Turnstats

Turnstats is a data collection framework for statistics gathered per turn in turn-based systems. It tracks how the state evolves from one turn to the next, enabling comparisons across players, strategies, or scenarios. It applies to digital and tabletop games, simulations, and research.

Data model: A typical entry records turn_number, timestamp, actor_id, actions_taken, resources_before, resources_after, score_change, and outcome_of_turn. Derived metrics include mean_turn_length, actions_per_turn, resource_change_per_turn, tempo, and turn-based win rate. Rolling windows (e.g., last 10 turns) support trend analysis.

Applications: Used in game design and balancing to measure pacing, identify effective or weak strategies, calibrate AI, and provide feedback. In simulations and AI research, turnstats support strategy evaluation and reinforcement learning on discrete timesteps.

Formats: Turnstats data can be stored in log files, databases, JSON, or CSV, with common visualizations like time-series plots and heatmaps.

Limitations: Discretizing time into turns can obscure rapid intra-turn events and bias metrics when turn length varies. Cross-domain standardization remains limited.

See also: turn-based systems, game analytics, per-turn analysis.