Computer simulations of interacting galaxies are used to determine the contribution of halos in disk galaxies. It appears that generally the halo is less massive than the disk. Redshift asymmetries are used to separate binary pairs from optical pairs in catalogs of galaxy pairs. It is found that true binary pairs give a rather low mass for the components. Redshift asymmetries also indicate that groups of galaxies are ill-suited for mass determination, either because they are in an unknown state of expansion or because they are not gravitationally bound at all. It is argued that so-called cluster missing mass cannot exist for dynamical reasons, and the high-velocity dispersion of member galaxies is more likely to indicate incomplete virialization than a high mass. The commonly quoted density parameter value Omega equivalent to 0.2 appears to be definitely too high, at least in scales less than 10 Mpc. >