Trying to split match list into parents groups.

[Essnell]

Hi Roots Chatters,   
 I am working on a particular match list .
 I believe it should connect to both parents -  but If those parents were siblings, would the match list contain only matches that lead back to their Parents and Grandparents.

I am trying to rule out this particular scenario. 

I have lots of matches that lead to one set of Grandparents.  There are lots of matches that do not appear to have any real connection to anywhere /  or limited connection to just a few others. 

I am about to try on Ancestry for this cluster option  may be??? 
Any ideas please.   

Essnell

[aghadowey]

Not sure if this helps but have you considered the possibility of intermarriage between two sets of siblings?
Some examples-
My great-grandfather and his brother married two sisters but the brother had no children.
Another great-grandfather and his sister married a brother and sister. Each had 7 children and my matches with descendants appear closer than they actually are (instead of being my 4th cousins they show up as possible 2nd cousins once removed, etc.)
Husband's grandfather and twin brother married two sisters which again accounts for more shared DNA.

[Essnell]

Hi Aghadowey,
Thanks for the reply.  No I had not considered that scenario as yet.  Problem is I am seeing who I think is the female of  this couple and that is where the most matches point . Also many are from the same small town in a quite distant area of where this is supposed to have occurred. 

On the other hand I am not getting anything like a cohesive group of families that would be on the male side . It's just a hotch-pot. So because of that I am not sure that my thinking along the question lines is right.

I reckon we need another person closer related to test.    UGH!

 I shall keep searching - there has to be someone somewhere. 
Thank you for your help. 
Essnell.

[Biggles50]

Have you worked through your known DNA matches and used the Colour Coding Groups feature.

If not then I suggest that doing so will probably help you far more than Clustering which at present is hardly out of Beta and what is available is limited in what it uses out of the DNA Match Pool.

The more appropriate DNA testers that you can include, the better.

[Essnell]

Hi Biggles, 
Short answer: Yes.     I have color coded as many as possible but I do not have 64 colors only 16 so it's getting a bit muddled.  I am seeing parts where some matches appear to be in two of those groups but can't see the links.  Others seem to go off on a tangent ..leading to another possible connection. It's just not jelling. 

Anyway I shall try putting all on a spread sheet.  Looking for a second set of Gt Gt Grandparents at minimum. 

Thank you for your reply - always helpful. 
Essnell

[Biggles50]

If you are using Coloured Groups for different systems than for 16 GGP’s then you could just use 8 of them, one for each Grandparent.

Including Shared Matches in the same Group.

It is those that are left unallocated that may reveal more.

[Essnell]

HI Biggles, Thank you for that.  I will try to get some of that done. It will take sometime, so eventually I shall be back with some sort of find. 

Essnell.