> Repeating it ad infinitum does that.

Take a moment to think about it. Why does the confidence interval go smaller with more samples? What does it say about the impact of luck in repeated tests?