Blog / Fifth edition rules
Duet D&D: running a campaign for one player
Duet play is one Dungeon Master and one player, running a full campaign together. Everything in D&D 5e works in that configuration except the combat maths, which was written for four characters and has to be adjusted by hand. That is the entire summary: it is a better format than most people expect and it has one real structural problem.
The format has been acknowledged in an official product - the 2025 anthology Dragon Delves flags several of its adventures as runnable for a single player, and offers an optional supernatural gift, the Blessing of the Lone Champion, that gives a lone character temporary hit points and more frequent Heroic Inspiration. That is a small thing on its own, and a useful signal that the adjustment below is expected rather than a house rule you are inventing alone.
Why duet is one of the best formats there is
Pacing. There is no table to poll. A scene that would take twenty minutes with five people takes three, because nobody is waiting for their turn to talk and nobody is on their phone. A duet session covers something like twice the ground per hour, which is why a duet campaign can tell a story with an actual middle rather than dissolving after the second dungeon.
Character depth. This is the real reason to run one. With four players, a GM has to ration attention: every backstory hook competes with three others, so most of them stay unused. With one player, the campaign can simply be about their character. Their sister, their debt, their old order, the thing they did before session one - all of it is available, all of it can pay off, and none of it has to be balanced against anybody else's screen time. Players who have only ever played in groups tend to be startled by what that feels like.
Scheduling. Two adults can find an evening. Five cannot. This is not a small point, it is the reason most campaigns die, and duet play is close to immune to it. A cancelled duet session is rescheduled by one text message.
Range of tone. A duet campaign can be a quiet political game, a slow horror story, a single-character tragedy. Those are hard to run at a full table because somebody always wants a fight, and correctly so. With one player you only have to satisfy one appetite.
What genuinely breaks
The encounter maths. Every difficulty guideline in fifth edition assumes roughly four characters. Run a "medium" encounter against one character and it is frequently lethal, because damage scales with the number of enemies while the target pool does not.
The action economy. This is the sharper version of the same problem, and it is the one that actually kills people. Three enemies against a party of four is three attacks against four. Three enemies against one character is three attacks a round on one body, and one action a round back. The player loses roughly three quarters of their tactical options at the same time: nobody flanks, nobody grapples the archer, nobody draws fire.
No party skill coverage. A party covers its own gaps by accident. A single character has whatever it has. A duet rogue simply cannot heal, a duet paladin simply cannot pick a lock, and if the adventure has a locked door on the critical path, the campaign stops at the door. This is a design constraint on you, not a flaw in the player's build.
The solo character death problem. In a party, dropping to zero is a setback: somebody stabilises you, somebody drags you out. Alone, dropping to zero is usually the end, because there is nobody to intervene and death saves against a monster that keeps attacking are not really saves. That changes the emotional weight of every fight - and one bad initiative roll ending a campaign built around one character is a worse outcome in a duet than it would be at a table where the story continues.
The fixes, in order of how much they help
Give the player a companion. This is the highest-value change available and it fixes three of the four problems at once. A sidekick NPC, a hireling, a loyal beast, a bound spirit - anything with a stat block and its own turn. It restores the action economy, covers a skill gap you choose deliberately, and gives someone the ability to stabilise the player at zero hit points.
The design rule is that the companion should be clearly less capable than the player character and should never make the decisions. It exists to absorb attacks and pass the healer's kit, not to solve problems. The best duet companions also have an opinion, because they double as the thing a solo player misses most: somebody to talk to in character.
Reduce enemy numbers before you reduce hit points. The instinct is to halve a monster's hit points and call it balanced. It is not, because the thing threatening the player is the number of attack rolls landing on them each round, not the durability of any individual enemy. One tougher enemy is a fight. Four weaker ones is a coin flip decided in round two. Cut the count first, and adjust hit points only after the count is right.
Give more consequence, not more combat. With one player, every choice they make is load-bearing, so let it land. The NPC they were rude to in session three should still be hostile in session eleven. The faction they helped should send help unprompted. This is trivially easier in a duet, because there is exactly one thread of causation to track, and it is the strongest thing the format has going for it.
Lean on the exploration and social pillars. Fifth edition combat is the part that needs the most repair for one character; investigation, negotiation and travel need essentially none. A duet campaign that is sixty per cent talking and exploring and forty per cent fighting plays better than the usual ratio, and it is not a compromise - it is the format playing to what it does well.
Use downtime deliberately. With one character there is no scheduling problem in letting weeks pass. Training, crafting, building a reputation in a town, tending a relationship: all of these give a single character somewhere to develop that is not the next combat encounter, and they are how a duet campaign gets a sense of a life rather than a sense of a quest.
If you are the player in a duet game
Two things are worth doing. Build for survival rather than for a niche: you have no one to cover a weakness, so the durable, self-sufficient end of the class list serves you better than the specialist. And do not wait for the GM to ask about your backstory. In a duet the campaign will bend toward whatever you show interest in, faster than you expect, so say what you want the story to be about. It is the one format where saying so is likely to work.
Where an AI Game Master fits
Not as a replacement for the person you are already playing with. If you have a friend who will run a duet campaign for you, that is the best version of this available and nothing here competes with it.
Where it does fit is when there is no one to run it. An AI Game Master runs the same one-player shape on demand, and the two structural problems above are handled in software rather than by remembering to handle them. On Rolegend the split is explicit: Keystone Engine enforces the rules and is the only thing permitted to change a number, rolling on the server, tracking hit points, conditions, spell slots and inventory. ROWAN narrates the world, is handed the result, and cannot ask for a different one. The honest gap is tactical positioning - grid movement, cover, opportunity attacks and flanking assume a battle map, and that is published rule by rule at /rules-coverage rather than left to be discovered mid-fight.
If a duet is what you want and the problem is simply that nobody will run one, the full set of options - oracles, solo-first systems, running a module yourself, and this - is laid out at playing without a DM.
What it will not do is the thing a good duet GM does best: notice, over months, which parts of your character you actually care about, and build the whole campaign around them on purpose. Consistency is not intention. If you can get the human version, get the human version.