Hmmm. Random interactions among otherwise-identical components
also result in this pattern. It's somewhat disquieting that a model
with so little information gives decent results.
> > ... They found the pattern in spots but nowhere as
> > consistently as with Internet stocks.
It's nice to know that other industries actually require a bit of
garbage in to derive rankings. It'd be even more reassuring if that
extra information had much to do with the nature of the businesses.