Using this matrix as an example, the post deals with commonly ignored problem that palaeophylogenetic matrices provide non-treelike, often ambiguous, partly conflicting signals. Cladograms, i.e. phylogenetic trees without branch-lengths (and here: branch support), cannot capture the strengths and weaknesses of the used data. The polytomies may be soft (e.g. one jumping taxon) or semi-hard (no differentiating signal). Networks, 2-dimensional and based on pairwise distances, or multidimensional, based on tree samples, can. Even if the general preference is a tree, one should always show a phylogram (or several alternative phylograms), in which the branch lengths indicate the (inferred) change.