PICKMYSIX
Pick my numbers Pick six

Lottery number patterns, and which ones actually matter

Do not play 1, 2, 3, 4, 5, 6. Do not play all odd numbers. Do not play a line that looks wrong. Those are lottery number patterns, and buried in the advice is a claim about the draws themselves — that these shapes come up less often than they should. It is testable, so we counted them across 11,469 rounds, and they come up about as often as the matrix says they should.

Two different claims, and only one of them is checkable

The first claim is that such a line is less likely to be drawn. It is false, and it is false for a reason that needs no data at all. A draw does not see the shape of a line. It produces 6 numbers from 49, and every one of the 13,983,816 possible sets is equally likely. 1, 2, 3, 4, 5, 6 has exactly the chance that 7, 19, 23, 34, 41, 45 has.

The second claim is about frequency, and it sounds like the same thing but is not. It says these shapes turn up in the record less often than the arithmetic says to expect. That is a statement about our archive, so we checked it.

What we counted

Each shape's probability is computed from the number matrix — exactly, in whole numbers, never estimated and never taken from the data it is about to be compared with. Then we counted how often it actually came up.

A shape appears in the table below only where the expected count reaches 20. Below that, observed and expected differ by amounts chance produces every day of the week, and printing the pair would invite you to see a gap that is not there.

Every shape, pooled across every game and matrix deep enough to support it. A z of about 1 is ordinary; you would need around 3 before anything was odd.
ShapeRounds Came upExpectedz
every number odd 9,002 150 148.2 0.15
every number even 9,002 146 131.5 1.28
contains two consecutive numbers 11,469 4,799 4,793.9 0.10
contains three in a row 7,958 326 331.5 -0.31

The most avoided shape of all happens in about half of every draw. Across 11,469 rounds spanning 12 different matrices, 4,799 lines contained two consecutive numbers. The matrices give an expected figure of 4,794. That is 0.10 standard deviations from expectation, which is another way of saying: exactly what should happen.

Read that again, because it is the whole guide. A consecutive pair is not rare and never was. In a 6/49 game roughly one draw in two contains one. People avoid the shape because it looks designed, and looking designed is not a property the drum can detect.

Three in a row is genuinely less common — 326 times in 7,958 rounds — but it is less common because there are fewer ways to arrange it, not because anything prevents it. The expected figure is 331.

All-odd and all-even are not the same probability, which catches people out. A pool running 1 to 49 holds one more odd number than even, so an all-odd line is slightly likelier than an all-even one. In our archive: 150 against 146, with 148 and 131 expected.

The same line, twice

Here is the one genuinely surprising thing in the archive. On Friday 23 December 2016, Bonoloto drew 21, 27, 30, 40, 43, 47. On Thursday 14 November 2024, almost exactly 8 years later, it drew them again.

Different draw numbers, different draw order, different supplementary balls, different jackpots. The same six numbers.

And it is not a coincidence worth calling a coincidence. With 3,538 draws held and 13,983,816 possible lines, the number of repeats to expect is about 0.45. We found 1. This is the birthday problem: you do not need anything like 13,983,816 draws before two of them collide, because every draw is being compared with every other one.

How much does a draw remember of the last one?

Nothing, and the record says so precisely. Across 3,537 consecutive pairs of Bonoloto draws, the average number of balls shared with the previous draw is 0.726. The arithmetic — 6 squared over 49 — gives 0.735.

Bonoloto: numbers shared with the immediately previous draw, 3,537 pairs.
Numbers repeatedHappened Expected
0 1,544 1,542.0
1 1,475 1,460.8
2 464 468.2
3 51 62.4
4 3 3.4

People do the opposite of both things here. Some strike out last week's numbers as used up; others play them because they are running hot. The record shows the draw doing neither. It repeats a number about as often as picking blind would.

What about playing a sum in a particular range?

This one is more interesting than it looks, because the underlying observation is true. For Bonoloto, the 6 numbers add up to 150 on average, and 90% of draws land between 96 and 204. Our archive: 151.0 on average, lowest 51, highest 255.

So a line summing to 60 really is unusual. But notice what that does and does not buy you. There are far more lines that add to 150 than add to 60, which is exactly why middling sums come up more often — and it is also why restricting yourself to them puts you in a bigger crowd. Each individual line is still equally likely. You have not improved your chance of winning. You have improved your chance of sharing.

That is the same arithmetic as the birthday problem, arriving from a different direction, and it is the only sense in which any of this is actionable.

Per game, if you want the workings

Every shape with enough expected weight to be worth a row, by game.
GameMatrixRounds ShapeChance Came upExpected
Bonoloto 6/49 3,538 every number odd 1 in 79 48 44.8
Bonoloto 6/49 3,538 every number even 1 in 104 33 34.1
Bonoloto 6/49 3,538 contains two consecutive numbers about 50% 1,774 1,752.0
Bonoloto 6/49 3,538 contains three in a row 1 in 21 169 168.7
Mega-Sena 6/60 3,086 every number odd 1 in 84 40 36.6
Mega-Sena 6/60 3,086 every number even 1 in 84 38 36.6
Mega-Sena 6/60 3,086 contains two consecutive numbers about 42% 1,296 1,299.0
Mega-Sena 6/60 3,086 contains three in a row 1 in 31 97 99.1
US Powerball 5/69 1,434 every number odd 1 in 35 41 41.4
US Powerball 5/69 1,434 every number even 1 in 40 38 35.5
US Powerball 5/69 1,434 contains two consecutive numbers about 27% 377 380.1
La Primitiva 6/49 1,334 contains two consecutive numbers about 50% 660 660.6
La Primitiva 6/49 1,334 contains three in a row 1 in 21 60 63.6
Mega Millions 5/70 944 every number odd 1 in 37 21 25.3
Mega Millions 5/70 944 every number even 1 in 37 37 25.3
Mega Millions 5/70 944 contains two consecutive numbers about 26% 243 246.9
El Gordo de la Primitiva 5/54 554 contains two consecutive numbers about 33% 175 182.8
Thunderball 5/39 142 contains two consecutive numbers about 44% 59 61.9
Loto 6 6/43 119 contains two consecutive numbers about 55% 54 65.1
UK Lotto 6/59 117 contains two consecutive numbers about 43% 52 49.9
EuroMillions 5/50 71 contains two consecutive numbers about 35% 29 25.1
Set For Life 5/47 71 contains two consecutive numbers about 37% 34 26.4
Loto 7 7/37 59 contains two consecutive numbers about 74% 46 43.9

Method

Computed by scripts/patterns.mjs from the archive published on our data page. Rounds are grouped by number matrix, because a shape's probability is a function of how many numbers are drawn from how many, and pooling two matrices makes both figures wrong while leaving both looking plausible.

Probabilities are exact combinatorics in arbitrary-precision integers. The count of lines with no run of three, for instance, is derived rather than simulated, and checked against brute-force enumeration on small matrices where every line can be listed.

Nothing here helps you pick. It is a record of what has happened, and every combination remains equally likely on the next draw. Generate a line if you want one, or read why hot and cold numbers do not work.

Last computed Wednesday 18 November 2026 across 11,581 rounds in 16 matrix windows.