17-07-2026, 12:17 AM
One problem I see here is that any value of the initial comparisons is diluted by the later comparisons that are clearly just coincidental. This makes it more likely (or obvious) that the earlier comparisons are also coincidental.

