This page serves supplemental information on my paper, "Relevance of Hydrogen Bond Definition in Liquid Water", *M. Matsumoto, J. Chem. Phys. 126, 054503 (2007). DOI:10.1063/1.2431168, in which I discussed why water is regarded as a network material while other (simple) materials are not, and proposed a consecutive way of tuning the thresholds.
In this section, error magnitude is evaluated for new hydrogen bond definitions proposed after my paper was published.
Their definition is simple. Ambiguity is eliminated because there is no control parameter (threshold) for determining the bonds. Error magnitude is small, but is not minimal compared with the simple criterion by intermolecular OH distance.