Every combination has the same chance
In a hypothetical game where you pick 6 numbers from 49, there are 13,983,816 possible tickets. If the draw is fair, each one is equally likely to come up. That means 1-2-3-4-5-6 has the same chance as 8-17-23-31-40-45: 1 in 13,983,816 each.
This surprises many people because the first looks “obviously not random” and the second looks “random”. But a draw does not produce an impression of randomness. It picks one combination out of the full list, and the process cannot tell which entries look tidy to a human. The fact that a result looks ordered is a fact about our eyes, not about the draw.
Why patterns feel meaningful
Humans are strong pattern spotters, and that usually helps. In a random setting it leads us astray, in a few predictable ways.
Special sets are easy to name. Only a handful of combinations look patterned, such as a straight run, a single column on the playslip or all multiples of a number. Many millions look unremarkable. When a patterned set is drawn, we notice it. When an unremarkable set is drawn, no one remarks that it was just as unlikely.
We judge by how typical something looks. A mixed-looking set resembles what we expect from a random process, so we call it more likely, even though it is not. This is a feature of how people judge chance, and it ties into how risk perception can differ from the arithmetic.
We expect short sequences to look balanced. People often feel that a draw with all low numbers or all odd numbers is “unlikely” to appear. Some such sets are rarer as a class than mixed sets, but the key word is class, as the next section shows.
Classes versus single combinations
A set of tickets can be rarer than another set of tickets without any single ticket being more or less likely. Here is a concrete case in the 6-from-49 illustration.
A run of six consecutive numbers can start at 1, 2, and so on up to 44, which makes 44 combinations. The chance that the draw is some run of six is therefore 44 in 13,983,816, which is exactly 1 in 317,814. That sounds rare, and it is: runs of six are a rare class.
But each of the 44 runs has the same chance as every other individual ticket. The class is rare because it is small, not because its members are special. A ticket covering one run has a 1 in 13,983,816 chance, just like any other ticket.
The same logic applies to smaller patterns. Some patterns, like having at least one pair of neighbouring numbers, are common. The number of six-number sets from 49 with no two consecutive numbers is 7,059,052, which is a standard result from counting. Subtracting from the total, 13,983,816 − 7,059,052 = 6,924,764 sets contain at least one neighbouring pair. That is about 49.5 percent of all combinations, so roughly half of draws include at least two consecutive numbers. Someone who avoids all consecutive numbers because “they never come up” is ignoring the fact that about half of the draws contain a pair.
Hot numbers, cold numbers and “due” numbers
Charts that list how often each number has been drawn are common. They show differences, because any random process produces uneven counts over a finite number of draws. Some numbers come up more than others and some less, simply by chance. Reading the chart as a prediction involves two mistakes:
- Hot numbers will keep coming. Without any evidence that the draw is biased, a number’s recent frequency tells you nothing about the next draw. How far past counts can tell you about the future is the question behind the hot hand belief.
- Cold numbers are due. The idea that a number must catch up is a gambler’s fallacy. The balls or random number generator have no memory, so a long absence does not raise the next probability. Over many draws counts do even out in proportion, but that does not mean the next draw compensates; this is the point of regression to mean and of the law of large numbers.
A fair draw has no pattern a player can use. If a machine were physically biased, it might be detectable, but a properly run lottery tests its equipment and procedures to avoid exactly that, and small unnoticed biases would not be something an ordinary player could exploit.
What patterns can change
Patterns do not alter the chance of winning, but they can alter the amount you receive if you do win. If many players pick the same visible pattern, they split a prize when it comes up, as shown in shared jackpots. That affects the payout, not the odds. It is therefore not a way to improve your prospects, and no choice of numbers turns a ticket into a good bet.
The same goes for the method of choosing. A quick-pick ticket, a favourite set of numbers and a pattern drawn on the slip all have identical probabilities.
Keeping it in perspective
Looking for patterns in lottery results is a harmless puzzle, and it can be an entertaining exercise in checking claims. It becomes a problem if it leads you to spend more because you believe you have found a system. Nothing you can find in past draws will make future ones predictable. Keep any spending within a limit you set beforehand, and do not buy extra tickets to make up for earlier ones. If gambling is causing you difficulty, a local support service can help.