1 On the other hand, we frequently find situations where a small set of rv’s, say \(W, X, Y, Z\) satisfy the Markov condition that \(\operatorname{Pr}\{Z \mid Y, X, W\}=\operatorname{Pr}\{Z \mid Y\}\)...1 On the other hand, we frequently find situations where a small set of rv’s, say \(W, X, Y, Z\) satisfy the Markov condition that \(\operatorname{Pr}\{Z \mid Y, X, W\}=\operatorname{Pr}\{Z \mid Y\}\) and \(\operatorname{Pr}\{Y \mid X, W\}=\operatorname{Pr}\{Y \mid X\}\) but where the conditional distributions \(\operatorname{Pr}\{Z \mid Y\}\) and \(\operatorname{Pr}\{Y \mid X\}\) are unrelated.
