My approach is merely statistical

One of the biggest uses for a program like I described is for determining whether or not lab results from tests and studies are true and not made up.

When you do an experiment on the melting point of say copper, you won't get the same result each time you measure it due to inherent error of the lab process. But you will get an average result that will be fairly close. BUT, some underhanded people have been known to falsify their results, perhaps in favor of their own theory. So you take the results you got and start messing with them, making up numbers to give you a more favorable result. However the human mind is drawn to patterns, and has a hard time making random numbers.

There are computer programs out there in use by professionals that look at these lab results and can give a statistical probability that they aren't random and were PROBABLY manufactured. These programs are designed to look for patterns in number systems that aren't random.

Oh by the way. The chances of you pulling those numbers out of a hat are:
1 in 4882200000000000000000000000000000000000000000000000000000000000