The real question should be: is there a clustering of purchases of those numbers for this draw that is different than the same numbers on other draws? If this was a planned scam, the distribution of purchases would be more tightly coupled than normal.