Skip to content

Hot and cold lottery numbers describe the past, not the next draw

Hot numbers have been drawn most often, cold numbers least. Those counts describe past draws. Every ball has the same chance in every draw, and you can test that yourself on real draw history.

Printed sheets of lottery draw results with frequency tallies marked in pencil, on a plain desk beside a calculator

Open the free draw statistics tools

How we source draw data

Are the most drawn numbers more likely to come up again?

No. A frequency count describes draws that have already happened. Each draw starts fresh, so every ball has the same chance as every other ball.

Random draws never come out even. Over a few hundred draws, some balls will sit ahead of the average and some behind, purely by chance. Those are the hot and cold numbers on any list.

A cold number is not overdue. The balls have no memory of earlier draws, so a ball that has gone quiet is no more likely to appear next time. Believing otherwise has a name: the gambler's fallacy.

Our counts start on the date a game's current rules began. For Oz Lotto that is May 2022, when the game moved from 7 from 45 balls to 7 from 47. Mixing the two would count two different games as one.

Numbered white balls resting in the collection tray of a lottery draw machine after a draw

Photo by Abbie Rowe on Wikimedia Commons (Public domain)

Where to test hot and cold lottery numbers yourself

Every figure in our lottery draw statistics comes from real draw history fetched from operator and government feeds. These free tools show the counts and test them, with no account needed.

Randomness tester

Runs a chi-square test on real draw history and reports a real p-value. It answers one question: are the counts more uneven than fair draws would produce?

See Randomness tester

Frequency and z-score analyser

Shows how often each number has been drawn and how unusual each count really is. A z-score is the distance from the average count, measured in units of the normal spread.

See Frequency and z-score analyser

Draw simulator

Simulates many draws to show how results spread out by chance. Run it once and hot and cold numbers appear in draws that are fair by design.

Oz Lotto statistics

Counts for every Oz Lotto draw since the 7 from 47 rules began in May 2022, taken from operator and government feeds. Reach it from our Australian lotto odds page.

See Oz Lotto statistics

Hot, cold and overdue lines

One of seven generator methods on every game. It builds lines from real draw counts and shows those counts. It does not change the odds, and no generator can.

See Hot, cold and overdue lines

Queensland Keno pairs

Shows how often pairs of Keno numbers have been drawn together. Read it with the lucky-pair answer below in mind: searching thousands of pairs always turns up a few standouts.

How a chi-square test tells luck from a real pattern

A chi-square test checks whether the counts sit further from even than chance alone would produce. Our randomness tester runs these steps on real draw history and shows the result live.

  1. Count every draw

    We take each ball's count from real draw history, starting the day the game's current rules began. That is the lottery number frequency you see on our statistics pages.

  2. Work out the even share

    If every ball were equally likely, each would end up with roughly the same count. That even share is the expected count: the total balls drawn divided by the number of balls in the pool.

  3. Measure the distance from even

    For each ball, take the gap between its real count and the even share, square it, and divide by the even share. Add those up and you have the chi-square statistic: one number for how uneven the counts are.

  4. Turn it into a p-value

    A p-value is the chance of seeing counts at least this uneven if the draws were completely fair. A large p-value means the unevenness is ordinary. A small one means it is unusual enough to check again.

  5. Ask one question, not thousands

    Every test has some chance of a small p-value by luck alone. Run enough tests and a few will look unusual. That is why we test the whole set of counts at once, rather than hunting for a standout ball.

Every claim is backed by a calculation shown on the page or a linked primary source; anything that cannot be sourced stays off the site.

Jackpot Metrics, Our sourcing rule

See what we never publish

What people ask about hot and cold numbers

Short answers first, with the reasoning after. Each one points to a tool where you can check it on real draws.

Frequently Asked Questions

Do hot or cold numbers improve my chances?
No. Every ball has the same chance in every draw. Our hot, cold and overdue generator builds lines from real draw counts and shows those counts, and our randomness tester runs a chi-square test on the same history, so you can see for yourself whether any pattern is more than chance.
Why does a lucky pair keep turning up?
Because someone went looking. Oz Lotto's 47 balls make 1,081 possible pairs, and some games make thousands. Check every pair and a few will have come up together far more than average by chance alone. Finding a standout pair is expected. Finding none would be the surprise.
Is a cold number overdue?
No. A draw machine has no memory of earlier draws, so a ball that has gone quiet has the same chance as the rest next time. Expecting it to catch up is the gambler's fallacy: the belief that past random results change future ones.
What does a p-value tell me, and what doesn't it?
It tells you how often fair draws would produce counts at least as uneven as the real ones. It does not tell you which ball comes next. A small p-value on its own does not prove a draw is unfair; it says the result is worth checking again on more draws.
Where do your draw statistics come from?
Only from real draw history fetched from operator and government feeds. Each game's history starts on the date its current rules began, and every page shows its source and the date it was last checked. We do not publish a count we cannot source.
Is there any reason to choose particular numbers?
Not for the odds, which no choice changes. Your choice can only affect how many people you might share a prize with. Lines that avoid birthdays and sequences have identical odds and fewer people to share a win with, which is what our unpopular-number generator builds.
Are the tools free?
Yes. All 25 tools are free, run in your browser and need no account, including the randomness tester and the frequency and z-score analyser.

Spot a count that looks wrong? Tell us.

Ask about a test result, a source or a term on this page.

Get in touch

See where to get gambling help

Errors are checked against the source and fixed.