Law of Large Numbers
The principle that observed averages tend to move closer to expected averages as the sample grows.
It explains why RTP simulations become steadier over thousands of plays while ten spins can look completely different. Convergence is gradual and never guarantees a short-session result.