Usually the lowest energy conformation of the largest cluster should be your docking solution. Generally it rarely happens. For this reason, if you have the crystal structure of your receptor co-crystallized with a ligand, you should try to redock the ligand in the receptor in order to check if the docked position is similar to the crystallographic one (rmsd < 2). If you're lucky the nearest docked conformation to the crystallographic one will correspond to the lowest enery convormation of the largest cluster.

You can perform control docking with the other default parameters in autodock. It is one of the best docking software. To visualize the docked component you should emphasize on hydrogen-bond beetween ligand and receptor if it exist. For the further analysis I would recommend Gromacs.

You must have knowledge over active site of your protein. When you're docking one ligand with crystallography structure solved, you can compare RMSD between it. However if you are docking a new compound, you should pay attention on active site pocket (or protein surface if site be in it), the patter of residues and type of interaction between them and ligand. Is a question of good sense and knowledge almost always!!!

you should know about molecular interaction in the complex (docking result) and the criteria for the best docking result is the free energy binding. for further analysis of docking result, you can do molecular dynamic simulation to know the stability of interaction (ligand-receptor). the tools ( AMBER, GROMACS, etc.).

The autodock python viewer is itself a visualization program. For the best result you have to pay attention about minimum energy and large number of interactions.