Author Topic: cM anomaly between Ancestry and Gedmatch - same kits??  (Read 634 times)

Offline Gadget

  • RootsChat Marquessate
  • *******
  • Posts: 57,138
    • View Profile
cM anomaly between Ancestry and Gedmatch - same kits??
« on: Wednesday 05 September 18 11:29 BST (UK) »
I have a match with a distant relative on both Ancestry and Gedmatch. My test was initially an Ancestry one which I later uploaded to Gedmatch. The match appears  also to have done an Ancestry test but much earlier, if the numeric is correct.

On Ancestry we have a 4-6 cousin match of 24cMs across 3 segments.
On gedmatch it's 4.0 gnerations to MRCA with 53 cMs across 3 segments (largest 37 cMs)

Could the type of Ancestry test have changed over the years and caused the discrepancy?

Can anyone give any other explanation, please?


Gadget


Census &  BMD information Crown Copyright www.nationalarchives.gov.uk and GROS - www.scotlandspeople.gov.uk

***Restorers - Please do not use my restores without my permission. Thanks***

Online Nanna52

  • RootsChat Veteran
  • *****
  • Posts: 924
  • Edwin WB Vincent, my actor, (1881-1940)
    • View Profile
Re: cM anomaly between Ancestry and Gedmatch - same kits??
« Reply #1 on: Wednesday 05 September 18 12:57 BST (UK) »
I don't think it is changes to Ancestrys test Gadget as my son and I did ours in May.  On Gedmatch his match to me is about 100 cMs more than on Ancestry.  I think it must be different ways of measuring or counting.
James -Victoria, Australia originally from Keynsham, Somerset.
Janes - Keynsham and Bristol area.
Heale/Hale - Keynsham, Somerset
Vincent - Illogan/Redruth, Cornwall.  Moved to Sculcoates, Yorkshire; Grass Valley, California; Timaru, New Zealand and Victoria, Australia.
Williams somewhere in Wales - he kept moving
Ellis - Anglesey

Gedmatch A327531

Offline Gadget

  • RootsChat Marquessate
  • *******
  • Posts: 57,138
    • View Profile
Re: cM anomaly between Ancestry and Gedmatch - same kits??
« Reply #2 on: Wednesday 05 September 18 13:21 BST (UK) »
Interesting, Nana52.

Thanks for the info.

I don't have such a discrepancy with other joint matches though  :-\


Gadget
Census &  BMD information Crown Copyright www.nationalarchives.gov.uk and GROS - www.scotlandspeople.gov.uk

***Restorers - Please do not use my restores without my permission. Thanks***

Online Nanna52

  • RootsChat Veteran
  • *****
  • Posts: 924
  • Edwin WB Vincent, my actor, (1881-1940)
    • View Profile
Re: cM anomaly between Ancestry and Gedmatch - same kits??
« Reply #3 on: Wednesday 05 September 18 13:43 BST (UK) »
I just looked at four other matches I have on both Ancestry and Gedmatch and they are all slightly higher on Gedmatch, between ten and thirty.  I think the closer the match the higher the difference.

I read somewhere that Gedmatch counts every little segment, but ancestry stop at a certain number.
James -Victoria, Australia originally from Keynsham, Somerset.
Janes - Keynsham and Bristol area.
Heale/Hale - Keynsham, Somerset
Vincent - Illogan/Redruth, Cornwall.  Moved to Sculcoates, Yorkshire; Grass Valley, California; Timaru, New Zealand and Victoria, Australia.
Williams somewhere in Wales - he kept moving
Ellis - Anglesey

Gedmatch A327531


Offline Gadget

  • RootsChat Marquessate
  • *******
  • Posts: 57,138
    • View Profile
Re: cM anomaly between Ancestry and Gedmatch - same kits??
« Reply #4 on: Wednesday 05 September 18 13:49 BST (UK) »
That's why I've looked at the largest segment and how many segments. I read somewhere that it's length of segment that might be most important. Can't remember where though as I've read so many bits and pieces  ;D

Another one:

Gedmatch - 34.2  1 seg, 4.4 gens
Ancestry - 37 over 2,  4-6

My closest matches are all on Ancestry or My Heritage.
Census &  BMD information Crown Copyright www.nationalarchives.gov.uk and GROS - www.scotlandspeople.gov.uk

***Restorers - Please do not use my restores without my permission. Thanks***

Offline Kimbrey

  • RootsChat Veteran
  • *****
  • Posts: 864
  • Census information Crown Copyright, from www.nationalarchives.gov.uk
    • View Profile
Re: cM anomaly between Ancestry and Gedmatch - same kits??
« Reply #5 on: Wednesday 05 September 18 15:30 BST (UK) »
Yes,Ancestry changed the alogrithm's they use in May 2016,the raw data from tests before then can be downloaded to FTDNA, but those done since then will not transfer without  the use of a "fix" worked out by an FTDNA Project admin which he has shared.

From comments on the FTDNA Forums it is thought that the total number of matches now are lower.

Gedmatch is not affected as their system is designed to accomodate several different companies .

Kim

Offline Gadget

  • RootsChat Marquessate
  • *******
  • Posts: 57,138
    • View Profile
Re: cM anomaly between Ancestry and Gedmatch - same kits??
« Reply #6 on: Wednesday 05 September 18 16:30 BST (UK) »


I used the fix before transferring the Ancestry data to the other sites and they uploaded with no problem.

It's really the Gedmatch and Ancestry variations that I'm interested in as I have corresponding matches. 




Census &  BMD information Crown Copyright www.nationalarchives.gov.uk and GROS - www.scotlandspeople.gov.uk

***Restorers - Please do not use my restores without my permission. Thanks***