I just finished having a conversation with one of my players, and was led to a question I'm not sure I know how to answer.

I'm running a sandbox campaign (D&D 4e, if it makes a difference) where there are a few major events taking place, but not set expectation that the players have to take part in them. Thus far, the party has tended to follow up more on smaller leads, like interesting caves, a village with a plague, etc., rather than major leads, like fire falling from the sky or the rediscovery of a massive underground dungeon. The leads are clear, and they are aware of them, but haven't followed up. I am totally fine with this.

But in talking to my player tonight, he said that he feels that the party is looking for a major story arc, rather than what they have been doing. Their tendency to follow up on little things is, he thinks, part of an ingrained video game completionist paradigm that wants to leave no stone (i.e., quest) unturned. The problem is that the way I have been running my game, there are an infinite number of stones. Want to follow up on x? I will make x interesting.

It's a weird thing where allowing the players to do what they want seems to lead to them doing what they don't want to do. How do I break the video game habit and encourage players to do what they want?

  • 14
    +1 for allowing the players to do what they want seems to lead to them doing what they don't want to do :-) – Martin Jan 13 '11 at 9:45

13 Answers 13

Thus far, the party has tended to follow up more on smaller leads, like interesting caves, a village with a plague, etc., rather than major leads,

Why do those leads end up being minor? Can you turn an interesting cave into a major lead? Maybe the cave leads into the massive underground dungeon that was rediscovered. Or maybe the cave has a family of villagers who fled for shelter when the fire fell from the sky. Just because you intended for these things to be minor doesn't mean they have to stay that way. Take the plots that the players show interest in and hook them into larger plots.

Depending on the game you want to run, you have the option of either merging these plots into the established plot or branching off that plot and making this new plot into the more important one. It can feel railroady to the players when they try to take things in their own direction, but still end up knocking on the same BBEG's Fortress o' Doom regardless of the road they took.

  • 6
    +1 All quests, no matter how seemingly unrelated, lead to the main threads. – F. Randall Farmer Jan 13 '11 at 3:50
  • 2
    "Just because you intended for these things to be minor doesn't mean they have to stay that way." -- +1. In a campaign I'm currently running, I intended for a mountain to be simply background scenery. My players took interest in it, and the mountain completely replaced the story arc I had intended, culminating in the discovery of an ur-god and the annihilation of said mountain. =) – Brian S Dec 4 '13 at 18:37
  • 2
    -1 Feels a bit like cheating to me :/ i.e. "whatever you do, you end up in the main plot". Pretty bad railroading, entirely stripping any sense from the players' decisions. – o0'. May 18 '14 at 10:50
  • 1
    @Lohoris you don't have to hook them into the main plot. Just hook them into something. If the players find an NPC interesting, you're free to fill out his background in a way that gives them lots of game time instead of the thirty minutes you had planned. All roads leading to the same plot can feel contrived, especially if the players see you doing it, but that doesn't mean that you can't explore in different directions or let them pick which plot becomes the main plot. – valadil May 18 '14 at 18:21
  • @valadil this is very good, but the actual answer says something different. I'd be happy to revert in case it is rewritten with this in mind :) – o0'. May 18 '14 at 19:22

In Sandbox campaigns, what the players are doing is the centre of the campaign. If they do not investigate and stop the fire falling from the sky, then there is a consequence and you should play it out as if real history is unfolding before them. I find that the best way to do this, is to image what the major characters of the world are doing (in shells around the party). Then have them do it, that is, invade here, doom there to fire from the sky, rob the local tavern etc. Only spend any time designing something when the players actually look like they might interact with it. At some point they will normally change what they are doing to go an hunt down the big bad. Sometime you will need to prod by getting the big bad be on there radar directly. This is once they realise they will have to do it themselves. You have given them something close to them that matters. Do not just handily place a muguffin to get them to the next plot point it breaks the sandbox. People in real life don't try to stop a mad man with vaste army destroying the know world, that takes you having an army too. But find a scouting party that leads to a secret way into a underground base, well that is another thing ...

  • 7
    +1 for If they do not investigate and stop [it], then there is a consequence and you should play it out ... – Martin Jan 13 '11 at 9:50

Even though the players are the center of the campaign in a Sandbox, they are not the only cause of action in the world. They want to investigate the village with plague. Let them. Meanwhile the fire raining from the sky continues (and unlike videogames, does not go into a static point of no extra destruction until the players show up). Maybe the Fighter grew up near the fire-rains, but thinks "we'll have plenty of time to get there, after we look into these plague-ridden villagers". Take that from him. Have his childhood home burn to the ground. Actions have consequences. Even lack of action should have consequences.

Plot lines should follow the following progression:

  • Seed (There's fire raining from the sky near the town of X)
  • Sprout (The fire-rains destroyed town X and are spreading towards town Y)
  • Bloom (The Fire-God's Wrath has spread to the whole state of Z [which includes both towns X and Y)
  • Fruit (Since the Fire-God's Wrath has spread, the towns of X and Y are defenseless, hungry, and plague has broken out in the devastation.)

You drop the seeds, and some will grow, others don't. It happens. When you drop a seed give it a linear progression of what will happen if the players don't react to it. Then think of a few ways the outcome changes based on how the characters react to it. When/if a thread comes to the fruit stage, go crazy with new seeds. Maybe you don't want to follow up too much on the plague idea, make it a minor plague that once food and medicine gets to the affected area, it clears up. But the warlord across the border thinks this is the perfect time to expand his territory, what you/the players want to follow up on. Go crazy and have fun with it. Sooner or later, your characters will come up and say, "Hey, there was a point a few weeks ago, and now that nothing major is going on we want to follow up on that." You now no longer need to worry as much about the plot because the players are helping you create it.

  • 3
    The is the same structure of the much-lauded 'countdown' threat system that Apocalypse World formalised (not "invented", to be sure!) and Dungeon World inherited. It works very well! – SevenSidedDie Dec 4 '13 at 17:49

The answer about turning the small leads into bigger leads is great, but may not fit what you want to do with the game. I've dealt with this in several ways, usually escalating:

  1. Just have boring answers to the leads. "Sorry, the caves turn out to be not that interesting," or "the caves are frozen over and would require weeks or months of work to chisel through... or you could just wait until the summer when it melts." Usually the players will start to be able to identify these subtle clues that they're heading off to an undefined area of the plot.

  2. A NPC can be great here, to charge the players with following a plot line -- having a person with a sympathetic backstory leading them the right direction can help direct them when, given spare time and nothing better to do, they might pick up on any smaller lead that interests them. Example: An old man needs them to get this medicine to another town for his granddaughter -- he would do it but he's afraid he might not make it in time. Oh, and they just happen to get there in time for the totally unrelated rain of blood. Note that overuse of this device can cause the party to become violently suspicious of the sympathetic and aged.

  3. Break the fourth wall (i.e., talk directly to the players about it). It's not something I like to do a lot, but if things are really heading in the wrong direction, I think it's a very reasonable thing to do, to drop out of character and say, "Hey guys. I think I've given you the wrong idea about where the interesting part of the plot is. You can keep heading this way and I can try to roll with it, but it's not what I had in mind for the game." Of course, if you get to this point (and I have a couple of times), it probably means that this set of players really wants or needs more direction than your game is giving them.

People sometimes claim to like pure sandbox play, but then they get bored or easily distracted for this exact reason, not knowing what to focus on. Really they want a little railroading. Some groups are self-aware enough to come out and say this, some aren't. But if you're starting to get that kind of feedback, it may be time to make some of those leads uninteresting/lead them back to "main" plots/otherwise contrive and railroad.

I personally like running the "you can do whatever" kind of game like you are. But I've gone through this with groups before, and they don't want to know they're being railroaded, but they want you to do it.

My current Pathfinder campaign was being run very sandbox, and in one session frustration with it came to a head. As part of the post-mortem, one player said "Can you just tell us if there's nothing more to be found somewhere or if something's not important, because we feel like we're banging our heads against stuff and aren't sure if there's just nothing there or whether we're failing at something, and so we take that as we're failing."

A more subtle approach is to not have all those little leads be around forever. Maybe someone else does them, maybe the problem is over, maybe another one of the quests is "on a timer". (Or even worse, one of the bad/important problems goes to hell while they're messing around with collecting boar parts.) "Our village has a plague! Never mind, we're better now." This can help clear the backlog and also help set the expectations that not everything is world-shaking or needs the PCs, and they should engage and discover which a lead is before setting off on it.

It may be that the types of characters that your players have created are not suited to this style of adventuring. Zak wrote an excellent blog post about this: Sandboxes And The Roguish Work Ethic

A hook isn't automatically a hook for a bunch of lovable rakes:

"A cleric has been found dead in the town square."

"Well why should we care?"

"The church is offering a reward of 600 gold pieces to find the killer."

"Um, couldn't we just sack the church and make more than that? I mean, who was this cleric anyway? Maybe he deserved it..."

When you're a thief, the world is your sandbox. When you're an Epic Hero, it's a big fire house you sit around in waiting for a fire.

Now, "rogue" here refers to an attitude, rather than any specific character class or game mechanic. These characters are highly self-motivated and proactive. Using their knowledge of your game world, they devise their own schemes for seeking wealth and fame. Usually these schemes are disruptive to the status quo, thus the "authorities" seek to stop them.

(the NPCs are then reacting to what the player characters do, rather than the usual quest where the PCs act at an NPC's behest or respond to an external event in the game world)

If both you and your players still have the goal of an enjoyable, sandbox-style game, then I can see two possibilities:

  1. Roll up new characters that have backgrounds and motivations that are more conducive to raising hell.
  2. Evolve the existing characters into more proactive denizens of your game world.

Now, many characters will evolve in this direction over time. As they become more powerful, they are more confident in their ability to make a difference with their actions. In the course of their adventures, they learn more about your game world and thus have more information to decide what to do. They should also become personally invested in your game world: pay attention to their favorite NPCs and locations (as well as NPCs whom they despise).

If you are going to introduce a plot hook, make it affect the player characters personally. Give them a reason to care. Perhaps their favorite NPC is kidnapped, or their local drinking hole is going bankrupt. That is one of the main advantages of tabletop roleplaying: you can tailor the adventure to suit the adventuring party.

One plot hook that I have used to great effect in my own sandbox campaign is from The Maze of Screaming Silence by James Thomson: one of the player characters has recurring nightmares. In their dreams, they see the location of the major story arc that you want to introduce. Of course, you also need to come up with a solid reason why that specific person should be having nightmares. Hopefully, it will pique their curiosity a little.

  • 1
    I never thought about the kind of character that works well in a sandbox (and since I've played a few Superman types wiht a bunch of rogues, it also explains why I always feel bamboozled sooner or later). – Pulsehead Jan 14 '11 at 14:39

I just read Vincent Baker's Apocalypse World RPGGeek, and like all of his games it contains more than one brilliant thing. But for this case, the brilliant thing I'm going to suggest that you take a look at is the Countdown Clocks he created. Without directly stealing the clocks (which I intend to do for my own games, even outside Apocalypse World), you could take the major actors in the world (rain of fire, new dungeon) and create progressions for what will happen if the PCs don't intervene, and / or what might make them move along the progressions.

Just an off-the-top of my head example for "rains of fire":

  1. Rain of Fire appears in the sky
  2. If PCs do nothing, villages begin to burn
  3. If PCs do nothing or fail, the richest forest in the land is set on fire
  4. If PCs do nothing or fail, the forest is consumed and the world is blanketed in darkness and ash
  5. If the PCs do nothing or fail, the kingdom is plunged into n years without summer and the frost giants move down from the utmost North...

This isn't exactly how the clocks work (his have much more awesome ladled over them), but it puts a flashing arrow into the player's picture of the world saying "BIG PLOT HERE!" and if they choose to ignore it, or suck at helping, you know where the consequences are headed.

Also - if your "giant dungeon rediscovered" is progressing on "If the PCs do nothing..." then you've presented an interesting dilemma - which part of the world do you want to save?

What if one PC has peasant roots and his ancestral village will be destroyed by the fires and one has royal roots and his throne-to-be is what will be destroyed by what's crawling out of the dungeon? Or what if they PCs are strangers and just have to choose between saving the peasants and saving the king?

I'm running a sandbox campaign where there are a few major events taking place, but not set expectation that the players have to take part in them. ... Major leads like fire falling from the sky or the rediscovery of a massive underground dungeon.

Have the major events advance while they play around with the side quests. Have effects from those events start affecting them indirectly and then directly:

  • Refugees fleeing from the fire storm swarm a city they're in.
  • Monsters bleeding out of the dungeon have destroyed a PC's hometown, kidnapped someone important, killed the NPC they're doing a quest for, etc.

If they're still not interested, have the events play out to their terrible conclusions.

It can be difficult in sandbox games for players to know what the characters are "supposed" to do. You may think the answer can be "anything they want", but in reality there are going to be things that work well and things that don't work out so great.

So my advice is to try to have the players create characters with strong motivations, as then they will tend to ignore things that don't fit with their drive for revenge, or wealth, or prestige, etc. and you'll have a better idea of how to create bits of sandbox that will engage their interest. May not be so useful to you this time, I guess, but you could bear it in mind next time!

You need to give the PCs some context in which to make decisions. Basically by defining their character backgrounds. Work with them on a one on one basis to craft the background to fit the campaign and to give them reasons for the party to be together.

By doing this they have a way of relating to the larger and smaller events of the game. If it came out that Abe the Wizard was raised in the City State of Isindur and five months into the game the party hear about riots in Isindur. There is good chance that Abe may try to convince the party to go check it out.

It also allows the campaign to have hooks into the characters and characters into the campaign. These can be a source of benefits or complications. Above they serve as a source of adventure.

Because you have an existing campaign you want to look at anything the players have done with the characters and use them as a skeleton to create their background. You will need to keep mind what they did in the past so the new material stays from ret-conning the past.

Again a sandbox campaign doesn't work unless the players having something to base their decision on. Otherwise they might as well roll the dice to see which direction to go or events to investigate.

Every lead you give is a promise something is going to happen. So, when you say "There's an interesting cave", you are promising the players there is something good within.

So, when you give leads, make sure you can back them up with something interesting. Conversely, hold back on the "minor" leads, because the players will think they lead to major things.

There shouldn't be 'major' or 'minor' plot points in a sandbox campaign, in my experience. Spread as much interesting information around as you can (keep a list somewhere) and then try to improvise something for whatever the players decide. You can make all the leads only point to a handful of major causes, which makes it easier to keep an overview of the region (e.g. the large number of orc raids, the large numbers of wolves, and the theft of the local mage guild's mini McGuffin are all instigated by the same evil warlord).

This gives the impression of a large, varied world and since you only have a handful of major factions to handle, you can keep a folder of dungeons and NPCs and drop them into any suitable situation the players find themselves in, while creating the story off-the-cuff. Let the players make the story - this helps keep them invested in your world. When they look back and see that all the problems they've dealt with were down to a covert war between two rival Thieves' Guilds, one of which they've helped win, you can quietly smile behind your screen and know that your world is steered by the players, rather than it steering them, and they'll know too (hopefully).

I've had players make INT checks, I call them "hunch" checks or intuition checks. Somebody's going to roll not-suck and I'll tell them "you get a feeling that the filigree on the clock face is a red herring."

Then I'll try to work it back into the main plot a few sessions later, and then use another INt check to point back at it. "While you're talking about the Dragon's Sundial you remember that clock face with the filigree."

  • If it's intuition, shouldn't it technically be Wisdom not Int? ;) – LeguRi Jan 12 '11 at 19:58
  • 1
    Depends on your definition. I think of like, Kantian intuition which is the application of your whole faculty for processing information. Call it a smarts check. – Ry St Jan 14 '11 at 15:32

Your Answer


By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.