Coin flippers

I see these Markov triples as deviation rates.  They are binomials distributions with an unfair coin.

The likelihood of picking one positon is 1/m, m beings the Markov number. When one of the rates need a coin toss, it flips it  m times, m being its rate.  Note the rates are actually interleaved but relatively [rime so I can segmen here. Any way, the counter has selected one, two, three  or nothing.

Each rate operates the same, but their coin toss rate must be relatively prime.  The mean vzlues of these expanded binomolial is always w * 1/w where we make p = w. The variance (w-1)/w.