Well from my understanding the core problem is the statistical models used to calculate their probability were completely wrong. So we had “hard” math “proving

Well from my understanding the core problem is the statistical models used to calculate their probability were completely wrong. So we had “hard” math “proving” they were once-in-a-million-years unlikely, and that lasted until we had overwhelming observational evidence to the contrary (in this case buoy data)