Behind the result - No mystery in the method
The count is based on the original text, not a normalized numeric display. Leading zeros and signs do not count. After the first nonzero digit, written trailing zeros count when a decimal point or e notation is present. For ordinary integers without either, trailing zeros are treated as placeholders. All-zero inputs use the written decimal-place count, with a minimum of one, and are flagged as convention-dependent. Rounding uses an exact power-of-ten grid, nearest integer with half ties away from zero, and a scientific coefficient padded to the requested figure count. Increasing the requested count only adds formatting; it does not add measured information.
This tool does not store inputs, fetch live data or replace a teacher or textbook explanation. Blank optional fields are omitted. Text is displayed literally, not interpreted as HTML.
Changing an input clears the old answer so it cannot be mistaken for the new calculation. Download text makes a local copy of the current working. Print / Save PDF uses the browser’s print controls. Review any note before sharing.
Before you go
How many significant figures are in 1200?
Under this tool’s explicit convention, plain 1200 has two. The trailing zeros are ambiguous without context. Write 1200. or 1.200e3 to make four figures explicit, or use the notation your teacher requests.
Is 0.000 counted in a universal way?
No. All-zero precision is convention-dependent. This tool counts the written decimal places, with a minimum count of one, and explains that choice. A count is a statement about writing, not proof of measurement uncertainty.
Which tie rule is used for rounding?
Half away from zero: 1.25 rounded to two figures is 1.3, and −1.25 becomes −1.3. This is not round-half-to-even. Carry is handled, so 9.99 to two figures becomes 1.0 × 10^1.
Does adding zeros make my measurement better?
No. A requested display such as 1.2000 does not establish that those digits were measured. The tool separates value change from formatting and cannot recover missing experimental precision.