Information in a sandbox

Here's another question: If you want to make exploration of the game world a priority in sandbox play, how much information is too much and how much is too little?

As far as I can tell, you want to make the players aware of the risks so that they can make their choices, but you don't want to reveal so much that they aren't surprised or aren't forced to make logical deductions about the game world.

Let's say that, in a large mist-shrouded lake, there's a local legend: no one who goes into the mist ever comes back. Their boats return, washed up on shore, with no sign of their crew.

The DM knows that it's because harpies on an island in the middle of the lake fly out and take the people back to their aerie to eat them.

You want to make the players aware of the level of risk but you don't want to give away any secrets that would be interesting to discover through play.

At the moment our legend doesn't work: the level of risk is unknown. We'd want to put something in there about the level of risk in a natural way. (The easiest way would be to say, "This area is level x", but eh.)


Another question, related to mechanics: Would it be a good idea to tie some information to skill checks?

You'd want to make the result of the skill check secret, I guess.

Would something like this work:
DC 20. Success: The PCs learn of a fisherman who did return but he went mad, babbling about "the bewitching song". PCs who talk to him may be able to extract more concrete information. Failure: The PCs hear about an alchemist who ventured to the edge of the mist and has been studying it; his theory is that the mist is a poison that drains the body of liquid, causing the fishermen to dive into the lake to slake their thirst, and eventually they drown.

I have a feeling I want to make each roll carry some kind of risk.

There are some dependencies to be considered.

Strangers should know the bare minimum informations only (major city, major trade route) and work their way up from there.
Depending on Edition, I´d really use knowledge and Bardic Knowledge checks, besides regular gossip. I´d say there is more than one way to get usefull information, all dependent on the angle the players use to aquire it.

I'm planning on starting a sandbox style campaign soon, taking place on a large chain of islands in an otherwise mostly water world.

My plan is to use Gather Information and Knowledge skill checks to give the PCs information on what's out there.

In your lake example, I would only include the "false" information on a severe failure, say 5 or more lower than the DC (in other words, the roll comes up a 15 or lower for a DC 20 check).

I'd also provide several different sources of possible information. The mad fisherman and the local scholar being the obvious choices, but also include a popular drinking song at the tavern that tells the tale of a local legend about women who sing you to you grave or who steal your heart and eat it.

To the original example, the only thing I'd add is that, "Bob, the famous warrior, and Phil, the powerful wizard, went into the mist to try to stop the menace and never returned." That should at least tell the characters that they better be a famous warrior or powerful wizard before setting out on the quest themselves.

(By the way... This is one of the little flavor things lost between editions. Level titles could always be used to give the players information like this... "Bob the Swordmaster" and "Phil the Warlock" would explicitly give the players a ton of information without resorting to "vulgarly" telling them classes and levels.)

LostSoul said:

At the moment our legend doesn't work: the level of risk is unknown. We'd want to put something in there about the level of risk in a natural way. (The easiest way would be to say, "This area is level x", but eh.)

Click to expand...

"The famed ranger Golchescue and his band went to that island. While he was able to beat the stone giant Thorlog, it seems that island was too much for him."

Also, I had a GM for a White Wolf game once, who had the habit of assigning a session a "threat level", and announcing this before we began play.

Information in a sandbox is a little tricky.

One of the tools as a GM you can use is to parcel out information about things that aren't exactly close and keep the bits that are close near or at the challenge level appropriate for the party.

Part of this is going to depend on how you're running the campaign. For example, the harpies. Is it a full fledged harpy dungeon or is it a single encounter? Single counters in 4e, much like in 3e, allow the players to expend a lot of their resources and make them effectively a 'higher' level in terms of what they're able to put into the fight. You don't have to worry about saving dailies if you're going to just have this one encounter.

In terms of the harpies, you can tease the players with it a bit.

1. They hear about it.

2. They explore some of the beaches around the fog. Perhaps they hear something out there.

3. One of the boats that washed up, despite it's battered and beaten frame, has some evidence of what's on the isle... feathers....

Hope that helps!

LostSoul said:

Here's another question: If you want to make exploration of the game world a priority in sandbox play, how much information is too much and how much is too little?

As far as I can tell, you want to make the players aware of the risks so that they can make their choices, but you don't want to reveal so much that they aren't surprised or aren't forced to make logical deductions about the game world.

Click to expand...

Just to get it out of the way: Everything the GM tells the players is information. This is the nature of the game.

IMG, the information the players receive is the most common reward the game offers. The quality and quantity of information the players have on hand, through memory, notes, maps, etc., affords them a greater ability to overcome any given challenge. Of course, our game is a deductive reasoning game, so you may not desire it. But in the end, the amount of information the players have about any given situation will alter the difficulty for them in facing it.

My advice is, when making up a challenge define the amount of information potentially accessible to the PCs and include that into factoring the difficulty for it. More intelligent foes will seek to limit information about themselves. This includes lairs, treasure, abilities, activities, and so on. More limited information makes the NPCs more dangerous. But this works for the PCs too, so...

Let's say that, in a large mist-shrouded lake, SNIP

Click to expand...

Harpies are a 3hd CE creature with 2 special abilities, magic resistance, flight, low intelligence, and 3 attacks (2 low, 1 avg). They are about a 5th level challenge IMO, which is pretty tough for just one. But they are not brutes. They lure with their song, Charm by touch, and devour their prey in safety. 2-12 are the No. Appearing, but this is throughout their territory and probably not all at once given their Int and Alignment.

Some good tactics to assign here include: not choosing to fight openly or directly unless pressed into a corner or guarding a nest, using their song SA from a hidden position, low morale, and attempting to leave with downed opponents, if others were not charmed.

IMG, four levels of enemy filled dungeon exist to be encountered or surreptitiously bypassed by PCs to reach the harpies' position. This could include the lake, the surrounding lands, or perhaps only the island itself up until the harpies' territory/nest is encountered. These foes don't need to be in friendly with the harpies, but they are their because it is safe ground for them, Chaotic territory. Many will likely have had interactions with the harpies and may know a good deal about them depending upon the time span, location, and regular activities involved. Questioning one or more is a way to gain information. Evidence of the harpies' regular activities in their own territory offers more clues to what they are and what they do.

Another question, related to mechanics: Would it be a good idea to tie some information to skill checks?

Click to expand...

You can. I prefer not to use them. I think it is offering up to chance and fiat the reward of potentially vital information rather than leaving it to the results of the players' decisions. You have a couple of good rumors there though.

Does a stone have spatial dimensions? Weight? Color? Texture? Scent? Taste? Does it produce a sound when struck? Does it break when struck hard enough? Do heat and cold affect it? Is it chemically reactive?

There are more answers in the world than time to give them. Leave it to the players to ask the questions!

I am quite at my wit's end in trying to figure out what "sandbox" is supposed to mean.

I can suggest, however, that when one lets the players decide what is important to them, then they will decide how much information is enough.

They will decide how much risk is too great, how much reward is too little, and how much time to spend on this or that activity.

If you have a "skill check" mechanism, then it seems to me the application ought to be straightforward. If not, then you're dealing with a design too convoluted for me. If someone's reading a book, then I'll use the "reading a book" factor; if searching for spoor, the "searching for spoor" factor; if listening, the "listen" factor; and so on.

How is "success = you get some real information" and "failure = you get misleading information" not straightforward?

You could elaborate on the virtues of actually playing through the information gathering: "Maybe one of these fishermen knows something more." That's a fair argument and the reason I brought it up in the first place. What are the virtues of doing it like that? What do you gain, what do you lose?

But that's not what you did, is it? Okay: we know you are contemptuous of this method. Great.

Why?

Lost Soul, I am just baffled. Contemptuous of what method???

A systemic problem calls for a systemic solution. "Doc, it hurts when I do this!" That it is a problem in the first place might reflect your having in hand a tool very well designed for a different job (which has never dissuaded D&D "hackers", of course).

It's unclear to me just what your difficulty is, though, unless it's uncertainty as to just what you really want to do.

Do you "want to make exploration of the game world a priority" -- or do you want "the level of risk to be known" without that exploration? Either way is not exactly rocket science, but trying to go in opposite directions at once might be frustrating.