Number Frequency
How often each number has come up
The dashed line marks what you'd expect if every number were equally likely. Real draws bounce around that line — they don't sit on it exactly, because randomness has noise too.
Chi-Square Test
Are the draws actually uniform?
This tests whether the unevenness above is the kind you'd get from pure chance, or something more. A p-value above 0.05 means no evidence of bias — exactly what a fair, well-shaken machine should produce.
Hot Numbers
Do they stay hot?
Split each dataset into an earlier half and a later half. If hot numbers were real, a number frequent early should stay frequent later. We correlate the two halves' counts — near zero means the past tells you nothing.
Regression
Does one draw predict the next?
For every draw, we sum its main numbers and try to predict the sum of the very next draw — the simplest regression you can run. Each dot is one (this draw, next draw) pair. A flat, shapeless cloud is what "no relationship" looks like.
All numbers here are real historical draws from official draw-history archives. Want to pick some of your own anyway? Back to Lottery Quick Pick →