Are Lottery Numbers Ever Due?

What independence actually means

Two events are independent when knowing the outcome of one tells you nothing about the other. Lottery draws qualify, and the reason is worth stating precisely.

Between draws, every ball goes back into the machine. The pool is restored to its full size and its full composition. Whatever happened last Saturday left no residue in the equipment, no counter, no tally. A ball is a piece of moulded plastic; it has no mechanism for remembering that it has been idle.

So the probability that ball 17 appears in a 6-of-49 draw is 6/49, or about 12.2%, in every single draw. After 17 has skipped a hundred draws in a row, it is 12.2%. After 17 has come up three draws running, it is 12.2%. The number does not move because there is nothing in the physical setup that could move it.

The gambler's fallacy, and how well documented it is

The belief that a due outcome becomes more likely has a name and a research literature. Clotfelter and Cook studied Maryland's daily numbers game, where players bet on three-digit numbers, and found that the money wagered on a specific number fell sharply immediately after it was drawn, then recovered gradually over several months.

Read that again with the arithmetic in mind. Players were systematically abandoning a number precisely because it had just won, on the belief that it could not win again soon. In a game where every three-digit number pays the same and has the same 1-in-1,000 chance every day, this behaviour costs nothing on average and gains nothing. It simply shows how widespread the belief is, in real money, at scale.

Tversky and Kahneman gave the underlying error its classic description in 1971, calling it belief in the law of small numbers: the intuition that a short run of outcomes ought to look like the long-run average. Chance has no such obligation. Balance emerges over enormous numbers of trials by dilution, not by correction.

The regression fallacy hiding inside it

There is a real statistical fact nearby, and confusing the two is what keeps the belief alive.

Over a very long run, each number's share of appearances does converge on the expected rate. That is the law of large numbers, and it is genuinely true. But it works by drowning early imbalances in a rising tide of later draws, not by staging a comeback.

Suppose ball 17 is 40 draws behind after 2,000 draws. To fix that gap by "catching up", 17 would have to be drawn more often than chance allows, which requires a mechanism nobody has ever found. What actually happens is that after 20,000 more draws the same absolute gap of 40 is a much smaller fraction of the total. The imbalance never gets corrected. It gets outgrown.

What an overdue list actually measures

Every lottery statistics site publishes a most-overdue table. It is not fabricated data. It is an accurate record of which balls have gone longest without appearing, and it is entirely useless for prediction.

Here is the part that surprises people: in a fair game, some number always has to be the most overdue. It is a mathematical certainty that the list exists and is non-empty. The gaps you see are the gaps a random process produces.

For a 6-of-49 game, a specific ball's chance of missing a given draw is 43/49, about 87.8%. The chance it misses twenty draws in a row is 0.878 to the power of 20, roughly 7.5%. Across 49 balls you would therefore expect three or four of them to be on a twenty-draw drought at any moment. Finding one is not evidence of anything. Finding none would be the strange result.

The hot-hand version of the same mistake

Some players run the logic backwards and chase numbers that have come up recently, on the theory that the machine or the ball set is favouring them.

This is the same error wearing the opposite coat, and it fails for the same reason: independence. Neither the drought nor the streak carries information about the next draw.

There is one narrow case where a streak would mean something, which is why operators test for it. If a ball set really were physically biased, through a manufacturing defect or wear, its favoured numbers would appear more often over the long run. Lottery ball sets and draw machines are regularly weighed, measured and audited exactly to rule this out, and draw equipment is randomly selected before each draw. Published audits do occasionally flag a period that fits the fair model badly: Coronel-Brizio and colleagues found two such periods in Mexican and Italian draw histories in 2008 and asked that the sources be chased down. Notice which way that cuts. Deviations get found and investigated, which is exactly what makes them useless to a player, and if a real bias ever were confirmed the operator would replace the equipment rather than let anyone exploit it.

Why the intuition is so stubborn

Knowing the maths does not dissolve the feeling, and that is worth explaining rather than scolding.

Ask someone to write down a plausible run of thirty coin tosses from imagination and they will alternate far more often than a real coin does, and they will refuse to write six heads in a row. Real coins produce that run regularly. Our sense of what randomness looks like is built from short sequences, and short sequences of genuine randomness look lumpy and wrong to us.

The deeper reason is that the intuition is usually correct. Almost nothing in ordinary life is independent. A bus that has not come for twenty minutes really is more likely to arrive soon. A shop out of stock for a week really is more likely to restock. A machine that has run for years without failing is more likely to fail, not less.

So the heuristic that produces the gambler's fallacy is a good heuristic in the world it evolved for. Lottery draws belong to a small and unusual class of systems with no memory and no schedule, and the rule that serves you everywhere else fails precisely there.

That framing also explains why the belief survives contact with an explanation. People are not being stupid when they feel a number is due. They are applying a reliable rule to one of the few places it does not apply, and no amount of arithmetic makes the feeling go away.

What to do with all this

The honest conclusion is freeing rather than deflating: there is no homework. Nothing to track, no spreadsheet of gaps to maintain, no cold numbers to monitor. Since every combination carries the same probability, the effort saved is pure gain.

If you enjoy having a system, pick one that costs nothing and gives you something you actually want. Our daily lucky numbers are seeded from your birth date and the date, so they are stable for the day and different tomorrow. They are not predictions and we will never dress them up as any. They are a ritual with the arithmetic shown.