SPE Concepts and Definitions

Use these definitions consistently when building SPE models in Albarena.

Variable role map

Driver: direct strategic entry point

Controllable by strategy · Entry or context

Strategy scenario impact: change applied to a Driver

Controllable by strategy · Mediates or measures

External Agent: external context for scenarios

Not directly controllable · Entry or context

Process or Outcome: mechanism or final metric

Not directly controllable · Mediates or measures

Use the role test before adding a variable to the model.

Driver

A Driver is an internal, high-leverage variable that is directly affected by strategic choices. It is the primary entry point through which strategies propagate effects into the model.

Drivers are directly or partially controllable by the decision-maker. A strategy scenario should impact Drivers directly. Drivers may influence Processes and, when justified, other Drivers.

Process

A Process is an operational or organizational mechanism through which Drivers affect performance. It represents how work is executed, coordinated, learned, controlled, or transformed into results.

Processes are not directly modified by strategy scenarios. They receive effects from Drivers and propagate effects to Outcomes or other Processes.

Outcome

An Outcome is a measurable performance result used to evaluate and compare strategic alternatives. Outcomes terminate the causal chain and should not propagate forward.

Examples include cost, schedule, profitability, safety, quality, stakeholder satisfaction, resilience, or strategic value. Outcomes should have a clear interpretation and, when possible, a scale.

External Agent

An External Agent is an exogenous variable that represents a force, actor, condition, or environment outside the control of the decision-maker. External Agents define context and are used to construct scenarios.

External Agents usually start at Level 0, before Level 1 Drivers.

Strategies should not directly modify External Agents. External Agents may influence Drivers, Processes, or Outcomes depending on the model structure.

Cross-Impact Matrix

The Cross-Impact Matrix records how a source variable influences a target variable. Each relationship is assigned a pattern such as Sig+, Mod+, Sli+, No, Sli-, Mod-, or Sig-.

The matrix is the bridge between the conceptual model and the simulation model. If the matrix is weak, arbitrary, or contradictory, the simulation results will be weak even if the interface is complete.

Initial probability

Each SPE variable begins with an initial probability vector over five states:

NN: very negative
N: negative
O: neutral
P: positive
PP: very positive

The probabilities must sum to 1.0. The default Albarena prior is uniform: 0.2 for each state.

Initial probability over five states

The default prior starts each variable equally likely across NN, N, O, P, and PP.

Patterns

Patterns describe the direction and strength of influence between variables. Positive patterns move the target distribution in the same direction as the source. Negative patterns move it in the opposite direction. The No pattern is a zero-effect relationship.

Seven default impact strengths

Sig-

Mod-

Sli-

Sli+

Mod+

Sig+

The sign gives direction and the magnitude gives strength. The No pattern has zero effect.