Survival Setting: What All I Need?

[7th son of sons]

So, in my upcoming campaign, I plan to be running my party through a Survival Campaign. Heavy importance on food, water, safe places to sleep etc. There are tribes scattered about, two very large warring factions (The Ozodrin Tribe and the Grimmoire Initiative), and I have a working world map of mountains, caves, dungeons, locations of some tribes, Random encounter tables for the varying areas of the world. I've also got a (very loose) plot line laid out, only very loose because of the openness of the setting, and the way my characters tend to go about things in... unusual ways. That's all taken me about three weeks, and I've got one week before I'm start before I'm running the setting.

What all am I missing? What more do I need to know about this world before I should loose some PC's Upon it? Any and All suggestions are greatly appreciative.

[TheStranger]

The hardest part of a "survival" game, assuming you mean wilderness survival and not zombie survival, is encounter design, not setting design. D&D really isn't the best system for this, because skill checks are so abstracted. A couple good survival rolls, and you never get lost, always stay warm and dry, and can live off the land. Unless it's always raining, freezing, snowing, scorching, etc., and even that only gets you so far. Basically, anything in D&D that is covered by a skill check is either a solvable problem (stack bonuses until you can take 10), or it's a binary pass/fail that's only interesting for about 30 seconds at a time. That's not a universal truth, and really good encounter design can make it work, but it's the shallow skill system that makes wilderness adventures hard to do.

This makes me sad. I love real-life wilderness exploration, and I wish I could incorporate that into games. There's nothing that says "adventure" more than a blank spot on a map; there could be anything and everything waiting there, and you live or die by your skill, wits, and ability to keep your head when things go wrong. It should be great - and it just isn't.

I've spent a lot of time trying to come up with an acceptable solution to this problem, and I haven't managed it (beyond "be a better DM"), but I've managed to define the problem a little better. My theory is that the "game" portion of D&D consists of decisions, action, and feedback. That's probably not unique to D&D - every game that comes immediately to my mind, from dodgeball to chess, uses that decision/action/feedback loop. Imagine a simple combat encounter - the party is attacked by a band of goblins on the road. So the players (and make not mistake, the goal is to engage the players, not the PCs) each have to decide how to use their abilities to "win" the encounter. They decide what to do, resolve the action, and get feedback - did they hit or miss? Did the hit kill the goblin? Did they miss and leave themselves into a flanked position? What spell should the wizard cast? Did the goblin make its save? Over the course of a single combat encounter, every player goes through that decision/feedback loop several times, responding to a situation that evolves as the combat progresses. And every player stays engaged most of the time, because they only have a few minutes until it's their turn again, and what's happening in the interim directly impacts their next decision. Or at least that's the goal. When combat fails to be interesting and starts to feel grind-y, it's usually because there aren't any real decisions to make. Either because of uninspired encounter design, or because you're just rolling the dice to mop up the last few goblins, the combat has been "solved" and the players are just going through the motions.

And that's the problem with survival-type challenges - the skill system in D&D is too easily solved. Imagine you're in a wilderness adventure; you've decided that the party needs to go through a jungle to reach a remote village. In theory, there are a million and one things that can go wrong along the way and make that an interesting challenge. But mechanically, the Ranger or Druid with maxed-out Survival never gets lost, easily finds food, and leads the party through the jungle with nothing more than a few random encounters (and the other players do nothing). Or you decide that you're going to track food, water, ammo etc. and make things engaging that way. But the "game" then becomes thumbing through the PHB and buying one of everything and a mule to carry it - and preparing to climb Everest has never been the exciting part. Either way, you don't get that decision/feedback loop that, IMO, is the key to engaging your players.

Ideally, this would be the part where I posit an elegant solution to all of this, but I don't have one. I've toyed with the idea of slowing things way down and having dozens of decision points in a single day of wilderness exploration. That might work, but it's a fundamental change in how the game is played and probably isn't right for most groups. You can increase the difficulty of challenges, combine them, undermine the players' assumptions - but the skill system you're using to resolve actions remains both shallow and solvable. And if you ratchet the difficulty up so failure is a real possibility, it's not like missing an attack roll - the skill system isn't that granular. If you fail a Climb check, for instance, by default you either fail to proceed and immediately roll again, which is basically a non-event, or you fall to your death. Multiple checks doesn't solve anything, because the decisions don't change, you just have to roll a bunch of times. Fail a Survival check, and maybe you don't die immediately, but you're hopelessly lost in the middle of nowhere - now what? In that situation, the players are too helpless and confused to make their decision unless they're wilderness guides in real life (and the DM is too, to properly adjudicate things), or you tell them what to do, and then it's often a matter of DM fiat or pure chance whether they get un-lost or not, not the result of their choices. They're not dead, but they've lost their agency because there are no meaningful options.

So, a good encounter, adventure, or even campaign poses a question and then goes on to answer it in a way that gives all of the players, individually or as a group, multiple iterations of that decision/action/feedback loop. They decide what to do, the action is resolved, the situation changes as a result, they decide what to do about that, etc. D&D combat does this moderately well right out of the box, and very well with good encounter design. For non-combat challenges, it takes masterful encounter design to do this even moderately well. The Angry DM has an excellent series on this. Look at the design of the chase scene he uses as an example - that's by far the best non-combat encounter I've ever seen. But I doubt most DMs can reliably design something like that (I know I can't), let alone do so often and quickly enough to make it a significant part of the game.

Clearly, I have spent far too much time thinking about this. Just as clearly, my failure to come up with a solution is because I'm an idiot. Discuss.

[FlayerIV]

I'd love to see that map.

Regardless:
In regards to the issue with dice rolls, don't roll things for the players or tell them when to roll. Because 'an appropriate check' can make everything work, use that against the players. If the players just assume that because they've worked out which direction north is they don't need to roll it again after being distracted by a large tree, then they're going to get lost (see mythbusters episode for an interesting point on this).

Food, water is good, but what about the locals? Are the party considered outlaws and thus can be killed by anyone who sees them with no legal consequence? Are they going to be loved by the locals Robin Hood style or automatically thought of as bandits on sight? Are there any locals, if so, what reason would there be for a random person to be spotted in the forest (Hunters? Charcoal makers? Political/military field agents?).

Similarly, what about the local wildlife, which bear holds what territory and if the party kills said bear how soon until the nearby wolf population adopts said territory (which could also wreak havoc on navigation if they happen to use the wilderness as a guide).

What knowledges do the players have before entering this zone?

[Yora]

A good approach to rolling dice is to make a roll to check how well the character succeeds at his task. Don't make the roll replace the task.

"I place some traps around the camp", is not sufficient. The player should have to say what types of traps he brought with him, what he wants to make from maerials found nearby, what the traps are supposed to do, and how he wants to chose the places where he sets them.
The more detail the player can provide, the better his chance of succeeding his dice roll. (Any discription should only give the character a bonus, even if something is really stupid it should only have no effect at the most, never make things harder.)

Also, as a rule of thumb, if there are no conditions that would make success meaningfully uncertain, there is no need to roll dice at all. For example, in D&D 3rd Edition or Pathfinder, anything with a DC lower than 15 could easily be an auto-success, unless there is something that would make the whole thing a lot more difficult. (Sneaking in heavy armor, climbing a tree with an unconscious person on your back, and the like.)

[Domriso]

Here's how I would do it. Make the checks less binary is the first task. Instead of a "roll above this to succeed, or less to fail," make it more like "roll below a 5 to fail miserably, everything goes wrong; roll between 5 and 10 and just fail; roll between 10 and 15 and you succeed, barely; roll between 15 and 20 and you succeed admirably; roll above 20 and you succeed amazingly."

Obviously you wouldn't use that metric wholesale, but the point is to make the degree of success important. Maxing out one trait will make it trivially easy, but it leaves you open to other issues. Look at Apocalypse World for a good example of a non-binary metric.

From there, other individuals have already put forth good ways to solve the issues. All of them work well.

[Moranica]

I'm a big fan of this setting also and ran a survival game with several groups here on the forums. It kinda blew up, but we quickly learned what did, and what did not.

Idea
DM thread
Then into recruitment etc.

Leaderboard This was to keep them from feeling alone in the world.


Cold Blooded Justice OOC
Cold Blooded Justice IC

The New Age Vanguard OOC

Disciples of Grulk IC

The Builders IC

The New Photosynthesists IC

Phoenix council IC

[boss45]

Remove food and water gathering from being a skill check unless it is setting a trap. Make it common sense(you have to plan, are you going to try to set traps(ask where and add a penalty or bonus to the check to set traps, make the penalty big if they are stupid or lazy and make the bonus big if they are using a spot they have come across or maybe seen animals in before) as about hunting, using the same bonus/penalty system and penalize them if they try to go too far from camp as they will spend too much time traveling and not enough sitting still waiting for prey, let them craft tools for gathering food that add stacking bonuses but take time and resources to make(bows for hunting, nets for fishing(requiring multiple people to work the best), snares and traps for trapping are a good start. What about bird decoys, stealth aids, animal calls, scent attractants and covers? Lets go a little further with animals in cages or shelters where they are kept(think chickens for eggs and slaughter),blinds, gardens, fishery ponds where you feed trash food scrapes to certain fish that can eat them.

Don't be afraid to spring some strange, fantastic creatures on them that have unique properties if your setting allows.

Reward hard work and creativity. Make food gathering a fun minigame(with emphasis on fun).

Overuse of an area should cause that area to have little game, extreme overuse could cause local extinction of an animal. Small setbacks should occur occasionally anyways if things get too easy.

Far as water goes. Wells take time to construct. River water should have parasites like real world water that are killed by 5 minutes of boiling and diseases that are killed just by bringing the water to a boil. Make it a chance of getting a parasite or sickness thing with a Con save for diseases but a very high con save for parasites(shiould be odd to succeed). Some vines or gourds or cacti could have drinking water. Water catchment could come into play(a leather tarp stretched out with a hole in the middle weighted by a circular stone or something over a barrel, big bowl burn/carved from a stump, or cauldron.

This is really nonstandard, so maybe the PCs should have a native guide who gets crippled by the plot early on and who only serves as a mouthpiece as its condition worsens and it dies.

[VoxRationis]

Also: keep track of all spell components and do NOT allow a spell component pouch or the like, if you're allowing any casters at all. Wizards in particular are surprisingly dependent on a society with lots of specialized craftsmen and merchants. Making it so that they don't have unlimited "splinters from siege engines" and other bizarre components will help put the party in the mind of resource conservation.