Markov's demonstration meant that observing convergence in social statistics does not prove the underlying decisions were independent, so such statistics don't prove free will at all; independence isn't even necessary to do probability.