Skip to content

The Long Game Project · Tools · TLDR

One piece of our wargame engine, the software that plays the wargames published on this site, is open for you to use today. It is free. It does not ask for an account.

From questions to a personaFlow diagram. 3 boxes: 54 questions; Fifteen scores; Your persona file. 54 questions leads to Fifteen scores. Fifteen scores leads to Your persona file.54 questionsFifteen scoresYour persona file
From questions to a personaFlow diagram. 3 boxes: 54 questions; Fifteen scores; Your persona file. 54 questions leads to Fifteen scores. Fifteen scores leads to Your persona file.54 questionsFifteen scoresYour persona file
Fifty-four answers become one persona file, a written profile of how your organisation decides.

The diagnostic

About eighteen minutes of questions. You get fifteen Ingredient scores (the traits we measure), a coherence rating (whether your answers about what you value and what you do agree), one of sixteen Flavours (the types we sort organisations into) and your blind spots. No AI scores it.

The persona file

The diagnostic also writes a persona file, about three pages on how your organisation decides. It is a system prompt, the standing instructions an AI model reads before it answers. Load it into any model that accepts one.

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

The Long Game Project · Tools

A tool from the practice

One piece of our wargame engine, the software that plays the wargames published on this site, is open for anyone to use. It is free for now. It does not ask for an account. This page says what it is and what it is not. It comes from Long Game Project, Dr Dan Epstein’s strategic-exercise practice, which runs this studio.

The strategic fingerprint, our name for the diagnostic

The fingerprint asks fifty-four questions about how your organisation makes decisions in practice. It takes about eighteen minutes, a long coffee. A fixed formula scores the answers, with no AI in the loop.

You get fifteen scores, which this site calls Ingredients. You get a coherence rating, which compares your answers about what you value with your answers about what you do. You get a match to one of sixteen archetypes, which we call Flavours. The most useful output is usually the blind-spot list: the places where those two sets of answers come apart.

It runs in your browser. Nothing leaves your browser unless you save the result to the public gallery. Saving is the one action on this page that asks you to sign in.

Take the diagnostic

How the scoring works, layer by layer, is in the diagnostic explainer.

The persona file

The diagnostic also writes a persona file: a written profile of how your organisation decides, about 800 to 1,200 words. It is a structured character brief. Hand it to a language model and the model plays the character the brief describes. Our wargame players use the same persona format. The diagnostic writes the player. The engine gives the player a game.

You download the file from your results page after the diagnostic. It is a system prompt, the standing instructions an AI model reads before it answers. Load it into any model that accepts one.

This site used to run a chat with your persona, called the digital twin. We retired the chat in September 2026. The persona file does the same job in the AI tool you already use.

Saved profiles live in the gallery.

The annex · Deep

What the tool keeps, what it sends, and what a persona file contains.

ToolRuns whereSendsKeeps
DiagnosticIn your browserNothing, unless you save to the gallery, which needs a sign-inYour answers in this browser session only

The persona file

  • Overview.
  • Strategic Ingredients.
  • Principles.
  • Win conditions.
  • Red lines.
  • Posture notes.

About 800 to 1,200 words of markdown, roughly three pages. It is a system prompt. Load it into any model that accepts one.

Whether a persona-driven answer matches what the real organisation would do is still an open question. The evidence page says what we have tested and what we have not.