When we do an experiment a large number of times the average result will be very close to the expected result.

In other words: in the long run random events tend to average out at the expected value.

Example: Flipping a coin

• Flip it just 10 times. We expect 5 heads. But 7 heads would not surprise us.

• Flip it 10,000 times. We expect 5,000 heads, but are *very* unlikely to get 7,000 heads. The result is likely to be within a few percent of 5,000.