Never Tell Them the Answer

A steering committee, three weeks ago, in a sixty-person plastics shop that supplies the automotive trade. The question on the table was whether to keep the ageing production application for another two years or replace it now. I had an opinion. I had it before the meeting started, and I was fairly sure it was right.

I said nothing for forty minutes.

They got there. Not to my answer, to a better one, because the production manager knew something about the August shutdown that I did not.

On the drive back I worked out that I had been doing that particular trick for forty-three years, and that nobody had ever taught it to me at work. I learned it at fourteen, behind a cardboard screen, in the youth club of a village in the Yvelines.

Three books in a language I did not have yet

1983. The three Advanced Dungeons & Dragons hardbacks had been out for a few years by then: the Monster Manual since December 1977, the Players Handbook since June 1978, the Dungeon Masters Guide since August 1979. The French translations were still years off, and when they eventually arrived they were poor enough that a generation of us skipped them. So I bought the American editions, along with the figurines and the strange dice, at the Descartes shop on the rue des Écoles, opposite the Sorbonne. Casus Belli, which had launched in April 1980 and was the only French magazine that took any of this seriously, came home in the same bag.

Then I went back to the Yvelines and read two hundred pages of English I could not read.

I was in collège. My English was the English you have at fourteen after a couple of years of it, which is to say almost none. I learned the rest from a rulebook about hit dice and saving throws, quickly, because for the first time there was something on the far side of the language that I actually wanted. School had given me English as a subject. TSR gave me English as a door.

Every manager who has ever wondered why a team will not adopt a tool should sit with that difference for a minute. I have never once seen a reason arrive because someone was given a manual.

The screen hides the dice, not the monsters

The Dungeon Master sits behind a screen. Three panels of cardboard, tables printed on the inside, the players seeing only a picture of a dragon. They assume it is there to hide the monsters. It is there to hide the dice.

I fudged rolls. Not often, and never to spare anyone the consequences of a choice they had actually made. But a wandering monster that turns up in the first ten minutes and kills a character somebody spent a month building is not a consequence of anything. It is arithmetic arriving before the story does. And a fight going so lopsidedly the party’s way that everyone has visibly stopped caring is not a triumph, it is a room emptying. Both times, I moved a number. Both times, I would do it again.

The distinction I could not have articulated at fourteen, and would defend now, is that I was never changing the outcome. I was protecting the conditions under which an outcome could mean something. Wanting the table to like me would not have qualified, and I knew that even then, which is exactly why I can still count the occasions.

Discretion is only tolerable when you could explain each use of it out loud, to the people it was used on, and have them agree. That is not a principle I picked up in management. It is one I worked out about myself at a folding table, and I have since watched a good number of executives fail it without noticing.

The scenario was never the thing

I would spend a week drawing a dungeon. Corridors, a flooded level, a faction of goblins with a grievance I had thought about more than was reasonable. Then five teenagers would arrive on Saturday, take the first left instead of the first right, befriend the goblins, and spend the afternoon somewhere that existed only as a blank square on my grid.

You have two options at that moment, and only one of them is the job. You can push them back toward the material you prepared, which they will feel and resent. Or you can build the room in front of them, from whatever you have, and let them believe it had been there all along.

I have written elsewhere about the adjacent possible, the idea that what is reachable next is determined by where you currently stand and not by where you meant to go. A dungeon is the crudest possible model of it. Every door is an adjacency. You do not get to decide which one they open, only what is honestly on the other side.

Most of what I now do for a living is that, with a budget attached.

Never tell them the answer

The hardest minutes of my adolescence were not the fights. They were the puzzles.

You write a riddle into a door. The door is on the critical path. Forty minutes later five teenagers are going in circles, the energy in the room is leaking out under it, and the answer is sitting in your notes, ten centimetres from your hand.

You cannot say it. If you say it, you have not helped them. You have finished their evening for them and called it generosity.

So you learn the smallest available intervention, and then you learn a smaller one. You re-describe the room, letting one adjective carry more weight than it did the first time. You have a torch gutter in a draught nobody thought to ask about. You let a passing character mention something that is not the answer but sits next to it. You ask a question rather than make a statement: what exactly did the inscription say?

And then you stop talking, which is the part that takes years.

When I later spent nineteen years running a company on sociocratic circles, the vocabulary was new to me and the gesture was not. Holding a decision open while a group finds its own way to it, resisting the pull to supply an answer you are already carrying, is the door with the riddle in it. The stakes changed. The move did not.

What surprises people is that this is harder than deciding, not easier. Deciding takes a minute. Holding takes the whole meeting, and you get no credit for it, because when it works the group is quite certain they did it themselves.

They did. That is the point. It is also the invoice.

The rules are the boring part

The Dungeon Masters Guide is two hundred and thirty-two pages, most of them tables. Encumbrance. Random wilderness terrain. The exact behaviour of a magic item nobody would ever draw. It is the most gloriously over-specified document I owned before I met enterprise software.

And then, on page 230, near the end of the book he had spent years writing, Gygax says that it is the spirit of the game and not the letter of the rules that matters, and that you should never allow some barracks room lawyer to force quotations from the rule book upon you.

The man who wrote the rules put in print that the rules were not the job.

I think about that page every time someone hands me a governance charter, a security framework, a set of ITIL processes, or a national cyber baseline with numbered security objectives. The document is real and I read it properly. But I have watched organisations treat the existence of the document as the completion of the work, and I have never seen that end well. The same instinct is why, when a client freezes on a regulation, I try to read the text back to them as a short list of operational questions rather than as a wall.

Rules are cheap. Rulings are the job.

The panic that never reached the Yvelines

The whole time this was going on, the press and the television were explaining that the game was dangerous. Impressionable adolescents, occult symbols, boys who supposedly went missing. It ran for years and it was not a fringe story.

I never met it. Not once. Not from my parents, not from my friends’ parents, not from a teacher, and not from the youth club that hosted five teenagers every Saturday for years and never asked us a single question about what we were doing in there.

I have since asked around. Nobody I know lived it either.

So there were two objects in the world at the same time. There was a media object called Dungeons & Dragons, which was a menace to the young. And there was an actual game, played on actual Saturday afternoons, which was five kids doing arithmetic and arguing about a door. The two never touched at any point, and the first one never once made contact with my life.

That is the oldest lesson in this piece and the one I use most often. The coverage of a thing and the thing are different objects, and the gap between them does not close on its own. Somebody has to go and look.

I get paid for this now. A dirigeant asks whether AI is going to replace a team, or whether a given vendor is a risk, or whether everyone is moving to the cloud. The question always arrives shaped by the coverage and almost never by the work. Half of my first meeting is usually spent establishing which of the two objects we are talking about.

The job description has not changed

The regular table ended with the lycée. I ran the occasional game into my late twenties, more rarely each year, and then the twenties ended and so did that.

What I did behind the screen, for those four years, was hold a world’s constraints and refuse to hold its outcome. I knew what was true. I knew what waited if they went left. I did not decide for them, I did not rescue them from a choice they had actually made, and when they walked into a room I had never drawn, I built it while they watched and kept the tone steady.

That is the whole of what I do now, at a different table, for a rate. The conductor in an earlier piece of mine has one advantage I never had, which is a score. The Dungeon Master does not get one. Neither does anyone steering a company through a decision that has not been made yet.

Nobody issued a certificate for it. It remains the best training I ever had, and I paid for it in pocket money at a shop on the rue des Écoles.

It is the spirit of the game, not the letter of the rules, which is important.

Gary Gygax, Dungeon Masters Guide, Afterword, page 230, 1979.