Your Composite Score is MindFrame's primary performance metric. It combines four components:
- **Accuracy (40%)**: The percentage of questions you answered correctly - **Calibration (30%)**: How well your confidence matched your accuracy (inverse Brier Score) - **Reasoning Quality (20%)**: AI evaluation of the quality of your explanations and reasoning - **Efficiency (10%)**: Speed relative to median completion time for the challenge
The weighting reflects the research literature: accuracy alone is a poor predictor of decision quality. A person who is 70% accurate but perfectly calibrated is more useful in real situations than someone who is 80% accurate but wildly overconfident — because the first person knows what they don't know.
Scores are percentile-ranked against all MindFrame users so you can see not just your absolute performance but your relative standing.