect that their decisions will lead to interesting, unexpected, dangerous and/or rewarding outcomes, and this is borne out in play, then the DM can get away with the occasional straight Yes or No answer and *still* have the players waiting for the other shoe to drop with the DM giving very little moe than a small secretive smile when the players make their decisions. To wit:

"You want to sail down the Grand Canal to Prrng, the City of Wizards, rather than booking passage with the Royal Dromedary Society? Okaaay..."

Played ths way, it relieves the burden on the DM to have his eye on so many small, relatively unimportant aspects of the adventure setting, as the players themselves provide much of the tension that makes the game enjoyable, wondering what the crafty DM has up his sleeve.

This is not to say that their decisions shouldn't be important to the outcome; I think you're absolutely right that offering the same outcome whatever the players' decision might be is at the very least lazy DMing. All I'm saying is that a minor bit of chicanery on the DM's part is a good thing, if it is in service to the players having a fun, suspenseful session. And if that also allows a busy DM to skip some preparation that may turn out to be unused and unnecessary, so much the better :)

SirAllen said...


iPhone app! Give free copies to your players and charge all your fans/blog readers. Contains writeups, facts, some photos, and is very searchable. Even though Apple is the opposite of Orange Goblin.

Blue Gargantua said...

Could you make a running list of rumors/local knowledge/places to go/people to kill and then build something custom for each player? Or stick them on index cards and just hand them out.

Example: Your players need to go from Alphatown to Betacity. You give them two choices: 1.) Go by boat up Peaceful River or 2.) go by horse through Cannibal Pass.

Put down the following items 1 to a card:

You know a guy who will give you free passage down the river if you protect him from river pirates

There's a 5gp bounty on every cannibal head brought in from the pass.

Peaceful River is low this time of year and you'll have to portage the boat.

Cannibal Pass was rocked by an earthquake recently. The roads might be out and aftershocks could be a problem.

...and other like that and then pass them out. Now everyone knows something different about their choices and everyone can argue over which path to take and you can develop encounters based on what they choose and what you've hinted might be there. Players can hang onto the cards and if you return to the area, they can make choices off of them again.

Anonymous said...

Something I'm going to try out this weekend is creating a setting WITH the players using Dawn of Worlds ( Hopefully this will mean that the players have a greater investment in the setting and will also remember details that they otherwise might not.

Zak Sabbath said...

None of my players are (yet) the kind of people who wouldn't tell other PCs what they know--so there'd b no point to giving them separate rumors. But the basic idea of giving them vague information about their choices is pretty much what I'm gonna do.

thekelvingreen said...

When I started my current game, I was caught in a similar quandary. I gave them a starmap with a number of locations, and for each one, I noted two or three points of interest, as well as the connection between them, and the quality of said connections. That was enough for them to have an idea of what they might find at a location, and whether it would interest them, but didn't involve me writing out a full campaign setting in advance.

So far it's been working well. I started them off with an introductory scenario, gave them the map and asked "where next?" I ask that question at the end of every week's session and it gives me a week to prepare, expanding on the brief notes I have on the chosen location, or the journey to it. Of course, if a good idea strikes me at another time, I'll put it together in advance so that I can have more detail "in the bank" as it were.

So that's where I found my balance. Something different will probably work best for you. I know that if I'd tried to map everything out from the start, I would have crippled the campaign from the outset; I've done it before, but I avoided that fate this time around.

Ripper said...

So I started reading this blog at first out of pure novelty. I quickly realized how well-written it actually was, and now I keep recommending it to people and sharing it in my reader. Today, I shared it with a friend of mine who is in Iraq, who also found it interesting, and had this to say about it:

"That blog is very well written, and his points are pretty spot on.
And it just depresses me
Somewhere there's a gaming party with drunk strippers.
No, drunk porn stars
And I'm here in Iraq sniffing burning poop, with hairy men trying to kill me."

I'm not sure if that's really sad, really funny, or both, but I thought you would appreciate it on some level.

Coopdevil said...

London to Birmingham is over 100 miles

Birmingham isn't helpful if you are trying to gauge relative distances between cities in the pre-modern era. It was only a small backwater town until the Industrial Revolution and only became important in the 17th century due to it's metal forging playing a role in the English Civil War.

Zak Sabbath said...

Seeing as how London-Birmingham is the middle example when I'm making the point that travelling between big cities takes x number of days, it doesn't really matter and my point is still true. But thanks for the history lesson.

Blair said...

I'm dying to see the hexmap Zak ends up painting on his wall...

Make that I'm dying to see the hexmap sysmbols Zak come sup with!

Norman J. Harman Jr. said...

It seems you are defining sandbox == The gameworld is defined as the players explore it. Or am I misunderstanding, again.

This is why big sandbox world authors Tolkien, Howard create a broad brush-stroke gazetteer/history of their world.

With that, it's much easier to know what kinds of dangers Conan will face when he decides to cut through Pict lands instead of the main road through Aquilonia.

If the east route and west route should have different consequences they should have different wilderness encounter tables.

Point being randomness/lack of foreknowledge of details does not preclude alternate choices from having different consequences or levels of risk/reward. You (NPC's, ranger skills, whatever) can tell them route A is slow and safe vs route B which is fast but dangerous without having to have preplanned what the danger is or what's along those routes.

A players "knowledge horizon" determines how far "out" meaningful decisions can be made.

Is it too late or can I still say I agree with your post. I think I've long ago jumped to the "mix" and don't see/have any issues with it.

Zak Sabbath said...

-Not exactly: I'm defining sandbox as "there's no presumed objective--the PCs can go where they want". It doesn't necessarily mean the world is defined or undefined ahead of time.

-the "two different wilderness encounter table" thing only makes sense in this context if the PCs KNOW that the encounter tables will be different (snakes to the south, bears to the north, etc.). If it;s the same table either way it;s not a meaningful choice.

". You (NPC's, ranger skills, whatever) can tell them route A is slow and safe vs route B which is fast but dangerous without having to have preplanned what the danger is or what's along those routes."

Well, yes. What I'm saying is I need to do that.

AGCIAS said...

Have to comment (sent from my sickbed so forgive me if I repeat what others have said). I have a continent-scale map and the players have information brought to them (someone who heard of them and wants to hire for specific job or share in loot), possibly false adventure kernels that they have heard or acquired (map to temple of X taken over by a medusa/journel indicating illegal slavery operation), possibly false rumors and facts that they know (ranger Oam patrols the Dreadwood and will hire backup from time to time)from thier histories. When they choose one, I can flesh it out while they are making preparations and getting there.

Anonymous said...

This... is a freakin' great post.

I proxied a character once in a game at KenzerCo (the Knights of the Dinner Table guys) and I was playing an elven ranger who was the local guide for the party. Every time we came to a new area or ran into a new faction or creature, I'd ask if I knew anything about it. I never did. This was pretty annoying.

I also couldn't hit the broad side of a warg with my arrows from five feet away (which I'm pretty sure I could do in real life), so that was embarrassing.

Anyway. You make some great points about "sandbox" play and how it straddles the line between obsessive setting wankery and just making shit up as you go. Thanks! said...

Dude, you blow my mind! Totally freakin awesome article about Conan. Need to do more roleplaying with barbarians. *sigh* Good old times...

Zak Sabbath said...

Somebody tried to leave this comment but couldn't, so he e-mailed it to me, I'm leaving it for him....

"We play in "fantasy California" (1st ed. AD&D). So, I rely to an extent on the players' letting their characters know basically what California looks like. This campaign (Gang Green; the previous one was Best Laid Plans and took place closer to Tahoe) started out on a larger version of Catalina Island that I could map and expand as needed and has now just reached the mainland for the first time (Baja). Some of the characters have come from L.A. (a druid from Thousand Oaks, e.g.) and so have more knowledge of what might be going on the mainland. The idea is that the actions of the BLP characters made adventuring cool again, something that kids on the farms of Reseda or Valley girl MUs with 10 INTs would dream of. And so to the relatively wilder pastures of Seasteed Isle.

"The balance between what's known and what's determined by gameplay and the needs of any module or adventure I want to toss in (or they stumble across, depending on how you look at) is pretty interesting to me, and I think to them. And since there are maps out there if they're needed, that part becomes a little easier.

"So what they've done (this campaign has lasted about three years of real time; longest-tenured PCs are 7th lvl., though both have lost a few levels to undead along the way) has determined some larger arcs that may or may not come into play. And what they've explored is the most decided territory. But there's just enough knowledge out there (San Diego = crazy town [Lankhmar]; San Fran = rich and decadent; Tahoe = young up and comer; Seattle = elves) to grant things some verisimilitude and allow for a large latitude of choices.

"Though generally they're so busy dealing with the obscene evil they've released in their own backyard that larger considerations and travel plans have to wait."

Rex Venom said...

Damn good stuff.
I feel the need to reach for my Dice!
Rock on!

DreamSeeker said...

Where the hell did you get that "500 miles between capitals"? It might be true with the big western european countries(Italy, France, Britain, SPain, Portugal), but it isn't anything like a common rule. Just look at nordic countries, for example.

Zak Sabbath said...


I am indeed running my game on analogies to "big western european countries". Duh.

Also: I used the word "generally".

Grey said...

"Just look at nordic countries, for example."

The nordic countries would make a good model for a more political campaign - the overall size of a large single nation but split into three by geography.