Choosing how to describe an estimate

The kinds of answer you can give, and why the kind decides which shapes are allowed.

For each uncertain number the app asks “How sure are you?” You never have to pick a statistical distribution; the app chooses a sensible shape from your answer. You can change it later.

The kinds of answer

What you say	What the app does with it
”It is a known, fixed amount” (you must give a reason)	Uses that exact value
”Lowest possible, most likely, highest possible”	Treats the lowest and highest as hard limits nothing can go beyond
”Low, middle, high for the likely range”	Treats low and high as a 1 in 10 chance below or above. More extreme values are still possible
The same, for a percentage	Same, but kept strictly between 0% and 100%
“A few distinct outcomes” (pass or fail, low, medium or high tier)	Each outcome gets its own chance. The chances must add to 100%
Observed data (a pasted list)	Builds the range from your observations
A formula from the function library	See Formulas
”Not sure yet”	Blocks the run. An unknown never silently becomes zero

Hard limits versus a likely range

This is the most important choice, because it decides whether your numbers are walls or only a likely range.

• Hard limits (lowest, most likely, highest): only shapes that stop at those limits are allowed. See Bounded shapes.
• A likely range (low, middle, high): only shapes with open-ended tails are allowed, so a surprise outside your range is still possible. See Skewed and percentage shapes.

The app refuses to mix these up. Treating a likely range as hard limits would quietly rule out the surprises you said could happen.

The read-back

Every estimate is read back to you in a sentence, for example “There is a 1 in 10 chance this is below 4 hours and a 1 in 10 chance it is above 12”. Check it before you move on.

Next: Bounded shapes