Powerball · Reality check
Six common beliefs, each checked three ways: against the real 1,984-draw historical record, against what probability math predicts, and against seeded random simulations that model a fair process. Every result is descriptive — none of it says what the next draw will be.
“Numbers that appeared in the previous draw are more likely to appear in the next.”
Observed 0.38 repeats per draw; simulated histories average 0.36.
Draws repeat numbers at exactly the rate a fair process predicts. Repeats are not a sign that certain numbers are “hot.”
“A number that has not appeared for a long time is more likely to appear soon.”
Cold numbers appear in 6.2% of draws — the same rate as every other number (7.6%).
Long-absent numbers appear at the same rate as everyone else, in real and simulated history. An overdue number is not more likely.
“The most frequent number pair appears more than chance allows.”
Most common pair appeared 21×; simulated histories top out between 19–23.
With 1,225 possible pairs in a 6/49 game, some pair will always “lead” — random history produces the same. That pair is not more likely next time.
“Draws almost never contain adjacent numbers like 17–18.”
28.1% of real draws contained a consecutive pair — close to the math expectation of 26.5%.
Consecutive numbers are common: about half of all draws include at least one adjacent pair, exactly as a random process predicts.
“What was drawn last time changes the odds of the next draw.”
Overlap between consecutive draws: 0.38 vs 0.34 between random pairs — a gap of 0.04.
Consecutive draws overlap no more than randomly paired draws. There is no evidence one draw influences the next.
“A draw with a balanced odd/even split (e.g. 3 odd + 3 even) is more likely than skewed splits.”
30.2% of real draws had a 2-3 split vs 31.7% expected.
Balanced splits are the single most common split — but only because there are more balanced combinations. Every specific combination is still equally likely.
When all three agree, the belief describes ordinary randomness, not a pattern you can act on. Historical frequency never changes the probability of a future draw.