Capturing assumptions and expectations at the moment of choice
A decision journal records the specific beliefs and conditions that make a choice seem rational at the time you make it. You write down your key assumptions, the options you considered, which option you chose, and why. You also forecast what outcome would confirm your decision was right and what would prove it wrong.
This forces explicit articulation of mental models that usually remain vague. What exactly are you assuming about customer demand, competitor behavior, or market timing? By naming these, you create measurable reference points. Later you can check whether those assumptions held.
Calibrating confidence and learning from results
When you record your confidence level (say, 75% sure) at the time of decision, you create the data needed to calibrate your judgment over time. If you're consistently right when you say 75%, then your confidence is well-calibrated. If you're only right 60% of the time, you're overconfident and should discount your future estimates.
Reviewing past decisions reveals which assumptions actually drove the outcome and which were irrelevant. Over hundreds of decisions, this feedback loop trains better intuition. You stop overweighting factors that sounded important but proved irrelevant, and you learn to weight the factors that actually mattered.