Science Analysis Template
These are the structural patterns found across all Scientific Disciplines
In the Interaction domain, parameterization specifies how multi-agent economic situations are encoded into a formal state description. Unlike Choice, where the state concerns a single decision-maker, Interaction requires representing joint action spaces, incentive structures, and information conditions across multiple agents. Parameterization here defines the strategic environment itself: who can act, how actions combine, what outcomes result, and what each agent knows. The sections below map the standard scientific parameterization patterns onto strategic economic systems.
1. Minimal State Variables and Phase Space
In Interaction, the minimal state variables must jointly specify a strategic situation involving multiple agents. The interaction state is defined by:
- Each agent’s action or strategy space
- Payoff or outcome functions linking joint actions to results
- Information available to each agent
Together, these define an interaction phase space, where each point corresponds to a fully specified strategic configuration. A change in any agent’s available actions, payoffs, or information constitutes a change in state.
2. Field Variables and Functional Representations
Interaction is parameterized by payoff fields defined over joint action spaces. These fields assign outcomes to every possible combination of agents’ actions.
Concretely:
- Payoff functions over strategy profiles
- Cost or utility mappings from joint actions to agent-specific outcomes
The interaction state is the entire payoff structure over the joint space, not any single realized interaction.
3. Statistical Distributions and Ensemble Descriptions
When behavior is uncertain, mixed, or heterogeneous, interaction states are parameterized statistically.
This includes:
- Mixed strategies (probability distributions over actions)
- Belief distributions about other agents’ actions or types
- Population-level interaction frequencies
The system state is encoded as a distribution over strategic profiles, with observed interactions treated as realizations from that distribution.
4. Dimensionless Numbers and Regime Parameters
Qualitative interaction regimes are governed by relative payoff and information ratios, not absolute levels.
Examples include:
- Relative payoff dominance
- Incentive compatibility ratios
- Coordination versus conflict thresholds
- Signal-to-noise ratios in information transmission
These parameters determine whether interaction regimes take forms such as coordination, competition, bargaining, or dominance.
5. Multiscale Parameterization of Unresolved Processes
Many determinants of interaction—social norms, reputation, bounded rationality, learning dynamics—are not modeled explicitly.
Their effects are compressed into:
- Reduced-form payoff adjustments
- Strategy constraints
- Type parameters
- Error or noise terms
This multiscale parameterization collapses complex social processes into effective strategic parameters.
6. External Conditions and Boundary Parameters
Interaction states include external structural conditions that agents do not control.
These include:
- Market rules and institutions
- Communication protocols
- Legal enforcement mechanisms
- Network or matching constraints
These parameters define the boundaries within which strategic interaction occurs.
7. Formal and Mathematical Encodings
Formally, interaction states are encoded using:
- Strategy vectors and profile spaces
- Payoff tensors or matrices
- Belief measures and information partitions
- Constraint sets defining admissible strategies
Only after this encoding is fixed are equilibrium or solution concepts applied.
8. Completeness and Parsimony
Interaction parameterization balances:
- Inclusion of all variables affecting strategic incentives and information
- Exclusion of redundant or behaviorally irrelevant detail
The goal is a minimal strategic state description that uniquely determines predicted interaction patterns.
In economic interaction, outcomes depend on how strategic environments are parameterized before any equilibrium or solution concept is applied. By encoding joint action spaces, payoff structures, beliefs, uncertainty, and external institutional constraints into a coherent state description, interaction models transform social situations into analyzable systems. Differences in observed strategic behavior follow from differences in the parameterized interaction state, not from the equilibrium concept alone. As with other scientific domains, the explanatory power of interaction theory rests first on disciplined and explicit parameterization.