Session zero: how to run it

Every campaign that quietly dies in the third month dies for the same reason: nobody agreed what game they were playing. Session zero is the cheap, unglamorous meeting that prevents it. No dice, no plot, just the table getting honest before the story can betray them.

What a session zero actually is

A session zero is the meeting you hold before the campaign begins, where the whole group decides together what they're signing up for. Not the world's lore, not the opening scene. The agreement underneath all of that: the tone you want, the lines you won't cross, how often you can realistically show up, and who these characters are to each other. It feels like overhead. It's probably the most useful hour you'll spend on the whole campaign, because everything after it inherits the assumptions you either checked or didn't. The whole point is that the next thirty sessions are play for the whole table, not a quiet tug-of-war over what game you're even in. Skip it and those assumptions are still there; you just won't find out they conflict until session eight, when one player wants gritty political tragedy and another keeps making puns at the funeral.

The campaign won't tell you who it's for. You have to decide that out loud, while you still can.

Expectations and tone

Start with the question nobody usually asks: what is this campaign for? A heist-comedy where everyone's a lovable disaster and a slow-burn tragedy about doomed loyalty are both great games. They're just not the same game, and a table accidentally running both at once will tear quietly down the middle.

So name it. Talk about the tone you're after, how serious the table wants to be, how much combat versus intrigue versus downtime, and how lethal the world is. Can a beloved character actually die, or is this a story where the heroes are meant to win? Then cover the un-fun-but-essential stuff: table talk versus in-character focus, phones, whether jokes break the mood or make it. None of this is a contract. It's a shared picture, and the point is simply that everyone is looking at the same one.

Safety tools: lines and veils

You're going to improvise a story with friends, and improvisation occasionally walks straight into something a person at the table did not want to be in the room for. Safety tools exist so that's a normal, low-drama course-correction instead of a ruined night.

The classic pair is lines and veils, a framing the hobby owes to designer Ron Edwards. A line is content the game won't include at all: it doesn't happen, full stop. A veil is content that can exist in the fiction but happens "off-screen", acknowledged but never depicted. Set a few of each during session zero, make clear anyone can add one later without justifying it, and agree on a quiet signal to pause or skip if a scene heads somewhere uncomfortable. The widely used X-Card, designed by John Stavropoulos, is exactly that signal boiled down to a single index card. This isn't about making the game softer; plenty of tables run brutal, harrowing stories. It's about knowing where the edges are so everyone can play hard inside them, trusting the group to stop on request.

The campaign pitch

Now the GM gets to sell the thing. The pitch is a short, vivid promise of what kind of adventure this is, a paragraph or two, not a worldbuilding dump. Where are we, what's the trouble, what role do the heroes play, and what makes this premise worth six months of Thursdays? "A frontier town on a dying coast, a debt the party can't repay, and a sea that keeps giving back the wrong dead." That's a pitch. It tells players what their characters should care about before they've made them.

A good pitch does double duty: it excites the table and it constrains character creation in the helpful direction. If the campaign is about a struggling river guild, the lone-wolf assassin who hates groups is going to fight the premise every week. Pitch first, characters second, and let the pitch quietly answer "what kind of person belongs in this story?"

Characters tied to the story, and to each other

The most common session-zero miss isn't bad characters; it's disconnected ones. Five strangers with no reason to travel together, hand-built in isolation, and the GM spends the first three sessions inventing contrivances to keep them in the same room.

Build characters at the table, together, against the pitch. Two questions fix most of it:

• Why is your character in this story? A hook into the premise (a debt to that river guild, a sibling lost to that wrong-dead sea) so the campaign already has a grip on them before session one.
• How do you know someone else at the table? One shared thread per pair: a debt, a history, a secret, a rivalry. A web of small connections turns five PCs into a group the GM doesn't have to glue together.

Settle the practical layer too (system, starting level, allowed options, party balance), but the relationships are the part that pays off for months. A party bound by history survives the boring sessions. A party of strangers needs the plot to constantly justify their teamwork.

Logistics: when, how often, where

The least romantic part of session zero is the one that actually kills the most campaigns. A brilliant premise and a beloved cast die anyway if four adults can never find the same free evening.

So decide, concretely: how often you play (weekly, biweekly, monthly), how long a session runs, what counts as quorum (does the game go ahead with one player missing, or not?), and how you'll reschedule when life happens. Pin down whether you're at a table, on voice, or hybrid, and who's responsible for the calendar. These answers feel trivial in the room and turn out to be load-bearing: the campaigns that last aren't the ones with the best plot, they're the ones that reliably meet. (Working out a recurring slot that fits everyone is exactly the kind of coordination Kompanion's scheduler exists to take off your plate.)

A session-zero checklist

Run it in roughly this order. You won't finish every line in one sitting, and that's fine. Split it across two evenings if you need to.

• The pitch. GM presents the premise, tone, and what makes it worth playing.
• Tone & expectations. Seriousness, lethality, combat vs intrigue, table-talk norms.
• Safety tools. Set a few lines and veils; agree the "pause" signal; make additions easy and judgement-free.
• System & characters. Rules, starting level, allowed options, party balance.
• Story hooks. Every character gets a reason to be in this campaign.
• Party bonds. At least one shared connection between each pair of PCs.
• Logistics. Frequency, session length, quorum, rescheduling, where you play.
• Say it out loud. If the table never actually discussed it, it isn't settled, no matter what one person assumed. The agreement is only what the group said together.

That last line matters more than it looks. The whole session zero is a set of decisions, and decisions evaporate. The table that "agreed on tone" in March is genuinely surprised by it in May. So once you've settled something out loud, put it somewhere everyone can re-read, a shared Google Doc, your Discord, whatever the table already opens, and the same habit pays off once the actual campaign starts and there's a whole world to keep straight.

After session zero: keep the agreement alive

Session zero is the start of a shared memory, not a one-off ceremony. The lines you drew, the bonds you agreed on, the slot you booked: they only help if the table can still find them in month three, by which point everyone's recollection of "what did we agree?" has quietly drifted. You don't fix that by writing more, you fix it by keeping the few things that matter somewhere the whole group can re-read, a Google Doc or your Discord, instead of in five separate heads. The hour you save chasing "wait, did we say lethal or not?" is an hour back at the table for the things only you can do: the next twist, the next voice, the next map.