On meaningful coincidences and why unscientific beliefs can actually be entirely rational