Ask someone to pick five random numbers between 1 and 43. Then ask ten more people to do the same. You will not see a uniform distribution across the range. You'll see clusters: numbers that feel lucky, numbers tied to birthdays and anniversaries, numbers that were recent winners, numbers that 'feel' more random than others, and deliberate avoidance of patterns that actually have the same probability as any other combination.
This isn't a character flaw. It's a documented quirk of human cognition that affects everyone, including people who understand probability well. The lottery industry has understood this for decades. Here's what the research shows.
The Birthday Bias
The most well-documented form of lottery number bias is birthday clustering. A large proportion of hand-selected lottery tickets contain numbers between 1 and 31, the range that maps to calendar days. This makes mathematical sense as a personal selection strategy (people pick birth dates, anniversaries, significant dates in their lives) but it creates a meaningful statistical consequence.
For games with large number pools (Powerball goes up to 69, Mega Millions to 70), avoiding the numbers 32 through the top of the range means your selections are drawn from a much smaller portion of the available pool than the draw itself uses. If you happen to win with birthday-clustered numbers, you are statistically more likely to share the jackpot with other winners who had the same idea.
Shared jackpots don't reduce your odds of winning, but they do reduce what you take home. In 2016, a $1.6 billion Powerball jackpot was split three ways partly because many players avoid numbers above 31.
The Avoidance of 'Obvious' Patterns
Most players would never choose 1-2-3-4-5 for a lottery ticket. It looks wrong. It feels like it obviously won't be drawn. But in a random draw, 1-2-3-4-5 has exactly the same probability as any other combination of five distinct numbers from the available pool.
Research into human randomness production (asking people to generate random sequences) consistently finds that people avoid repeated digits, avoid long sequential runs, and produce distributions that are detectably non-random. When people try to act randomly, they introduce patterns that a statistician can distinguish from true randomness.
In lottery terms: the combinations most people refuse to play (pure sequences, all-even numbers, all numbers ending in the same digit) are no less likely to win than any other combination. By avoiding them, players are making decisions based on pattern aversion, not probability.
The 'Hot' and 'Cold' Number Fallacy
It's common for lottery players to track which numbers have appeared recently (hot numbers) or haven't appeared for a long time (cold numbers), and to use that history to inform their picks. The reasoning varies: hot numbers are 'on a streak,' or cold numbers are 'due.'
Neither of these strategies is grounded in probability. Each lottery draw is an independent event. The balls have no memory of previous draws. A number that has appeared in the last ten draws has exactly the same probability of appearing in the next draw as a number that hasn't shown up in a year.
The 'due' number fallacy is a variant of the gambler's fallacy, the mistaken belief that independent random events are influenced by their history. It's intuitive but mathematically wrong.
The 'Lucky Number' Effect
Across cultures and individuals, certain numbers carry a sense of luck or significance: 7 in Western cultures, 8 in Chinese tradition, numbers tied to religious significance, and so on. These cultural associations are real in terms of human psychology. They're meaningless in terms of lottery draws.
In practice, numbers perceived as 'lucky' are overrepresented in hand-picked lottery entries. This means that if one of these numbers is drawn, more players share in any prize at that tier, which reduces individual payouts.
Why This Matters for Jackpot Splits
At the jackpot level, these biases affect how many people share the prize when a specific combination is drawn. A jackpot combination made up entirely of numbers between 1 and 31 will statistically be shared among more winners than a jackpot combination that includes several numbers in the 32-69 range.
This doesn't mean you should deliberately pick high numbers to avoid sharing a jackpot. The overwhelming probability in any given draw is still that you don't win at all. But it's a real phenomenon that affects jackpot split frequency.
What a Random Generator Actually Does
A properly implemented random number generator sidesteps all of these biases by drawing from the complete legal pool with uniform probability. Every number is equally likely. The generator has no cultural associations, no memory of previous draws, and no aversion to 'obvious' patterns.
This site uses your browser's cryptographic random source, the same API used by password managers and HTTPS connections, designed to be statistically unpredictable and uniform. Each time you generate, you get a combination drawn from the full pool with no bias introduced by human psychology.
To be clear: using a random generator does not improve your odds of winning. The lottery draw itself is random, and no selection method changes your probability of matching those numbers. But a generator does ensure your picks aren't systematically biased toward combinations that are more likely to be shared. It also removes the mental overhead of trying to 'think randomly,' which research shows humans are not particularly good at.
Your odds of winning the lottery are the same regardless of how you pick your numbers. A random generator is a convenience tool. It's not a strategy, not a system, and not a way to improve your chances.