The single-bit method can actually be seen as an instance of the huffman-coded method, where the match lengths are all assigned equal entropy estimates and the

The single-bit method can actually be seen as an instance of the huffman-coded method, where the match lengths are all assigned equal entropy estimates and the probability of "match vector" versus "literal character" is hard-wired to 0.5.