I suppose one could buy a 100 tickets and perform a chi-squared test ;) Even a cursory glance of Powerball frequency charts shows no discernible pattern. Look

I suppose one could buy a 100 tickets and perform a chi-squared test ;) Even a cursory glance of Powerball frequency charts shows no discernible pattern. Looking at the last winning numbers, they certainly seem clustered at the high end. Still uniform random for a given sample. But certainly not evenly distributed. Question you really want to be asking is: does my likelihood increase if I select the most frequently picked numbers? Advantage is probably minuscule. Although, I believe there is a strategy in table roulette that exploits "runs" in which the same number is drawn consecutively. Which for small sample sizes has a much higher probability of occurring than a long non-repetitive sub-sequence. It would be the scandal of the century if the entire lottery were faked. Centralized database of numbers played. And then a drawing that results in a "lottery actor" winning for show. Of course the true scandal is that the real winner of every jackpot is the IRS!