[crimsonalucard]

Correlation does not imply causation.

Little known trick to establish that A causes B is to trigger A at will and see if a B is observed.

If you trigger A 100 times and it precedes a B 98 times then you have established causation to a degree. Similar to the correlation coefficient there must be some causation coefficient.

[rightbyte]

You might imply it, but not triggering A should give less B than triggering A, too.

"Turning on my TV at 5 a'clock causes channel 4 to broadcast The Simpsons."

[crimsonalucard]

Good point. If B was just happening all the time, then a B would always happen after A. Have to make sure the inverse is true.

Though if B happens all the time, but also A causes B. Then there's really no way to establish causation.

"The Channel four studio has electrical sensors that will trip and broadcast the simpsons either at 5pm or when I turn on my TV at 5pm."