How to Run a West Marches Campaign: Complete D&D Guide

Your D&D group has eleven people in the group chat and roughly four of them can ever be in the same place on the same night. Somebody starts a poll. The poll dies. Three weeks pass. Eventually someone asks, gently, if the campaign is dead.

West Marches was built for exactly this group. It drops the idea that a campaign needs one fixed party on one fixed night, and replaces it with a shared world that a rotating cast of players drops into whenever they're free. No poll. No "who's in tonight." Just a hub, a frontier, and a standing invitation to go see what's out there.

It's not a new idea. Ben Robbins coined the format back in the mid-2000s, and it's had a resurgence lately, with Critical Role's current campaign built around the same rotating-table structure. Here's how to actually run one, and the one problem the format never quite solves on its own.

How to Run a West Marches Campaign (The Short Version)

Build a hub, not a plot. A safe home base where nothing dangerous happens and every session starts and ends.
Draw the frontier in layers. A handful of wilderness regions with a few named points of interest each, expanded only as players explore them.
Recruit a big pool, expect a small table. 10-20 players signed up, 3-6 showing up on any given night.
Run a mission board, not a story arc. Rumors and hooks players choose from, not a plot you push them through.
Treat every session as self-contained. Whoever shows up is the party. No waiting on absent players' characters.
Solve the DM-availability problem separately. The format fixes player scheduling. It does not fix the DM being one person with one calendar.

What a West Marches Campaign Actually Is

Strip a West Marches campaign down to its parts and you get three things: a hub, a frontier, and a pool of players bigger than any single table.

The hub is a town, a keep, a trading post, anywhere safe. Nothing dangerous happens there. It's where sessions begin and end, where players spend downtime, sell loot, and pick their next job off the board. The frontier is everything outside it: dangerous, unmapped in most places, and the actual subject of play.

The player pool is the part that makes the whole thing work. Instead of a fixed party of four or five, you recruit ten to twenty people who all have characters living in the same hub. Any given session, whoever's free that week fills the table, usually three to six players. Nobody's character advances unless that player actually shows up and plays them. There's no absent-party-member problem, because there's no fixed party to be absent from.

This is why West Marches solves the exact failure mode a normal D&D campaign runs into: everyone needing to be free at once. Scheduling collapse is one of the most common reasons D&D groups quietly stop meeting, and a rotating cast sidesteps it entirely. You don't need everyone. You need enough people, tonight, to go on an adventure.

Build the Hub First

Most new West Marches DMs want to start with the map. Start with the hub instead.

The hub is where every session opens and closes, so it needs to actually work as a location players want to return to, not just a loading screen between dungeons. Give it:

A handful of recurring NPCs. A quartermaster who buys loot, a tavernkeeper who trades rumors for drinks, a scribe who tracks the mission board. Three or four is plenty. Players will form opinions about them fast.
A reason to linger. A market, a temple, a faction with its own agenda. Somewhere a session can end early without feeling like it just stopped.
A visible sense of stakes. If the frontier is dangerous, the hub should occasionally show it: refugees arriving, a bounty board that changes, a shrine to adventurers who didn't come back.

Keep the hub small on purpose. A West Marches hub that's trying to be its own mini-campaign steals prep time from the frontier, which is where the actual game happens.

Draw the Frontier in Layers, Not All at Once

The single biggest prep trap in a West Marches campaign is trying to build the whole map before session one. Don't. Build three to five wilderness regions, each with a theme and two or three points of interest, and expand outward only as players actually go explore.

A point of interest doesn't need much: a name, a one-sentence hook, and a monster or complication if the party pokes at it. "The Sunken Chapel, half-submerged in the bog, bells still ring at low tide" is a full point of interest. You don't build the chapel's interior until a party actually walks up to the door.

This layered approach does two things. It keeps your prep time proportional to what's actually getting played, and it keeps the map feeling alive rather than fully known. Players comparing notes back at the hub, "did you find the thing past the ridge," is some of the best organic marketing a West Marches campaign gets, and it only works if the map still has real gaps in it.

The Mission Board: How Players Choose What Happens Next

A West Marches campaign runs on player-driven exploration, and the mission board is the mechanism that makes that real instead of aspirational. Keep a running list, physical or digital, of:

Rumors. Half-confirmed leads. "Something's been raiding the northern farms" might point at a point of interest you haven't detailed yet, which is fine.
Standing bounties. A merchant wants an escort, a noble wants a relic, a farmer wants their goat back (that one's never as simple as it sounds).
Open threads from past sessions. Something a previous party found but didn't finish. This is how continuity survives a rotating cast: the world remembers even when the party doesn't repeat.

Before each session, whoever's playing picks a job off the board. That's the whole prep loop. You don't write a plot. You write a world with enough hooks in it that a plot emerges from what players decide to chase.

⟡

Running Multiple Small Groups Instead of One Fixed Party

This is the part that trips up DMs coming from traditional campaigns: you're running a shared setting for several small, mostly-independent groups, not one campaign.

A few habits make this manageable:

Cap party size at four to six per session. Bigger tables slow down in a format that's already juggling a rotating cast.
Keep a simple world-state log. A shared doc noting what's changed: this bridge is down, that bandit camp is cleared, this NPC is now hostile. Ten minutes after each session, not a full session report.
Let characters level individually. XP tends to fit the format better than milestone leveling, since a session is a self-contained trip rather than a chapter of one continuous story, and it rewards a player for the risk they took that specific night regardless of which region their table happened to explore.
Resist syncing every table's timeline. Two groups can be in the same week of in-world time without ever meeting. Don't force overlap that isn't fun to track.

Imagine two different Thursday groups both push into the same forest region three weeks apart. Neither table needs to know what the other did in detail, just that the world remembers: the bridge one group burned is still burned when the second group gets there. That continuity, without either table needing the other's session notes memorized, is the whole trick.

The Real Problem With West Marches: It Still Needs a DM Every Session

Here's the part most West Marches guides gloss over. The format solves the player-scheduling problem beautifully. It does nothing about the DM-scheduling problem, because there's still exactly one of you.

A rotating pool of twenty players means there's almost always a table's worth of people free on a given night. It does not mean there's always a DM free to run for them. If you're the only DM, every session still runs through your calendar, and a popular sign-up sheet can create more DM demand than a normal campaign, not less. DM burnout doesn't disappear just because players stopped needing to sync schedules with each other.

The usual fixes are real but partial. Co-DMs split the load but introduce inconsistent pacing and tone between tables. A strict "one session per week, first-come sign-up" rule caps your exposure but also caps how many of your twenty eager players actually get to play. Neither one changes the underlying math: player availability scaled up, and DM availability didn't.

Where an AI Game Master Fits Into a West Marches Campaign

This is the gap StoryRoll was built to sit in, not as a replacement for your human-run West Marches nights, but as the thing that runs when your calendar is the bottleneck and a trio of players just wants to go explore.

Picture a setup where the hub, the region map, and the open mission-board threads are the shared canon your human sessions run in. On a week you genuinely can't run, three of your player pool still want to push into the ruins past the river. Instead of that slot going dark, they run it through an AI dungeon master instead: same hub, same frontier, same open threads. What they find, spend, and burn down folds back into the same world-state log the human sessions use.

StoryRoll's D&D mode keeps full classes, races, and 5e mechanics, so a session run this way still feels like the same game, not a lighter thing bolted onto the side. It doesn't replace what a human DM brings, the voice, the callbacks, the read on your players. It just means the sign-up sheet doesn't sit empty on the weeks your calendar says no.

If you're just getting the format off the ground, a smaller playtest campaign with your core group first will surface hub and map problems before you open sign-ups to a bigger pool.

West Marches Mistakes to Avoid

Overbuilding the map before session one. You'll re-theme half of it once players actually start exploring. Build light, expand as they go.

Starting the player pool too small. Five interested players isn't a West Marches pool, it's a normal campaign with extra steps. You need enough people that a full table is likely most weeks even with typical no-shows.

Forgetting the hub needs personality. If the safe town is just a shop menu with a signpost, players will rush through it instead of treating it as part of the game.

No consequence for danger. If dying or failing has no real cost, the frontier stops feeling dangerous and starts feeling like a vending machine. West Marches works best when the map genuinely bites back.

Running it exactly like a normal campaign with more scheduling flexibility. The format's power comes from player-chosen exploration and a persistent, reactive world, not just from having a bigger group chat. Skip the mission board and you've mostly just built a normal campaign with commitment issues.

Frequently Asked Questions

What is a West Marches campaign?

A West Marches campaign is a D&D format with no fixed party and no set schedule. Players share a safe hub town and a dangerous frontier map, sign up for whichever session fits their calendar, and each session is a self-contained trip into the wilderness with whoever showed up that night.

How many players do you need for a West Marches campaign?

A healthy player pool runs 10 to 20 people, with 3 to 6 actually showing up to any given session. The pool is large so the DM always has enough interested players to fill a table, even though most of them won't play in any single week.

Do West Marches campaigns use XP or milestone leveling?

Most West Marches games use XP, because sessions are self-contained trips rather than chapters in one continuous story, and XP rewards showing up and taking risks regardless of which subplot a given table happened to explore that night.

Can you run a West Marches campaign with an AI Game Master?

Yes, and it solves the format's core weakness: a West Marches campaign needs someone available for every session a group wants to run, which is exactly the constraint an AI Game Master doesn't have. StoryRoll can run a hub-and-frontier structure on demand, so a trio who wants to explore tonight doesn't have to wait for the human DM's calendar to line up.

What's the difference between West Marches and a regular D&D campaign?

A regular campaign follows one fixed party through a continuous story with a set schedule. West Marches drops both: the party rotates session to session, there's no recurring night, and continuity lives in the shared world and mission board instead of in a single unbroken party narrative.

How big should a West Marches map be?

Start smaller than feels ambitious. Three to five wilderness regions, each with its own theme and two or three points of interest, is enough for months of play. You can always extend the map outward once the hub and the mission board are working.

Try These Free Tools

Encounter Calculator: balance a frontier encounter on the fly for whatever party size shows up that night.
NPC Name Generator: keep the hub and mission board stocked with names without stopping to invent one mid-session.
XP Calculator: track individual leveling across a rotating cast instead of a single party.

The Verdict

West Marches is the best fix available for D&D's scheduling problem, because it stops asking everyone to be free at once. But it quietly trades a player-scheduling problem for a DM-availability one: the format only works on weeks you can run it. Build the hub with real personality, layer the map in as players explore, and run the mission board as the engine instead of a plot outline. Then solve the part West Marches doesn't: what happens on the weeks the sign-up sheet fills up and the DM's calendar doesn't. That's the gap worth filling, one way or another, before a popular West Marches campaign turns into the same burnout problem it was built to avoid.

How to Run a West Marches Campaign (Without Burning Out the DM)