Yeah - I kept looking for the meat of the argument. The graph of random data looks a little correlated and I’d love to know why.

That being said I loved the mercury/Glasgow explanation. Anecdotally I see that all the time.