Three Common Lottery Patterns That Are Mathematically Meaningless

By the photocopier, a colleague leaned over to tell me the number 17 was 'due a good run.' She meant it kindly, and she is far from alone, because half the staffroom has a theory. Here is the short version before any of the lottery probability gets going: the three patterns people swear by are mathematically meaningless, and a little data skepticism saves you a lot of disappointment. I play EuroMillions myself, so this is not me peering down from some statistical high horse. It is just what the numbers actually say once you strip out the wishful thinking and the mathematical fallacies that get dressed up as systems.

Every meaningless pattern shares one flaw

Strip away the colour-coded dashboards and the talk of streaks, and all three patterns I am about to describe fail for the same reason: each draw is independent. The machine in Paris has no memory. Those plastic balls do not know what they did last week, last month, or last year, and nothing about a past result nudges the next one. That is the whole of it, independence, and it is the single idea I most want a EuroMillions player to carry around.

Think of a coin. Flip a 10p piece and get heads five times running, and the sixth flip is still an even probability of heads or tails. The coin feels no pressure to balance itself out. Lottery balls work the same way, with every number carrying the same chance on every single draw, no matter how the recent history looks. Once that clicks, most of the systems sold online quietly fall apart.

'Overdue' numbers are not actually due

This is the one I hear most, and the honest answer is no. The belief that a number missing for twenty draws is somehow building up pressure is the textbook Gambler's Fallacy, and it is the coin-flip mistake wearing a different hat. A number that has not shown in months is no more likely to appear on Friday than one that turned up last week.

I learned to distrust the idea the practical way. At one point I downloaded an app that promised to track 'overdue' numbers due for a reappearance, flagging them in urgent red as high priority. It looked authoritative. It was also, in plain terms, selling me the gambler's fallacy behind a tidy interface, because the flagged numbers came up no more often than anything else, and chasing them changed nothing except how much attention I wasted. For the full, unglamorous comparison, I have written it up in The Teacher's Notebook: 6 Months of Testing AI Against the EuroMillions, where the 'due' numbers never once outperformed plain random picks.

The usable rule is simple. If a tool's main selling point is telling you which numbers are 'due,' treat that as your cue to ignore that feature entirely.

Hot streaks, cold spells, and what they actually tell you

A 'hot' number is just the overdue myth turned over, the number on a tear rather than the one in hiding, and it was exactly what my colleague had pinned her hopes on. Whether you misread a run as hot or as cold, you are making the same error about independence, and I have gone deeper into why hot and cold numbers get so badly misunderstood in a piece of its own. For here, the short version holds: a recent streak predicts nothing.

None of that makes counting pointless. The frequency chart pinned above my desk gets brought up to date in red biro after each result, and I once spent a fair while building my own EuroMillions frequency chart from the public draw history, the laptop fan winding itself up every time I set it sorting through years of CSV files. It is genuinely interesting work. It simply does not predict the next draw, and that distinction is the whole game.

Here is the one place the data earns its keep, though not in the way the sellers claim. Picking numbers above 31 will not improve your odds by a hair, but it does change who you might share a prize with. Loads of people stick to birthdays, which caps them at 1 to 31, so a winning line built from higher numbers tends to be split among far fewer people. The win itself is no likelier; it is simply worth more if it lands.

A friend from my undergraduate maths days, now a software developer in Leeds, is thoroughly sceptical of lottery AI for this very reason. She points out that a model trained on random data can only ever learn the randomness, a point I have chewed over separately on whether lottery AI is real or just better-presented maths.

The shape you draw on the slip doesn't matter

Crosses, zig-zags, the four corners, a neat diagonal down the middle, people love a tidy shape on the play slip, and there is real comfort in symmetry. The draw machine, however, cannot see your slip. A run of 1, 2, 3, 4, 5 is exactly as probable as any scatter of awkward-looking numbers, because every combination sits at the same odds.

The catch is the birthday problem in disguise. Thousands of players choose the obvious shapes every week, so on the rare occasion one of them lands, the prize gets sliced into very thin pieces. Draw your art if it makes the ritual fun, since it costs you nothing in probability, but know that the prettier and more popular the pattern, the more strangers you would be splitting with.

Playing EuroMillions with your eyes open

So where does that leave a player who still fancies a line each week? Roughly where they started, which is the honest part. Across all the prize tiers, EuroMillions pays back only around 40 percent of the money staked, so every ticket carries a negative expected return by design, and that is built into the game rather than a flaw you can pattern your way around. Run the expected value on any pattern-based strategy and the underlying return does not budge, because the maths is simply indifferent to your method, something I work through properly in a separate piece on the real odds.

One reader, Neville, a retired civil engineer over Wigan way who logs every draw in a paper ledger, is adamant that the Lucky Stars should be tracked apart from the main numbers, and on that narrow point he is right, since they are drawn separately and deserve their own column. It still will not tell either of us what falls next.

Right then, my honest position as someone who still plays: enjoy the data if you find it interesting, treat any 'pattern' as entertainment rather than a promise, and if you want the one edge the maths actually supports, choose numbers other people avoid so a win does not get shared half a dozen ways. It will not make lightning likelier to strike. It just spares you splitting the prize with half the city.