Known and hidden state

State is the factual content of a scenario: everything that is true about the situation, split by who can know it. It comes in two kinds, and the distinction is the backbone of scenario design.

• Known state is information the persona knows about themselves. They reveal it selectively, depending on what the operator asks and how much trust the operator builds.
• Hidden state is information the persona genuinely cannot know. The operator discovers it through their own professional process.

State does the most work in conversation scenarios and journey simulations, where the operator has to draw it out. In decision tasks, the equivalent facts arrive up front, usually as documents.

​

The litmus test

For any fact you want to add, ask one question:

Could the persona say “yes, I knew that”?

• Yes - it is known state. The persona has it; the question is when they share it.
• No - it is hidden state. The persona cannot have it: an exam finding, a test result, a record they have never seen.
• It is a textbook fact the operator should know from their training - it is not state at all. Domain knowledge belongs to the operator, not the scenario.

A worker knows they have been doing side jobs, so that is known state. They have not seen their own imaging results, so those are hidden state. The clinical meaning of those results is domain knowledge, so it does not go in the scenario.

​

Known state

Each known state item carries a reveal trigger that controls when the persona shares it: This is where skill shows. A strong operator asks the right direct questions and builds enough trust to surface what the persona is holding back. A weak operator leaves rapport-gated facts on the table and never knows they were there.

​

Revelation mechanics

Reveal triggers say when a fact can come out. Revelation mechanics set the overall difficulty of getting the persona to open up. There are five modes, forming a difficulty ladder:

​

Hidden state

Hidden state is everything the persona cannot know. The operator reaches it by doing their job: ordering the test, pulling the record, running the analysis. Examples:

• Investigation results - lab values, imaging findings, assessment scores
• Records - a prior claims history the persona has no access to
• Red flags - inconsistencies only visible through professional analysis

​

State over time in journeys

In journey simulations, state items can carry a recording frequency that controls how often the fact changes as simulated time advances: Pain levels might move daily, work status weekly, and a formal review monthly. Between sessions, the state evolves in response to what the operator did, so the situation the operator returns to reflects their earlier choices.

​

Next steps