To avoid some counterexamples, I'd prefer to use as my explanandum the more complex claim that ∀n∃m(m>n&m is odd and the light is on after t−m).
Press the back button on your browser to go back to the post