All my previous GMing experience is with tightly-plotted railroads, and my players are used to (and generally happy with) riding the rails. Right now I'm setting up a sandbox campaign to break out of those habits, but I don't know how to prevent my players from treating elements of plot as the story and jumping on the non-existing Plot Train as soon as I introduce anything.

Specifically, I want to have some active villains in this sandbox. (There are 21 entities based on the Major Arcana, plus a prophecy saying that seven of them would end the world with the help of the rest, even though some of them aren't inherently evil.) I want to keep them fairly background to start, but I know that as soon as the players catch wind of the mere existence of a villain they'll drop their own plans and everything else to follow "the story" where they think I must be leading: to go fight that villain.

How can I have villains in my sandbox without my players assuming they are "the plot"? I want my players to follow their PCs' own built-in motivations and subplots for the most part, and not accidentally railroad them into a plot about saving the world just because they are the player characters. I don't want them jumping on every wagon, thinking that it's the Plot Wagon, when it's really just a plain ol' wagon. Especially because I know that if they jump on a wagon, I'll jump on it too and it will become just another railroad-plot game like I've always run before.

  • 2
    \$\begingroup\$ Do you provide any reason whatsoever for them to not go after the villain they know about? If you start a new game of Skyrim, and someone tells you a thief in a specific cave stole their jewel, and you don't have anything else to do yet... \$\endgroup\$ Aug 1, 2014 at 19:07
  • 3
    \$\begingroup\$ Given that this is a departure from your regular style, the style they're used to, make sure you let them know that you're designing this campaign to not be "on rails" and to serve the interest of the characters rather than necessarily those of the world. \$\endgroup\$
    – Doc
    Aug 2, 2014 at 4:22

5 Answers 5


Tell Them Your Goals

If you haven't already, I would start by telling them essentially what you just said here. That there is no "one true plot". Tell them that introducing an evil person / problem does not make it the overriding campaign unless they want it to be. Tell them that you are willing to follow along with their character's background goals.

If they want to play that style of game, there is a good chance that will fix a lot (but not all, habits are hard to break) of it right there.

But its worth remembering that not everyone wants a sandbox. I prefer to play in games with good plots with well laid out characters. While there are some exceptions, GMs frequently are willing to invest more time into plotting and developping good NPCs if they are assured the PCs will focus on them. For that reason, I am perfectly happy to with what you might call "light rails". I want some character autonomy, but I am perfectly happy to follow "the one true plot" if it gets a better story and better NPCs.

By telling them your goals directly, they will either agree to go along, or directly ask you to put some rails back and then you can choose as a group how much of the game should be on rails.

Drop many plot hooks, make them choose

In the game itself, one way to force them to avoid rails is to drop so many potential plot hooks they have to choose which ones to focus on. You can even show that the world is a living breathing place by showing them the resolutions to the plot hooks they didn't take.

For instance, they sit in a bar and overhear of some great evil. They may be tempted to go after it immediately, and there are no rails, so they can! But then someone comes up and asks them for help urgently. And I don't mean Skyrim style urgently where everything is urgent but will wait for you without ever changing, give them an actual deadline measured in something like game days.

Now, they have a choice. Go after the big bad or help this person? In a game on rails, the answer is first help then go after the big bad. Obviously, the person asking for help is just there so they can level up. But you can break that by making it clear the big bad isn't standing still. If they go after him now, he won't have completed his McGuffin of Death and they can stop it before its an issue. But if they wait, they will have to deal with him and the McGuffin. Now, they have a real dilemma and rails won't help them.

To take it even further, highlight something unresolved or some goal on the character sheet. Point out that the world isn't standing still and any time they spend either helping this person or going after the big bad won't help them deal with that personal (to the character) matter. It too may get harder if they put it off.

Time is a great limiting factor in real life, it can be in games too.

Now, they have choices to make and no rails to help, they really have to choose. That is the first and biggest step.

After they get used to dealing with that, especially if you keep highlighting personal goals from the character sheet (and insist they add some if there aren't some obvious ones) they will eventually start adding more of their own, like becoming a dragon slayer. Then they will start actively trying to achieve those goals without further prodding.


How to run this, procedurally

Consider this - every week, the players show up and they manage to improvise and play, without having to preplan every "if this happens, then I'll do this". They simply look at their character sheet and improvise based on a basic understanding of their character, right?

As a GM, you can do that too. Set up your characters, give them motivations, simply look at the list of motivations and pick a character or two who either would take an interesting action or reaction to what has happened and set scenes around that.

Second, if your players give you a short list of 3-4 things they would like the story to revolve around, personal goals of their characters, issues, etc. you basically look for the intersection of their goals with the motivations of the NPCs.

"I see Bill's character really wants to find out what happened to his missing brother, and he's also trying to clear his name of a crime he didn't commit. NPC X has been starting a revolution, over here... It'd be cool if his new henchman IS Bill's PC's brother, and he also believes that he DID do the crime and is feeling all torn up about what his brother has become..."

You can usually sit down for anywhere from 10-30 minutes between sessions, look for connections like this, and then improvise during play for the rest.

Communicating with the Players

Now, I've had the situation you talk about, where players are looking for clue-directions as to where the "plot" is. The solution is simple - tell them exactly how the game works and remind them.

"There is no plot. I have no specific thing planned - the NPCs have their own goals, and you have your own goals. They're acting and reacting just as much as you are acting and reacting. If you find a clever way to solve your problems and stop NPCs from doing things you don't want to happen, great! Write down your (Flags, goals, etc.) and tell me what you want to focus the game around. Change them if you decide you want to change focus."

If players seem lost, which may happen a lot early on, remind them - "Hey, what's on your list of goals? That should help remind you what you're after, right? If it turns out you're not that interested in it, maybe this thing X over here might be more what you want to pursue?"

As you're doing all of this, don't make it a long, terrible clue trail to get to those issues - go for them early and consistently. Don't worry that you're "using up the good stuff" - more cool stuff will come out of it. You may want to go through the Same Page Tool to have a clearer idea of what play will look like.

  • 2
    \$\begingroup\$ Also, specific to your game - don't try to track all 21 Arcana NPCs right off the bat, that's too much. Pick like 3-5 NPCs and start there. \$\endgroup\$
    – user9935
    Jul 31, 2014 at 22:00

There are a variety of ways to do this:

Focus on the prophesy itself. If you cast doubts on the validity of the prophesy, the players may be more likely to leave it alone. For example, well-respected representatives for the forces of good declare that some of the named parties in the prophesy couldn't possibly be involved in something nefarious like that. Accusations that the whole prophesy is intended to sow discord start to gain credence. Or perhaps people dismiss the prophesy out of hand because of its source, a disreputable party that has repeatedly told of things that never came to pass.

Focus on what others are doing. The deadliest, most powerful heroes in the land have heard of the prophesy and are acting to thwart it. Two or three of the supporting bad guys have already been killed (of course you know they haven't really - rumors are so often wrong). The conditions necessary for fulfillment of the prophesy have already been negated. Nothing to see here, folks. Then, much later, when you are ready, the player characters encounter one of the supposedly dead henchmen. They dig deeper and find out that the heroes never really did wipe out the threat. Were any of the heroes in on some sort of conspiracy? Hmm... .

Keep adding to the gumbo. In my experience the most important thing is to push the prophesy into the background gently, without calling much attention to it. While guarding the Queen during the royal wedding, the player characters find out in passing that the Lord of Greenvale just returned from a battle against several of the villain's henchmen, where he defeated two of them in combat. In a tavern one night the PCs hear a pair of wizened sages chuckling about the latest crazy prophesy being spread around the kingdom. Sprinkle these elements in every once in a while so none of them stand out, and be sure there are plenty of other threats, prophesies, and rumors being spoken of, so the players get the sense that every bit they come across isn't supposed to be a call to action.


The simplest solution is to build them up over time and introduce them along with other conflicting requirements at the same time.

For example my current sandbox campaign has a were-rat bad guy. He was only active on the full moon (at least at first) and I very deliberately never introduced him in person (they still haven't encountered him, although they have run into a few minions). Because of this he's become part of the game, the players try and figure out his plots, foil a few here or there but he isn't all of the game. They go and do other things too. Other things that I introduced first and that they also want to progress on...

Some times he's more central to the plot than others, and there might be a few missions in a row related to him, other times other plots rise to prominence.

But the players realise that sometimes all they can do is wait.

For example if they work out that the bad guy will be at a certain point in 2 months time then for sure - they can just wait for 2 months, but its far more likely they'll go and do other things for a few weeks and then get in position.


I ran a campaign a few years ago that was pure sandbox. Only one vague goal was the plot (Get Home). Basically I had a gigantic word document that listed every place, who lived there, and what the players did there last. I would then simply update this document before and after the game, and also sometimes while the players bickered about what the one true plot line was.

The learning moment was when they finally went back to a missed hook from a few (real-time) months prior. I pulled that dusty entry, saw a few ingenious updates I had made, then launched them into the biggest fight ever -> The badguys got away with 'it', and had way more power than the players were accustomed to in this region. They quickly learned to take time in every location and weigh their options, not just go for the most lucrative/plot-like one.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .