p_of_hit=Uniform(0,1) p_of_hit binomial=Binomial(7,p_of_hit) Copy_of_p_of_hit=Uniform(0,1) Copy_of_p_of_hit Copy_of_binomial=Binomial(7,Copy_of_p_of_hit) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 Copy_of_Node5 Node5 Copy_2_of_Node5 Copy_3_of_Node5 Copy_6_of_Node5 Copy_5_of_Node5 Copy_4_of_Node5 Copy_of_binomial Node1=Node5+Copy_2_of_Node5+Copy_3_of_Node5+Copy_4_of_Node5+Copy_5_of_Node5+Copy_6_of_Node5+Copy_of_Node5+Copy_of_binomial Copy_2_of_p_of_hit=Uniform(0,1) Copy_2_of_p_of_hit Copy_2_of_binomial=Binomial(7,Copy_2_of_p_of_hit) Copy_2_of_binomial 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 p of hit 118 199 167 229 Observed hits 118 479 172 512 Example A: the node "observed hits" is abinomial function which accepts evidence (0 to 7), the node "p of hit" (unifrom function) returns the most probable probabilities of hit given the number of hits observed. 37 558 224 642 p of hit 494 200 543 230 Observed hits 766 239 820 272 Example B: Everything similar to Example A, but I want to progressively offer evidence to the binomial node. The seven observers can "hit" or "miss", they are added up in the node "number of hits, that is the output of the binomial node... therefore by setting all the observers to miss, the "number of hits" node equals 0 (it works... so far). But its value is not used as evidence for upgrading the binomial "observed hits", and the "p of hit" 927 151 1122 319 Observer 6 1020 454 1068 484 Observer 7 969 377 1017 407 Observer 5 1013 539 1061 569 Observer 4 861 577 909 607 Observer 1 500 449 548 479 Observr 2 494 533 542 563 Observer 3 676 578 724 608 Number of hits 727 423 843 453 Copy of p of hit 399 826 448 856 Copy of Observed hits 767 783 821 816 Observation 1060 795 1131 839