Replies

So I challenged my son to calculate the odds of this.....he used SSN last four digits rather than Date of Birth since DOB would have 365 days times roughly 55 years of eligible voters gets the numbers into the stratosphere.....bear with the math....

The odds of two people having the same last four digits is 1 in 9,999

There are 150K surnames in the US, the top 100 of which account for 46M people, roughly 1/6. Let's be really generous and say that they're all top 100 surnames, and the odds of two people having the same surname are also 1 in 10,000

There are 5K different first names in the US. The top 100 account for 60%. 6 names cover 1/6 of the male population, so let's say 12 names total. 1 in 144

This is all way underselling the improbability

Odds of all three occurring together given the parameters: 1 in 14.4 Billion.

So there may be a few with similar names and dates of birth just by coincidence.....but not 35,750 times. No way

Your math is off by several orders of magnitude.

The odds of two people having the same SSN (last four digits) is close, not 9,999, but exactly 10,000 (you have to include 0000).

The odds, given your assumptions of only 100 last names are 1 in 100, not 1 in 10,000. (If I have surname #1 and start drawing out other people’s names, I’ll average surname #1 once in every 100 draws since there are only 100 choices—the 1 in 10,000 would be the odds of drawing the a specified surname twice in a row...not the same problem)

Similarly, if you assume 12 first names, the odds of two people having the same first name is 1 in 12. Again, the 1 in 144 would apply only if you specified the first name to be drawn, say John. There’s a 1 in 144 chance of drawing John twice...but this is a different problem. Given that my name is John, there’s a 1 in 12 chance of someone else being named John.

So, that cuts the odds to 1 in 10,000x100x12, or 1 in 12 million. Given that there’s 100 million or so in the database used, you should expect to find 8 or 9 that match. But again, you used generous assumptions, so more like 1-5 probably.

As for birth dates, assuming 365 days x 60years you would get 21,900 options. Replacing the 10,000 SSN figure with 21,900 would give around 1 in 25 million. Either way, a lot of people voted twice.