Project Decision Framing Tool

This tool helps you structure a recurring project decision using Warren Powell’s five-element framework: state, decision, new information, transition and objective. Read the source framework. This questionnaire organises your assumptions; it does not validate a model.

1. Identify the Decision Context

What is the recurring decision you need to make? How often do you typically make this decision (Time Step 't')?

2. Identify State (S_t): Information Needed AT the Decision Point

Physical State (R_t): What describes the tangible status of the project?

Describe the resources and physical position known at decision time; estimates and unknown quantities belong in the information or belief description.

Informational State (I_t): What other known factors influence the decision?

This is known information, not physical resources, that affects your choices or outcomes.

Belief State (B_t): What key factors are uncertain, requiring estimates or beliefs?

This captures your knowledge about things you *don't* know perfectly.

Which elements (from above) are most uncertain or lack sufficient information?

3. Identify Decision Variables (x_t): Choices Made

What specific actions/choices do you make? What constraints limit these actions?

4. Identify Exogenous Information (W_t+1): New Info Arriving AFTER Decision

What new, relevant information typically arrives *after* you make the decision but *before* the next one?

These observations and shocks are unknown when the current decision is chosen. Their distribution may depend on the state or action—for example, more monitoring can reduce measurement noise.

How does this new information (W_t+1) help reduce the key uncertainties identified earlier?

5. Describe the Transition (S^M)

How do the current state, chosen action and newly arriving information produce the next state?

Write a rule or plain-language update: S_(t+1) = S^M(S_t, x_t, W_(t+1)). Distinguish an action’s intended effect from its observed outcome.

6. Identify Objective / Metrics

How do you measure the success of this decision over time?

7. Cadence Check

Does critical new information (W_t+1) often arrive *much faster* than your decision frequency? Consider the uncertainties you want to reduce (from Step 2). How feasible (time/cost) is it to get *better* information *before* your next decision point?

Aligning decision frequency with the arrival and value of new information is key.

Project Decision Framing Tool

1. Identify the Decision Context

2. Identify State (S_t): Information Needed AT the Decision Point

3. Identify Decision Variables (x_t): Choices Made

4. Identify Exogenous Information (W_t+1): New Info Arriving AFTER Decision

5. Describe the Transition (S^M)

6. Identify Objective / Metrics

7. Cadence Check

Sequential Decision Framework Summary

State (S_t) - Information at Decision Time:

Decision (x_t) - Actions Taken:

Exogenous Information (W_t+1) - New Info Arriving After Decision:

Transition to the Next State:

Objective:

Cadence Considerations:

Next Steps & Focus:

Project Decision Framing Tool

1. Identify the Decision Context

2. Identify State (St): Information Needed AT the Decision Point

3. Identify Decision Variables (xt): Choices Made

4. Identify Exogenous Information (Wt+1): New Info Arriving AFTER Decision

5. Describe the Transition (SM)

6. Identify Objective / Metrics

7. Cadence Check

Sequential Decision Framework Summary

State (St) - Information at Decision Time:

Decision (xt) - Actions Taken:

Exogenous Information (Wt+1) - New Info Arriving After Decision:

Transition to the Next State:

Objective:

Cadence Considerations:

Next Steps & Focus:

2. Identify State (S_t): Information Needed AT the Decision Point

3. Identify Decision Variables (x_t): Choices Made

4. Identify Exogenous Information (W_t+1): New Info Arriving AFTER Decision

5. Describe the Transition (S^M)

State (S_t) - Information at Decision Time:

Decision (x_t) - Actions Taken:

Exogenous Information (W_t+1) - New Info Arriving After Decision: