Probability Range vs Point Estimate
Use a probability range when evidence supports a region but not a precise number, and show which assumption controls its width.
Example: renewal probability
Historical renewal is 70 percent, but a pricing change and small sample widen the credible region to 50 through 75. Both options are tested across that interval instead of using 68 percent as fact.
Precision can exceed knowledge
A 63 percent forecast looks measured even when inputs support only something like 45 to 70 percent. A range makes uncertainty visible and prevents minor numerical differences from driving a decision.
Create the range
- Define the event and deadline.
- Choose a base rate.
- Vary the strongest uncertain adjustment.
- Record the central estimate and bounds.
Ranges must still be disciplined
A zero-to-one-hundred range avoids responsibility. Bounds should correspond to plausible evidence conditions and narrow when evidence improves.
The decision rule
What to record before acting
Use the narrowest range justified by evidence, then test whether the decision changes within it.
Apply the rule in context with compare options across ranges. Keep the evidence and assumptions with the decision so an outcome review can distinguish reasoning from luck.
Continue within evidence and uncertainty
How Confidence Intervals Inform a Decision
Use a confidence interval to understand sampling uncertainty, not as the probability that the true value sits inside one computed interval.
How Confidence Intervals Inform a DecisionCompare Evidence From Small Samples
Evaluate small-sample evidence through case selection, measurement quality, effect size, consistency, and how much the decision depends on it.
Compare Evidence From Small Samples