Value of Flexibility example

This synthetic example compares a policy that can choose a feasible strategy for each timing scenario with the best strategy committed before the timing scenario is known. It is an adjacent option-value analysis, not EVPI or EVSI.

Run the normative input from the repository root:

voiage --format json calculate-value-of-flexibility \
  specs/frontier/value-of-flexibility/v1/fixtures/normative/input.json

The example uses scenario probabilities 0.2, 0.3 and 0.5. Strategy A has value 8 in every scenario; strategy B has values 0, 10 and 12. The flexible value is 10.6, the best fixed commitment is strategy B with value 9.0, and Value of Flexibility is therefore 1.6 net-benefit points.

The result explicitly reports both policy paths, their named scenario and strategy axes, the commitment baseline, provenance, option value, units and exact-enumeration assurance. Because the scenarios are mutually exclusive, exercise_decisions is unsupported (null); changes between adjacent ordered scenario choices are reported separately as ordered_scenario_policy_changes. Its information_value_component is zero and decomposition_status is information-value-excluded; do not add EVPI or EVSI that is already embedded in the payoff surface.

Version 1 accepts mutually exclusive timing scenarios only. It rejects lifecycle-period aggregation, missing or extra named probabilities, non-increasing event times, infeasible commitment strategies, empty units and embedded information value. Non-zero discount, irreversibility and lock-in controls fail closed until their units and policy-dependent semantics have a separately governed contract. Rust, R and Julia execution are currently unsupported, and Mojo remains external.