Confidence interval
What is confidence interval?
A range around a forecast expressing how uncertain it is. A 95% interval is one expected to contain the true outcome 95 times out of 100.
Also known as: error bar, prediction interval
A difficulty forecast quoted as a single number does not show how far from the outcome it could be. Early in an epoch, when few blocks have been mined, that gap can be large.
A range calculated from theory and one fitted to historical data both need testing against outcomes. Compare how often the ranges contain the outcome with their stated confidence level.
Quick answers
- What does a 95% confidence interval mean?
- Across many forecasts, about 95 in 100 such intervals are expected to contain the true outcome. It shows how uncertain a forecast is, which a single number cannot.
- How do you check a forecast range?
- Compare the published ranges with the outcomes over a stated evaluation period. Report how often the ranges contained the outcome alongside their widths. Fitting a range to past data does not by itself show how it will perform on new forecasts.
Last reviewed 2026-08-20
Last reviewed . Figures marked illustrative are placeholders, not live data.
