Luke_A_Somers comments on Why Bayesians should two-box in a one-shot