[Taxacom] Molecular shared derived characters (was sine's line's)

[Jason Mate]

I believe that the statement was regarding how one chooses an outgroup: 
''the outgroup is not
informative unless one has a basis for predicting that the outgroup
condition is primitive''In any case I am not sure if you are referring to the outgroup per se or the idealised common ancestor.

On the other hand the selection of characters as described by your stated prediction method:

> By basis one might have a theory or principle by which the former state
> of a base is determined (predicted). But then I am not trying to make
> such a prediction myself so I don't know what basis one would use to do
> so.

It appears to be a guesstimate rooted on prior experience/data (perfectly OK): i.e. based on what I have seen in A, B, and C there is a pretty good chance that D will be like such and such. Models of base substitution work on a probabilistic basis that is rooted on empirical data, so they are comparable. The cladistic analysis of data is essentially the same process regardless of what is being analysed. Of course I may not have understood correctly.

Best

Jason