There is now 50 years of research documenting how cognitive biases distort human judgement and lead to worse decisions. Behavioural science needs a simple, outcome-focused benchmark that can be used to measure the impact of interventions to debias and improve decision-making. This paper proposes one: calibration. To be well-calibrated means your confidence in your judgement aligns with accuracy. 

The goal of calibration isn’t to eliminate cognitive bias, but to enable good decisions despite it. This report offers:

• a clear framework for understanding and measuring calibration
• practical tips for individuals, teams and leaders to achieve it
• design principles for building calibration into organisational systems.

Key insights

• Overconfidence is rampant in many domains from surgery to strategy – people are more sure in their judgement than they should be and this leads to bad outcomes.
• Well-calibrated thinkers are rare but superforecasters, bridge players and weather forecasters show that it’s possible, and give us hints about what types of environments foster and reward it.
• Calibration is trainable. Like physical fitness, it can be built through habit, feedback and repetition.