Consider 4 lights. Each pattern they form is a microstate.
Entropy is the count of microstates similar to the one whose entropy we seek. This is low entropy, since there are only 2 like this (all on, or all off).
This is high entropy, since there are 8 like this (only 1 on or 1 off).
In the actual definition, we use a log of the count instead of the count. The log converts the count into the number of questions we need to ask the universe to determine which exact microstate a system is in, when we know its macrostate.
Viewed this way, the second law is a statement of mathematics.
If a light flips, the 2 low entropy microstates are more likely to transition into one of the other higher entropy ones than vice versa.
Equity of transitions is where physics comes in.
“Chance is a more fundamental conception that causality”
— Max Born