Interactive demonstration · about 2 minutes

Rehearse one decision.

Choose a system, change five assumptions and compare three moves across 120 test worlds. You will see what appears to hold, what fails and which condition matters most.

No accountStart immediately
No personal questionsNo personality test or profile
Inputs stay hereSimulation controls are not sent

A working illustrative simulation

Change the system. Watch the recommendation move.

This public model is deliberately simplified and rule-based. A commissioned study adds your evidence, stakeholder participation, alternative model structures and bounded LLM agents only when they improve the decision.

JavaScript is needed to run this local demonstration.

You can still read the method or discuss a real decision. No controls below are submitted.

The local demonstration could not load.

No simulation controls were sent. You can still read the method or discuss a real decision.

01 Choose a decision to rehearse

BusinessLaunch or expandDemand, adoption and operational exposure Public policyImprove policy uptakeTrust, access and unintended responses Social + ecologicalAct under scarcityAllocation, compliance and unequal impacts

02 Describe the conditions

These controls describe groups and context—not you. Try the defaults first, then move one assumption and run again.

Trust & legitimacy55 DistrustedTrusted Sensitivity to loss60 Loss tolerantLoss sensitive Peer influence55 IndependentSocially responsive Capacity to act50 ConstrainedWell resourced External volatility65 StableHighly volatile

03 Stress-test three possible moves

The same synthetic population and shock sequence are applied to every move.

Illustrative—not a forecast. More agents reduce simulation noise; they do not validate the assumptions. No personal profile is created, and simulation controls are not sent to Consilience.

Illustrative result · business launch

Run the model to compare the moves.

0/120test worlds met all criteria

Swipe or scroll to compare all five columns.

Comparison of all three strategies in this illustrative run
Strategy	Held	Central 80% outcome band	Median downside	Median impact gap

“Held” means the move crossed this demo’s declared outcome, downside and impact-gap thresholds in that test world. The outcome band is P10–P90; outcome, downside and gap are comparative 0–100 demo indices, not observed units or probabilities of real-world success.

See the demo rules and limits

Held rule: —
Band rule: —
Ranking: —
Comparison: —
Local sensitivity: —
Version and seed: — · —

MAIN SENSITIVITY

—

FAILURE HYPOTHESIS

—

EARLY SIGNAL

—

This is where a real engagement begins.We would replace illustrative assumptions with evidence, test competing model structures and agree what a decision-grade claim can carry.

Discuss this decision

From public demo to client work

What changes when the decision is real.

The simulation becomes useful evidence only after the assumptions, population, data and validation standard are specific to your decision.

01
Measure
Define constructs before collecting data. Use validated measures, observed behaviour, interviews and participation—with licensing, consent and cultural adaptation where needed.
02
Calibrate
Build group distributions, networks and constraints from evidence rather than treating synthetic personas as real people.
03
Challenge
Compare simple rules with bounded LLM agents; test sensitivity, subgroup error, alternative structures and historical or held-out patterns.
04
Decide
Carry forward robust moves, named conditions, early signals and explicit points to scale, pause or change course.

Psychometric boundary: the demo models hypotheses about group behaviour. It does not assess, diagnose or infer the personality of any visitor or real person.

Read the measurement and validation method →

Bring your own decision

Replace the demo assumptions with evidence from your world.

Tell us what you must decide, when it matters and what failure would cost. We will assess fit and propose the smallest credible study. Scope and price are agreed later, in writing.

Discuss a decisionNo checkout, order or payment on this website.

Rehearse one decision.

Change the system. Watch the recommendation move.

Run the model to compare the moves.

—

—

—

What changes when the decision is real.

Measure

Calibrate

Challenge

Decide

Replace the demo assumptions with evidence from your world.