ok has anyone accually thoughtthis out.. more than likely its just gonna be a 5 randoms and the results of those randoms = the combination.
AND if that is the case the odds of winning rank 5 are 1,000,000,000,000,000 to 1 (1000*1000*1000*1000*1000) which honestly i dont think ANYONE is gonna win it.
If that were the case, then your possible results would not be from 0 to 99,999, but would be from 0 to 999,999,999,999,999. Your math's faulty.
A realistic possible implementation along your lines, would be 5 randoms of 0 through 9 and get the answer one digit at a time. But that would be 10*10*10*10*10 = 100,000, which is precisely the odds that we know it is anyway, so it's really just an overcomplicated way of getting to the same point.
There are 100,000 individual numbers between 0 and 99,999 inclusive. Precisely one of those numbers is the grand prize, therefore a single contest entry has 1 in 100,000 odds of winning the grand prize. You get into slightly (not significantly) more complicated statistics when you discuss your odds of winning with more than one contest entry, of course.
If you repeatedly pick the same number, the odds are unchanged. You don't have any more chances to win, but if you do win, you win bigger. (It seems that one can theoretically win a billion gil per character if the person buys 10 tickets with the same number, and those all turn out to be the winning number. However, the odds of the win are still 1 in 100,000.)
And if you pick a number guaranteed to be different than any of your previously picked numbers, then your chances in 100,000 increase by 1. Ten different numbers means a 10 in 100,000 (i.e. 1 in 10,000) chance of winning the grand prize.
Finally, if you pick several distinct numbers, but then get more than one ticket for at least one of those numbers, then you just apply the same logic as above. Ignore duplicates, and divide the number of unique numbers you have by 100,000, and there are your odds.
The math only really gets messy if you don't specify whether you're picking different or the same numbers. (All the above cases have later numbers *dependent* on earlier numbers. That is, you are deliberately picking something different or deliberately picking something the same. The math is a little messier with *independent* numbers.)