Skip to content

The Long Game Project · About · TLDR

A studio that runs wargames with AI players on live strategic questions, then publishes the run with the working shown. One person signs every report.

Who

Dr Dan Epstein. A medical doctor with a PhD, and director of the Long Game Project, where he has designed more than 130 business and strategy wargames. He commissions, edits and reviews every report here.

What for

The questions we would go deeper on if we had the time. We publish them to start a better argument. We do not publish them to call the outcome.

Your question

A commissioned report uses your actors and your evidence and is never published. A tabletop puts your team in the seats. Prices are on the commission page.

Switch to Normal for the full page, or Deep for the working.

The Long Game Project · About

About Strategy Soup

A studio that runs wargames with AI players on live strategic questions, then publishes the run with the working shown. Here is who runs it, what it is for, and why it is called soup.

Who is behind it

Strategy Soup is a studio of the Long Game Project, Dr Dan Epstein’s strategic-exercise practice. Dan started designing tabletop games at eleven and never stopped. Along the way he became a medical doctor, earned a PhD in decision science from Monash University, and spent years consulting across public and private sectors, from pandemic preparedness to behavioural economics.

He has designed more than 130 business and strategy wargames across defence, tech, financial services and consumer goods, and he leads every engagement himself, from scoping through facilitation to debrief. He built the Long Game Project because he kept seeing the same problem: capable teams making poor strategic decisions, not because they lacked information, but because they had never practised thinking under pressure.

The machine plays. Dr Dan Epstein is the author and reviewer of every report here, and where Claude drafted text from a run record, the report’s masthead says so. The full bio is on the Long Game Project site.

What this site is for

These are the wargames and research questions we would go deeper on if we had the time. We publish them to start a better argument, and sharper questions, while a situation is still moving. We do not publish them to call the outcome.

The engine behind the reports grew out of the method the Long Game Project uses in the room with clients. In a matrix game, players argue for each move and a referee rules on it. Under that sits the unfashionable belief that teams get better at decisions by rehearsing them.

Nothing on the site claims the AI players predict what a real organisation would do. Only the evidence page may answer that question, and today its answer is “open”.

Some questions deserve the full treatment: your actors, your evidence, your people in the seats. That is a commissioned report, or a tabletop. The commission page says what each one is and what it costs.

Why “soup”

Fair question. The internet is full of AI slop, and this is a site of AI-played games. Slop is a model free-associating into the void. No question, no rival, no referee, no record, and no one accountable for a single sentence. Soup starts with named ingredients, follows a recipe, and has someone tasting the whole time. The recipe is public, the ingredients are cited, and when the cook overrules the kitchen we write it down.

So: soup. Made from stock we can show you.

The annex · Deep

The rules every report follows are stated once, on Methods, under “The rules we do not bend”. Two rules for the site as a whole.

  • Nothing on the site claims the AI players predict what a real organisation would do. Only the evidence page may answer that question, and today its answer is “open”.
  • No em dashes. (A house rule. It matters to the author.)