A small, chaotic gnome with a fondness for party hats. A half-shark barbarian of limited patience. A wizard with a private agenda. The game is run by a Dungeon Master - part world-builder, part storyteller, part referee - who is responsible for creating the environment, voicing every character the players encounter, and keeping the whole thing from descending into chaos. Or at least they try.
I've been playing D&D as a player and as a Dungeon Master for over ten years. I also build tools and training for people who want to thrive in a culturally diverse, global landscape. Here's what the first thing taught me about the second.
D&D is fundamentally a cultural simulation. It forces you to inhabit perspectives that aren't yours, navigate environments with their own norms and power structures, and work within a team of people whose instincts and values regularly conflict with each other. The rule "stay in character" isn't just a social contract for the table - it's a constraint that makes the learning real. You can't just describe what your character would theoretically do. You have to do it, in the moment, with other people watching and reacting. And that requires what we at Country Navigator call “cultural intelligence” - not just in theory, but in action.
Before a single dice is rolled in D&D, you create your character. And that process is very much an analogue for culture.
You start with a race - the traits and tendencies you were born into. An elf and a dwarf don't just have different ears; they've absorbed different values, instincts, and assumptions about the world before making any conscious choices. Origin shapes outlook.
Then you have a background - what happened to you. A character who grew up as a criminal navigates authority differently than one raised in a noble court. Not because of who they fundamentally are, but because of what their experience has taught them to expect.
Finally, you choose a class - what you decided to become. The discipline you committed to, the role you chose to play. Your job.
Your cultural profile works the same way. Where you came from, what you've lived through, and the choices you've made about who you want to be all factor in. Most cross-cultural misunderstanding happens because we're reading one of those layers and treating it as the complete picture.
Every D&D character has a set of ability scores - strength, intelligence, wisdom, charisma, constitution. No character is exceptional across all of them. What matters isn't any single score; it's the profile, and what it implies about how that character functions in different situations.
Our Worldprism™ model of culture works on the same logic. It doesn't rank you - it maps where your cultural tendencies sit across nine dimensions. High on directness. Low on risk tolerance. Group-oriented. Linear in your approach. No position is inherently better than another, but each comes with its own strengths and blind spots.
Which brings us to one of the most important (and most amusingly ignored!) rules in D&D: never split the party.
It's a running joke in the community because it happens constantly - and almost always ends badly. The moment the group divides everyone becomes more vulnerable - the mission unravels, and someone ends up trapped in a dungeon alone wondering where the cleric went. You need the fighter, the wizard, the rogue and the healer operating together, because what one can't do - another can. A party of all fighters doesn't usually get very far.
This is what we're working on when we run cultural intelligence programmes for diverse teams. Not trying to smooth out difference, but helping a team understand its own profile: who covers what, where the gaps are and how to use the range of perspectives available rather than work around them.
Worldprism™ maps cultural tendencies across nine dimensions. Here’s how I’ve seen three of them play out in some of my favourite D&D campaigns.
Lilli was a gnome arcane trickster. Tiny, cheerful, wearing a party hat at all times… and completely and utterly unable to leave a shiny object alone. In Curse of Strahd - a gothic horror full of cursed artefacts and malevolent magic - this was (unsurprisingly) a recurring problem. By the time we reached the end of the campaign, Lilli had so many curses the rest of the party didn’t know which was which.
Helena was her best friend. A paladin - leader, protector, healer - who never rushed into anything. Where Lilli moved toward uncertainty with delight, Helena moved toward it with a plan. She was the one who assessed the room, watched the exits, and calculated what it would take to get everyone out. When Lilli acquired curse number three, Helena didn't abandon her, lecture her or lose patience. She just adapted, again and again, and figured out how to help.
As a pair, they represent the extremes of the Risk Taking / Risk Avoiding dimension. In a workplace, you can imagine the friction - the one who wants to pilot everything immediately, and the one who wants the risk assessment first. But what made Lilli and Helena extraordinary as a duo was that they learned to use their differences. Lilli found things out that no cautious character ever would. Helena was there to manage the consequences. The party was better for having both of them, but only because they'd stopped trying to be the same.
Mano is an excellent character to have in your party when the objective is physical. He is loyal, effective in combat, and deeply committed to forward momentum. He is also always hungry. Ravenously, ferociously hungry. He operates in a mental state that sits somewhere between "ready for a fight" and "fighting".
The difficulty with Mano emerges the moment you need to talk to someone.
In one session, the party encountered a strange, aggressive knight - visibly unsettled, possibly cursed, and exactly the kind of person you'd want to approach carefully. We looked at him and saw a lead. Mano looked at him and saw lunch. What followed was a split-second decision. Some of us began talking, trying to engage the knight before things deteriorated. Our French warlock Linus, however, made a different call. Rather than attempting to negotiate with Mano (which, given his current “hanger” … was probably not going to work), he cast the spell “Charm Person”. Mano was suddenly at his mercy - and so was the knight. Sometimes we think being culturally intelligent means adapting to others' preferences. But knowing when to back yourself and use your own strengths? That’s another form of flexing your cultural muscle.
The Worldprism™ dimension here is Task versus Relationship. Mano sits at the far task-oriented end: there is an objective, and everything else is an obstacle. For the party to function in a situation that required relationship investment, everyone else had to be creative and “flex” - not to change Mano, but to compensate for the gap. In any diverse team, some version of this dynamic exists. The insight isn't that task-oriented people are a problem. It's that a strategy built only around their instincts will damage relationships that matter - and the team needs to know when to create some space.
Halmo the Wizard had, it turned out, been doing some private research.
For weeks of campaign time, he had been quietly acquiring the knowledge and components required to summon a demon - through means the rest of the party would absolutely not have sanctioned, had they known. His target was Count Strahd himself: the ancient vampire lord at the heart of the campaign, the kind of Big Bad Evil Guy that requires creative solutions. A demon, if it could be kept under control, would be a significant weapon. If it couldn't… well. It would turn on the party. And we would all die.
The party found out about Halmo’s plan. We confronted him, angrily. Halmo heard us out. And then… he summoned the demon anyway.
During battle, Halmo’s control held. The demon fought for us like he was one of our own. The summoning proved immeasurably useful against the villain – and in no small part thanks to Halmo’s audacity, the party survived. Halmo's judgment, by any measurable outcome, had turned out to be correct.
But the conversation that happened afterwards - about trust, about what it means to make a unilateral decision on behalf of a group that expected to be consulted - was harder. The Individual / Group dimension in Worldprism™ isn't about selfishness versus teamwork. It's about how much weight you give your own judgment relative to the collective. Halmo wasn't wrong, and the outcome was good. But the trust he sacrificed to get there didn't recover quickly. In an organisation, this plays out whenever someone makes a unilateral call on something the team had a stake in. The result might be fine. But the relationship damage is unavoidable - and the team does not always recover easily.
When you play D&D, you have to inhabit someone who is not you. Their risk appetite is different. Their instincts in a conflict are different. Their relationship to authority and to the group is different. The "stay in character" rule exists precisely to stop you defaulting to your own instincts when your character's would differ. Over time, this trains something important: the ability to genuinely operate from a different cultural position, rather than just intellectually understanding that one exists.
But the learning doesn't stop there. Your character still has to navigate the world - and that's where the CultureFlex ADAPT process plays out. You analyse the gap between how your character would instinctively respond and what the situation actually needs. You decide how far to flex. You apply it under social pressure, with other players reacting in real time. You process the feedback - did that land, did the room shift? And you tune on the next interaction based on what you learned.
The reason this builds genuine skill rather than theoretical awareness is that you actually did the thing. You didn't reflect on what you might do in a cross-cultural situation. You made a choice and experienced the consequences. That's a learning loop that formal training invests a lot of effort trying to replicate.
Being the Dungeon Master is a different kind of experience.
As a DM, you don't just navigate cultural complexity - you construct it. Every faction, city, and society in your world has implicit norms: a relationship with time and hierarchy, a communication style, an expectation about how decisions get made. You make those choices whether you’re conscious of them or not, and players notice almost immediately when a world lacks cultural coherence.
You also voice every character who isn't a player - which means code-switching constantly, in real time, without preparation. A formal council. A suspicious trader. An NPC who doesn't trust outsiders and has earned that position. Each carries a different register, and you have to inhabit them convincingly enough that the players believe the world is real.
And when the party starts to fracture, you make judgment calls without a script. Do you adapt the situation to where the group is? Do you create a moment that pulls competing instincts together? Do you let the tension play out because the story needs it? These are leadership decisions, made in real time, about a team with fundamentally different profiles. It's the closest analogue I've found to the experience of leading a genuinely diverse team through a real, high-stakes moment.
None of this is purely theoretical. My friend Rob Cox has been running Dungeons & Dragons sessions for corporate teams for years through his company The Games Initiative. Rob designs each session around the skills a team actually wants to work on - communication, negotiation, decision-making under pressure, leadership - and follows it with a structured debrief that connects what happened at the table to what happens in the office. The Game Master handles everything - participants just have to show up and make decisions together.
What Rob does with D&D, and what we do at Country Navigator with our cultural simulations and coaching, is built on the same foundation: people learn by doing, not by being told. The research on simulation-based learning is consistent - when you practise a skill in a context where the stakes feel real but the consequences are safe, the learning transfers in a way that instruction alone rarely achieves. Whether it's a fantasy campaign or an AI-powered coaching scenario, what you're creating is a space where people can try things, get feedback, and adjust - without the cost of getting it wrong in a real meeting, a real negotiation, or a real team conflict.
The deeper reason D&D works as a learning environment - and the reason Rob's sessions are so successful - comes down to psychological safety.
You're not admitting that you struggle with risk. Your gnome does. You're not working on your tendency to override the group's judgment. Your wizard is. The distance between you and the character is what makes it feel safe enough to actually try something different, rather than defaulting to familiar patterns.
This is the same design logic behind our AI simulations and Carla. You can rehearse a difficult cross-cultural conversation - a negotiation, a feedback moment, a situation where your instincts are likely wrong for the context - in a space where getting it wrong has no cost. Try the version you'd never risk in a live meeting. See how the feedback lands. Try again. The simulation creates exactly the condition that makes the table so effective: you can actually practise, rather than just think about practising.
The fiction isn't there to make the training more palatable. It's the mechanism that makes it work.
So, the next time your team is struggling to function across its differences - when the range of instincts and approaches feels like friction rather than resource - it's worth asking what role everyone is playing, and whether the party has everything it needs to stay together.
And if your barbarian is eyeing the diplomat, step in. Fast.
Country Navigator is a cultural intelligence platform helping organizations build high-performing global teams. Learn more at countrynavigator.com.