# Argument selection and reproduction

For a standard normal random variable $Z$, $\Pr(|Z|\leq n)=\operatorname{erf}(n/\sqrt2)$ and $\Pr(|Z|>n)=\operatorname{erfc}(n/\sqrt2)$ CITE{NaturalArguments}. Integer standard-deviation thresholds select these arguments. They are conventional probability references, not arbitrary probability quantiles or mathematically unique cutoffs.

it holds the origin and the one- through six-standard-deviation references $x=n/\sqrt2$, $1\leq n\leq6$, under CITE{argument-selection}; the selection is not exhaustive over the real domain

References: https://dlmf.nist.gov/7.20.iii

Install numberdb, python-flint and mpmath. Run `python generate.py` to verify the stored selection or `python checks.py T197 --live` for independent checks. The default never writes. `--fill` sends entries; `--remove-unselected` is a separate explicit choice. It does not approve or publish a table. Historical reference grids remain available in the website revision history.