DMs: thoughts on sandboxes and balance?

[HidesHisEyes]

I'm thinking about running sandbox style campaigns in 5E, and I'm wondering how others who do it handle CR and balance and all that?

My instinct is that if I'm running a proper sandbox I want to give the feeling that each place is real and it's not just a big theme park for the PCs' amusement. That means I don't want to go the Skyrim/Oblivion route and have threats scale to their level everywhere they go. Also, when I say a sandbox I don't mean a largely undefined world that I fill in as the campaign goes on (I do this in Dungeon World, I don't like the thought of it for 5E). I mean rather a sandbox where some amount of prep work has been front-loaded to create regions and adventure sites, and then the actual adventures that happen in that space can be sort of layered on top, if that makes sense?

But of course, if different regions have monsters and threats suited to different levels/tiers of play, that somewhat limits where the PCs can reasonably go. Now I'm actually not too worried about regions that are currently too hard for the PCs, since I'm fine with "oh don't go into Murderdanger Forest, it's too dangerous", and having them more or less confined to a smaller corner of the map at low levels. I'm actually more interested in when the PCs are higher level, maybe have access to powerful transport spells that let them zip around the sandbox pretty quickly, and I might want an adventure to take them around various distant locations. Assuming they don't just fly and teleport everywhere, when they do travel through "lower-level" regions, how do you keep things interesting when they're in places that aren't particularly dangerous for them?

Should I be thinking in terms of terms of non-combat encounters? Am I worrying about nothing?

Last little request: If you've interacted with me on this forum lately you might notice I have opinions about RPGs and I'm easily tempted into a philosophical debate about them. But this thread is meant to be a genuine request for guidance with a particular style of campaign. Could I please ask that if your opinion is that sandbox campaigns are bad, or that what I'm describing can never work, or whatever, feel free to PM me about it but please don't put it in this thread. I don't mind people questioning any assumptions I've made in the above, or pointing out problems in my thoughts here, I'm just not asking for a debate (on this occasion).

[False God]

Big and powerful creatures are FAR more rare in sandboxes, just because of their environmental effects and needs, than they are in standard adventures/campaigns or in theme parks.

It balances well, but it can be dull. Reality is full of a lot of empty space populated by largely nomadic animals who travel over tens if not hundreds of miles. It's rare to find places that are "teeming with life", much less life that will give combat challenges. There should be a lot more exploration and social challenges in a sandbox.

As far as high level goes, it means they're going to have to find challenges. There's stories of a mighty dragon in the Pinnacle Peaks, but there are scant details. Maybe that semi-active volcano? Maybe that frozen peak? I tend not to worry about making low-level threats relevant at high levels, the players have worked hard to reach high level! When those bandits who don't know the party from Peter or Paul jump out of the shrubs, let the party have fun with it.

As far as low level goes, it means they're going to have to put effort into avoiding the dangerous areas. World design shouldn't confine them to "low level" areas, because that's essentially a theme park (think WoW where every zone has a set level range vs Skyrim where the whole world scales up with you). World design should be far more mixed, yes there's the Dark Forest where everything will totally wreck you, but right next door to it is the Twilight Woodlands, which is less dangerous, but runs the risk of powerful threats wandering in.

Players most importantly need to understand that retreat is a perfectly valid option(I find this is hard with D&Ders). They're not trapped in a dungeon where the only path is forward. They'll figure out how dangerous parts of the world is fairly quickly if they're allowed to dip their toes without being forced to dive in.

[OldTrees1]

I'm actually more interested in when the PCs are higher level -snip-, when they do travel through "lower-level" regions, how do you keep things interesting when they're in places that aren't particularly dangerous for them?



Typically in a sandbox the PCs have a reason for what they are doing and the reason is rarely "I can fight stuff". Make sure your adventure areas have enough interesting parts such that the party find an interesting reason to be there. If they want to reestablish the local wine industry, then it is not as big a deal if any incidental combat is non challenging.
Sometimes a combat is decided before the last foe falls. At some point in their travels I will start to narrate non threatening combat (during the night you were attacked by wolves, your rest took an extra hour) rather than roll initiative. This shortens the less interesting parts and helps keep the attention span.
Sometimes the region reacts to the apparent strength of the PCs. Hungry wolves are more likely to attack frail humans than high level characters that look like demons. This also works for bandits. This also helps skip less interesting parts and keep the attention span.
Sometimes the PC actions have consequences that spill over into other regions. Unlike Skyrim/Oblivion's abstract level scaling, these are specific outcomes of PC choices that fit verisimilitude and the consequence is independent of the PC's level. I would not ensure every region is affected, I would just affect the regions that would be affected.

Maybe ignoring an area long enough causes the situation to progress. Leaving the nearby village and bandits alone might cause the bandits to conquer the village and legitimize themselves as its new government.
Maybe defeating a threat in one area causes it to scatter and pick up pieces in other areas
Maybe there is a non localized threat that is becoming increasingly interested in the PCs, as time goes on. It might mobilize forces near the PCs.
Maybe the PCs unleash a greater threat that starts to wander the lands (I was surprised when they did this)

[Goobahfish]

I run a sandbox campaign.

Some things to keep in mind.

#1: You need to have a party that isn't too 'gung ho'. Sandbox adventures work best when the 'difficulty' of the encounters are highly variable. A difficult encounter with good preparation can be made medium, a surprise low level encounter can be made difficult etc etc.
#2: To subserve (#1), you need to establish that the players can lose encounters, and have little chance of winning them in a straight up fight. Unfortunately 5e doesn't really support this because 'running away' doesn't seem to be a 'well codified' option in the game. Not that you can't make stuff up, but it is difficult to make satisfying rules for this kind of thing on the fly. I wrote an entire set of chase rules which work very differently from 5e DMGs (basically there is a generic distance which has nothing to do with combat, lots of obstacles etc etc.). Likewise 5e's rules for stealth aren't well suited to 'satisfying' encounters. My recommendation here is to make sure the 'not combat' options players have (talking, running, hiding, tricking) have the allure of 'non-DM fiat' so that players are encouraged to use them to solve problems. This lets you throw 'too hard' encounters at them.
#3: It helps to throw the party a 'too hard encounter' where the punishment is not instant-death. Perhaps capture/mistaken identities or something. Or, the party gets bailed out. Just something which gives them the impression that random fights for no reason are a really bad idea. NPC: "You were lucky this time, but you should change the way you deal with the world or else it will chew you up and spit you out... hint hint".
#4: Make most of the challenges in the game not directly solvable through combat.
#5: Throw the parties a challenge where it is transparently explained to them that the threat is non-trivial (i.e. straight combat will not work - some kind of nasty monster which has killed even more experienced looking soldiers). Have the players try to 'figure it out' rather than go fight it.
#6: Avoid training your players by having many low-level RNG encounters to 'soften them up'. It creates the expectation they will win any fight they encounter. They should learn to fear encounters and often over-estimate weak enemies.
#7: Also remember that 5e is pretty bad for sandboxing because of that whole 5-minute adventuring day thing. So, the goal here is to try to find a balance between 5 minute adventuring days (which will be the norm, because only idiotic characters would not push for 5 minute days) and overloaded events. If resting in the wild spawns low-level encounters, eventually it will wear the players down. The problem is it might not be 'fun'. So take a challenge the players have had before. Say an Owl Bear. And multiply it by 10. Ten Owl Bears. The sentry spots them. They are heading this way. It doesn't seem like the Owl Bears have noticed you yet. Even the densest player will realise that fighting 10 Owl Bears (assuming one was a non-trivial challenge) will be a deadly encounter. Their sleep will be disrupted by this, but the goal might be to stealth or lure the owl-bears away. Remove spells that trivialise this from the game.
#8: And perhaps the most importantly. I always throw a lot at players (quests/missions/plot points) and then see what sticks. Players will choose their own encounter thread. Then I complicated everything they didn't choose (i.e. bad stuff happens when you ignore it). This will give the false impression that the world is reactive and 'big'.

There are probably more, but these have worked thus far.

[Telok]

Now I'm actually not too worried about regions that are currently too hard for the PCs, since I'm fine with "oh don't go into Murderdanger Forest, it's too dangerous", and having them more or less confined to a smaller corner of the map at low levels. I'm actually more interested in when the PCs are higher level, maybe have access to powerful transport spells that let them zip around the sandbox pretty quickly, and I might want an adventure to take them around various distant locations. Assuming they don't just fly and teleport everywhere, when they do travel through "lower-level" regions, how do you keep things interesting when they're in places that aren't particularly dangerous for them?

"And three weeks later..."

No, really, that's mostly it. On occasion throw something in, travelling titan scaring the begebus out of everyone & shutting down trade, off course purple worm eating random villages, low end ignored cult ghouled an entire town. But only about every 4th to 5th trip or something. You can, if they stop in towns, have random people asking them to clear rats out of a basement or a mayor having a "ghost goblin tribe" crisis (under strength brigands in masks & shouting memorized gobbo catch phrases) or something.

But mostly "And three weeks later..."

Edit: forgot to be extra precise. Stuff that's cr appropriate (if thats your care) that travels, stuff that's genuinely too low to threaten but you can play for laughs & comedy, abusing hand-wavium npc skills & magic to flood bounded accuracy with mooks in huge numbers, and of course anything that's not just a combat/right spell solution still works.

[Tanarii]

My recommendations are:
- don't do a true sandbox if it's only for one group of players. It's not worth the time and effort.
- set aside a lot of time for initial prep. At least 100 hours.
- steal content from everywhere you can, and use that.
- when you custom design an adventuring site, do it for a expected level range, not a specific party.

I highly recommend the Alexandrian articles on design for hex map campaign. Even if you only take core concepts, like spreading out adventuring sites to find and stealing content to populate your area.
Start of the series: https://thealexandrian.net/wordpress/17308/roleplaying-games/hexcrawl
Stocking areas: https://thealexandrian.net/wordpress/17442/roleplaying-games/hexcrawl-part-10-stocking-the-hexes
& https://thealexandrian.net/wordpress/17451/roleplaying-games/hexcrawl-part-11-more-hex-stocking

If you're going to put in actual exploration, you'll need to build some game structures for it. Again, see Alexandrian.

If you're going to run an open table, which is the best way to make all this effort pay off, figure out how you're going to handle ending a session. The party can't end in the wilderness or adventuring site unless you've figured something out.

Edit: Oh yeah, highly recommend the adventurers retire at level 11. And later on you can set up a special events / hard to access and limited time adventuring site where players can bring characters out of retirement.

[Psyren]

You can make the difficulty of specific areas in a sandbox as mismatched with the party as you want. For example, there's an Ancient Dragon on top of Ancient Dragon Mountain, and if the PCs go up there, they'll face Ancient Dragon - whether that's an appropriate encounter for them or not.

The key is to make sure that:

(a) Difficult areas are telegraphed. You don't have to put CR numbers all over your map, but the players getting no warning about dangerous areas is punishing.
(b) Difficult areas are avoidable. If the only way for the players to reach Easy Kobold Cave is to climb Ancient Dragon Mountain, that's not good design.
(c) Retreat from difficult encounters is possible. If the players bite off more than they can chew - especially if it's because you didn't do as good a job with #1 and #2 as you should have - then a TPK can just be discouraging for the party* and should be avoided.


(*Some groups don't mind wiping and making new characters, Tomb of Horrors-style. If high lethality and comedic deaths are what they want, go nuts!

[HidesHisEyes]

Thank you everyone, this is all great and very much appreciated.

[Chronos]

What makes an area "safe"? Most likely, it's the presence of a large concentration of people protecting it. A large city, for instance. But cities bring with them their own distinctive problems, many of which can't be solved by "be powerful enough to beat everything up". Sure, a party in the teens could probably take down Xanathar in a fight... if they can find it.

[JonBeowulf]

Lots of really good stuff in here that I fully support.

I suggest letting the party occasionally stomp a low-CR encounter. When done right, it can be one of the things the group remembers forever.

We were between levels 7-9 when we encountered a small band of kobolds led by a Pirate Captain <impressive name>. We were set to just roll over them, but Captain <whatever> was super arrogant and we started doubting our assumptions. We probably spent 15 minutes talking to this guy before we gave up and rolled initiative. The fight was over before they had a turn. Turned out he was just a cocky kobold who found some decent-looking gear and convinced a few others that he was awesome. Low-level NPCs treated us like heroes when we told them we killed him while more powerful NPCs were like "What? That guy was a nobody." and we'd be like "IKR?"

We talked about that encounter for weeks.
Hand-wave the typical stuff but make a few of them interesting and play them out even though the party is in no danger. It obviously doesn't work in an area in which the party is well known but it's great when they're in a new part of the world.

[kazaryu]

I'm thinking about running sandbox style campaigns in 5E, and I'm wondering how others who do it handle CR and balance and all that?

My instinct is that if I'm running a proper sandbox I want to give the feeling that each place is real and it's not just a big theme park for the PCs' amusement. That means I don't want to go the Skyrim/Oblivion route and have threats scale to their level everywhere they go. Also, when I say a sandbox I don't mean a largely undefined world that I fill in as the campaign goes on (I do this in Dungeon World, I don't like the thought of it for 5E). I mean rather a sandbox where some amount of prep work has been front-loaded to create regions and adventure sites, and then the actual adventures that happen in that space can be sort of layered on top, if that makes sense?

But of course, if different regions have monsters and threats suited to different levels/tiers of play, that somewhat limits where the PCs can reasonably go. Now I'm actually not too worried about regions that are currently too hard for the PCs, since I'm fine with "oh don't go into Murderdanger Forest, it's too dangerous", and having them more or less confined to a smaller corner of the map at low levels. I'm actually more interested in when the PCs are higher level, maybe have access to powerful transport spells that let them zip around the sandbox pretty quickly, and I might want an adventure to take them around various distant locations. Assuming they don't just fly and teleport everywhere, when they do travel through "lower-level" regions, how do you keep things interesting when they're in places that aren't particularly dangerous for them?


So, the first thing to keep in mind. most areas in the wild theoretically are going to have some pretty major threats. Its not a question of if those threats exist, its a question of if the PC's are going to encounter them. For example: say the PC's wander into an area that you happen to know is an ancient dragons favored hunting ground. that doesn't mean they *must* encounter that dragon. Its entirely possible they go by unnoticed. or perhaps they are noticed, and the dragon instead stays at a distance spying on them. after all, perhaps they'll lead it to a source of treasure within its own borders, etc, etc.

second: i definitely wouldn't suggest telling the PC's 'don't go there, its too dangerous'. In the epic fantasy genre that is exactly how to signpost 'this is where PC's go to prove they're badass, by surviving the unsurvivable forest'. That is...unless you're explicitly saying it to the Players, rather than your characters.

as far as making the lower level areas interesting. as i alluded to above, there's unlikely to be an area that is *actually* too low level for the PC's. there's always potentially larger threats to face. And in 5e, these larger threats can easily be made up of lower level monster, because of bounded accuracy and the importance of action economy. as an example: dire wolves are only CR 1. but they have 37 HP, so they're not necessarily going to get one-shot. you can even buff their HP/stats a bit if you want, although its not neccesary. So say you send a pack of dire wolves at the PC's that are like...level 10. obviously this is not a fight that the dire wolves are going to win long term. but the direwolves are hungry, not murderers. meaning that the dire wolves goal in that fight is different from the PC's. instead of trying to wipe the party...they try to bring on party member down and drag them off into the woods, while the other pack members attempt to ward off the PC's. perhaps they do an ambush when only 1-2 PC's are awake. "trying" to kill that one PC quickly and just drag them off (of course they can't tdo that...but they're wolves, they don't know what HP is.

either way, the point is to make it clear that the wolves will be perfectly content to just kill 1 person, and they are immediately a much more dangerous threat. because any death in the party is unlikely to feel good...and thus preventing it can


another option to think about is temporary threats. perhaps something happened nearby in the area, leaving behind some demon(s) that is/are running around killing stuff. its not a native threat to the region..but there are other people in the world. maybe this demon/elemental/whatever you want it to be was summoned to assassinate someone else that was traveling the area, and the PC's come across the campsite mod attack, or just post attack. maybe some wizard was trying to achieve immortality, but was tricked, and the spell actually just brought forth some powerful outsider that then killed the wizard. things like that.

lastly, there's always the option (table dependent) of just...not worrying about it. hand wave the journey as being 'not worth really focusing on. Let me know if any of you want to have specific RP moments, but other than that here's how the trip goes:....'

[Xervous]

Seems all the points I’d hit have already been addressed, but I’ll add a little anyways.

Sandbox campaigns offer broad choice and freedom to the players. It’s a freedom to choose what they want to do, where to go, and what they interact with. In order to make good choices, to feel like they are making worthwhile choices, they need to be informed.

If Fiery Doom Summit is suicide by adventure, they know that, and they choose to go anyways, that’s what they’ve chosen. If they know the bandits are a joke for their L15 party and they choose to go to Muggers Marsh anyways, that’s what they’ve chosen. The players deem something important to their characters, and then play out their characters acting on that important something.

Sandboxes work best with invested players whose characters have things they come to care about.

It’s not about finely tuned combat, it’s about exploring and interacting with a world. The goblins are a joke when you’re 15th level. The dragon is certain death when you’re 3rd. The players should understand they won’t get a dragon dropped on them without reason. They should also understand you’ll drop hints or make it outright clear when they are doing stuff that would bring a dragon down on them. When the players understand they need to buy in on some hazardous situations, combat balance becomes less of a concern. They will push as far as they are comfortable, and take risks they deem appropriate for their current goals. If they encounter something that is initially insurmountable all you need to do is suggest other ways they might go about securing the capabilities to deal with the obstacle. In this way players can end up designing their own quests as they break large tasks into series of smaller ones.

[Sigreid]

What I do is I don't have high and low level areas. All areas have a mix of levels of opposition. Then I have in world signs to clue the party in on what they're heading into and give them ample time to make the decision that "this is for later". That said, if they see the signs and decide to continue on into the dragons' maw, well, that's on them...

[Sorinth]

In terms of combat difficulty the players have to understand that they will on occasion have to run/hide from a fight. As a DM your job is to telegraph those situations, don't spring those types of fights on the party via random encounters where the monster(s) are 30ft away and it's straight to rolling initiative. So the players are travelling through Dragon Forest have them be miles away when they spot it flying in the air. There's no chance of an actual fight but you've shown the dragon is there and is an option down the road when they think they are ready.

It's also a good idea to understand how the group will generally look to handle an encounter. If bandits waylay the party is the first instinct to fight, or is to try to talk their way out. If it's the latter it's easier for you to throw in those difficult challenges, whereas if it's the latter then you have to be more careful or risk a TPK.


In terms of region, I think you should have a varied level of threats, the big ones are foreshadowed and relatively easy to avoid. So in Dragon Forest there could be low level encounters like goblins/giant spiders even though there's also a high level encounter with the dragon as a possibility. So at low levels the dragon pretty much ignores the PCs because they aren't worthy of it's attention, at high levels the goblins/giant spiders either avoid the party because they know they are too weak to take them on or you can just narrate encounters and how the party easily defeats them as part of the travel.

[ImproperJustice]

Check out Worlds without Number by Kevin Crawford.

It has tools to help you build a great sandbox campaign like in a matter of minutes or a few hours vs. hundreds.

It’s free, give it a read. Just it’s world building advice is great.


Just get a skeleton framework around your party and keep filling in details as the game develops and your players motivations grow.

I’ve done the sci fi version and it really is a great way to play. I prefer sandbox style these days.