Skip to main content

Distribution Fitter overview

Distribution Fitter takes a column of real data, identifies which probability distribution best describes it, and saves that distribution to a shared library you can plug into every simulation tool in PocketStats.

What it does

You upload a CSV or XLSX file, pick a numeric column, and the Fitter runs a panel of candidate distributions (normal, lognormal, gamma, weibull, poisson, etc.) against your data using maximum likelihood. It returns a ranked table — best fit first — alongside a histogram with the fitted curve overlaid and three diagnostic plots (Q-Q, P-P, CDF). When the answer looks right, you save the fit to your org's Distribution Library with a name like Monthly revenue (lognormal). From there, Process Map, Spreadsheet Sim, and Probability Playground can all insert it with one click.

Who it's for

Anyone running a simulation who has been guessing at distribution parameters. If you've ever stared at a Monte Carlo assumption and typed in mean=100, std=20 because it felt right, the Fitter replaces the guess with data.

The hero use case: turning a column of historical task durations, monthly revenues, or daily call volumes into a grounded assumption for downstream simulation.

note

Distribution Fitter requires a Pro plan or an active trial to create new fits. Anyone in the org can read existing library entries even after a trial expires.

Next steps

  • Core concepts — the terms and mental model behind fitting, scoring, and the library
  • How-to — picking a column, configuring candidates, re-fitting, and using the Distribution Library
  • Reference — AIC, BIC, KS, the diagnostic plots, and the supported distribution catalog