Random Wheel Tools
← Back to the main wheel

Distribution diagnostic

Randomness Distribution Test

This page repeats an equal-probability selection many times and compares observed counts with the mathematical expectation. It explains natural random variation; it does not certify a browser or drawing process.

Run local test

This is a local diagnostic simulation, not an audit or certification.

What the simulation can show

  • See why small samples rarely divide perfectly evenly.
  • Compare outcomes after 100, 1,000, or 10,000 independent local draws.
  • Check that every supplied outcome can be reached by the current browser session.

Run and read the test

  1. Enter two to twenty unique outcome labels.
  2. Choose a trial count; larger samples usually have smaller percentage deviations.
  3. Run the test and compare actual counts with the expected count, while remembering that a new run will differ.

Diagnostic, not certification

  • A balanced-looking table cannot prove that future selections are unpredictable or tamper-proof.
  • A large deviation can occur by chance, especially with few trials or many outcomes.
  • Independent review is required for regulated, financial, legal, or high-value drawings.

Randomness test questions

Should every outcome have exactly the expected count?
No. Exact equality is uncommon; random sampling naturally produces positive and negative deviations.
Why stop at 10,000 trials?
The cap keeps the page responsive and prevents an accidental long-running browser task.
Does this test send results anywhere?
No. Outcomes, counts, and calculations remain in the current page session.