Schedule a consistent 30-minute block to skim recent entries, score outcome quality versus decision quality, and capture one improvement experiment. Tag anomalies and surprises. Ask what you would repeat, refine, or retire. This ritual keeps feedback loops short, prevents drift, and ensures that even small wins are noticed and reinforced before they fade.
Aggregate your confidence ratings and outcomes to compute calibration and resolution metrics. Celebrate accurate uncertainty, not lucky hits. Plot Brier or log scores and identify domains where your sense is sharp versus foggy. Update base rates and refine your default ranges. Invite a trusted peer to poke holes in recurring stories and confirm real progress.
Choose three pivotal decisions and reconstruct the reasoning with brutal honesty. What evidence was missing, and which incentives whispered in your ear? Simulate the counterfactual: what would a well-informed opponent argue? Document durable principles and turn them into brief checklists. Share highlights with collaborators to institutionalize learning, not just personal epiphanies.
Write as if the decision already failed spectacularly. List the plausible causes and early warning signs. Then backcast from success to identify necessary milestones. This dual view reveals fragile links and encourages practical safeguards. Capture two prevention steps and one detection step you will actually implement before enthusiasm erases uncomfortable possibilities.
Before deciding, require at least one opposing source, counter-dataset, or expert critique. Ask what would change your mind quickly and what evidence would make you walk away. Track whether you truly updated or rationalized. This habit tames motivated reasoning, increases intellectual humility, and improves the signal-to-noise ratio of your notes.