GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2007-10 > 1191286572


From: (John Chandler)
Subject: Re: [DNA] East Anglia and the Vikings
Date: Mon, 1 Oct 2007 20:56:12 -0400 (EDT)
References: <KHEKIJEABJGJEKDPFEDMEEGEEGAA.Elizabethod@eircom.net>
In-Reply-To: <KHEKIJEABJGJEKDPFEDMEEGEEGAA.Elizabethod@eircom.net>


Elizabeth wrote:
> I know you discount Family Tree's mutation rates altogether,

Not altogether. Just a factor of two (or less, depending on which
panel you look at).

> Also, I know you responded to me a bit ago mentioning a 500-1000 year scale
> for using your own rates with 'fudge factors'. Could you see them being
> stretched to 2,000 years?

The fudge factors are an attempt to account for the population history
when you are giving a purely statistical treatment instead of a purely
genealogical one. The problem is that the population history is, in
general, not known, and so you simply can't tell what factor to apply.
This is true for any time scale when you don't know the family tree
connecting the test subjects. Even 100 years would require fudging
if you have to guess at the relationships. Conversely, if you have
the tree, you don't need to fudge ever.

John Chandler

This thread: