BestResponseResult

Source:

src/ometeotl_core/game/best_response.py

Local role: Result container for one best-response computation: the dominant strategy for a focal actor given a fixed opponent profile, plus a ranked list of all available responses.

Big-picture role: Rational behavior in a competitive context is not “pick the best strategy in isolation” — it is “pick the best strategy given what others are doing.” BestResponseResult is the answer to that conditional question: what is the optimal choice for this actor, and how do all other options compare, when opponents are assumed to play a specific way?

Inheritance:

standard dataclass

Fields:

actor_id: str — the focal player
opponent_profile: StrategyProfile — the fixed opponent strategies used in this computation
best_strategy: Strategy — the utility-maximising strategy
best_utility: UtilityFrame — the utility frame for the best strategy
all_responses: list[tuple[Strategy, UtilityFrame]] — all focal strategies ranked descending by utility, ties broken ascending by strategy id

Methods:

to_dict() -> JsonMap

Example:

calc = BestResponseCalculator()
result = calc.compute(
    actor_id="actor-1",
    opponent_profile={"actor-2": actor2_strategy},
    game=game,
)

print(result.actor_id)
print(result.best_strategy.id)
print(result.best_utility.scalar_value)

# Full ranking of all available responses, descending by utility
for strategy, frame in result.all_responses:
    print(strategy.id, frame.scalar_value)

data = result.to_dict()

Ometeotl