It’s easy to mix up numbers and combinations — but in lottery math, they mean entirely different things.
A number is just one ball. A combination is what you get when several numbers are grouped together into one ticket entry. You can’t enter a lottery using a single number. The game requires a full set — a complete combination.
For example, 3, 15, 27, 39, 41, and 49 are separate numbers. Put them together and they become one combination — in this case, a 6-odd composition.
All Numbers Have Equal Probability
Any serious conversation about a lottery formula has to set aside “hot” and “cold” number theory. Random number systems don’t keep streak advantages going forever. Frequencies stabilize as the sample size grows.
This has a name: the Law of Large Numbers (LLN). The more times you run an experiment, the closer your actual results get to what probability predicts.1 Critically, the LLN describes convergence over many trials — it says nothing about what any individual draw will produce.
Canada Lotto 6/49 is one of the games I studied as a real-world example of this behavior. I ran the numbers against historical draw data from June 1982 to September 2018 — 3,688 draws across 36 years. In the early draws, I observed large gaps in how often each number appeared — some balls showed up frequently while others barely registered. As the draws accumulated across decades, those gaps quietly closed. That’s randomness doing exactly what the math predicted it would, just on its own timeline.
Look at the gap between balls 18 and 49. The pie chart below shows just how different their frequencies are during the first 30 draws of Canada Lotto 6/49.
Balls 11, 18, 28, and 35 jump out early. But as draws keep coming, the quieter numbers chip away at the gap. By about 100 draws the chart already looks much more even.
By around 500 draws, the frequencies even out further. By 1,000 draws, the chart is nearly dead-on what probability predicted. By 2018, nothing has changed. No bias, no rigged balls, no lucky numbers — just randomness, doing what randomness does.
When we track how often each of the 49 balls appeared across 3,688 real lottery draws, the distribution looks just like the chart below — every number ends up with the same chance of being drawn.
So where should we actually look, once lucky numbers and hot-cold superstitions fall apart? The answer starts with the combinatorial composition behind every possible lottery combination.
I built a PHP simulation using PHP’s random_int() function to generate large volumes of lottery combinations and observe how randomness distributes outcomes across many draws. The results are described visually at How Truly Random Lottery Games Exhibit Long-Run Statistical Behavior — a page that shows how combinatorial compositions behave and what a genuinely random process looks like when you run it at scale.
References