This is starting to get real old, real quick. I understand that science moves forward and as we continue to research things will change. Lord only knows I have lived through three versions of AncestryDNA updates and a couple at FTDNA. But Ancestry should be ashamed of this NADs fiasco. To unleash the barrage of NADs on the heels of the v3 update without truly testing how the v3 would effect those NADs is nothing short of irresponsible.
How can I say that you ask?
I am one of those lucky folks to have tested both parents so in theory I should benefit from Ancestry phasing our autoomal DNA test. I'm not uncomfortable at this point with the v3 update although I still have high confidence false positives.
But those NADs I was hanging around five NADs before the update. When the NAD update hit that number jumped to 17. But the rub was only nine of them were shared by my parents. Eight of them were false postives. I immediately sent a support message to Ancestry. and now the big update again, I'm down to seven and my parents have all those NADs.
So here is the fallout of the bouncing NADs - pre v3 update - v3 update - current:
My test 5-17-7
My father 10-35-8
My mother 4-14-5
My wife 17-28-15
Wife's 2nd cousin 18-33-18
My 2nd cousin 9-14-4
My original article on the false NADs posted on my family blog at http://family-genealogy.blogspot.com/2016/05/nads-and-nads-and-yes-even-more-nads.html
Let's hope this is the last update for awhile at Ancestry, but I fear we have at least one more coming (that is the rumor I am hearing).
Come on folks I need a chance to catch up my documentation and I am tired of adding columns to my spreadsheet documenting your ups and downs. Please give this a rest for a while.
Here is Ancestry's response to my Facebook post on their FB page:
"Hi Larry, we apologize for any frustration you may have experienced with the New Ancestor Discoveries. In order to determine New Ancestor Discoveries, we created an algorithm with criteria that connects people to DNA Circles based on their DNA matches. This algorithm was created last year when we launched New Ancestor Discoveries and with the rapid growth of the DNA database, we are finding it needs to be updated.
"As DNA Circles get larger and more DNA matches are delivered, more people are connecting into the DNA Circles, which results in more New Ancestor Discoveries, but with a decrease in accuracy. So, we are updating the criteria to make it more conservative and increase the accuracy of New Ancestor Discoveries. This means you’ll need more connections into a DNA Circle to get a New Ancestor Discovery.
"These updates will result in a significant decrease in the number New Ancestor Discoveries, but with an increase in accuracy. We will continue to monitor and adjust this as necessary to ensure these provide meaningful discoveries for our members."
And then there is this explanation from Anna Swayne of Ancestry DNA:
"Previously, you needed to match at least 2 members of a known DNA Circle to be given a New Ancestor Discovery. Now, users must match at least 3 members of a small (15 members or less) DNA Circle to be given a New Ancestor Discovery. For larger DNA Circles (16+ members), users must match 20% of that Circle to be given a New Ancestor Discovery. For example, if there is a DNA Circle of 10 people, you will need to match at least 3 people to get a New Ancestor Discovery. And if there is a DNA Circle of 30 people you will now need to match 6 people instead of 2."
Welcome to the Family Roots and Branches Genealogy Blogspot. This site is devoted genealogical hints, tricks, tips and news for family historians. You will also find info on genealogy classes and meetings here in western North Carolina. This blog is run by Larry and Gayle Van Horn and you can reach us via email at familyhistorian at frontier dot com. Copyright 2006-2023 by Family Roots and Branches, a division of Teak Publishing.
Showing posts with label v3 update. Show all posts
Showing posts with label v3 update. Show all posts
Thursday, June 2, 2016
Sunday, May 22, 2016
NADs and NADs and yes, even more NADs! But some of them aren't real?
AncestryDNA has gone nuts with NADs. The sad part is they just can't be trusted for anything useful in some cases aka false NAD matches.
So Larry you ask how do you know this outrageous fact. Hint here dear reader, I've tested both my parents and if they do not have those NADs that I have they can't be real right? Remember that Ancestry made a big deal about the fact that those of us who also tested our parents have had our results phased with their test to produce more accurate results and get rid of false positives - right!
So let me illustrate dear reader. Here are the current crop of NADs (they put a whole bunch on the street earlier this week).
Larry Van Horn 17 NADs (the son)
Elisabeth Defeld (1819-1901)* Maternal
Virginie Devillier (1821-1878)
Jane Dunlap (1811-1873)* Paternal
Elinor Henson (1787-1873)* Paternal
Faustin Hollier (1812-1876)
Berryman Isom Jones (1820-1896)* Paternal
John Doyle Lee (1812-1877)
John Lynch (1778-1863)* Paternal
Warren Lafayette Lynch (1839-1916)
Jeremiah Meek (1806-1856)* Paternal
Jeremiah V Meek (1818-1893)
Johann Peter Meuth (1809-1857)
Margaretha Meuth (1843-1910)* Maternal
Frank Seidel (1827-1893)* Maternal
Elizabeth Vest (1820-1897)
Lueticia Ward (1845-1883)
Licenia Watkins (1824-1909)* Paternal
My Father 35 NADs
Henry C Begley (1792-1854)
Levi Brookshire (1830-1916)
John Laddie Bullock (1844-1926)
Peter Richard Clement (1840-1914)
Henry Heslip Davis (1849-1931)
Emma D. Dixon (1839-1909)
Jane Dunlap (1811-1873)
Alfred Dye (1842-1918)
Francis Kincannon Elliott (1813-1884)
James Millard Gilliland (1856-1939)
Laura Emma Hall (1860-1932)
Elinor Henson (1787-1873)
Adah Belle Higby (1853-1925)
Berryman Isom Jones (1820-1896)
Nicie Elizabeth Kinnaird (1859-1925)
Anne Friedericke Kock (1827-1894)
Reuben Layton (1836-1892)
Louisa Jane Long (1840-1933)
Eliza Clementine Lynch (Born 1873)
John Lynch (1778-1863)
Jonathan Brooks May (1857-1923)
Jessie Alexander Mayhall (1835-1912)
Lucy Jane McAleer (1856-1933)
Jeremiah Meek (1806-1856)
Mary Cordelia Meek (1847-1928)
Eliza Ann Plaster (1839-1915)
Elizabeth Roberts (1792-1837)
Mary Elizabeth Selvage (1859-1947)
Morgan Joseph Smith (1834-1911)
Johann Heinrich Staack (1817-1894)
Licenia Watkins (1824-1909)
Elisabeth "Lizzie" Yeilding (1852-1884)
My Mother 14 NADs
Elisabeth Defeld (1819-1901)
Ruben Theodore "Fred" Farthing (1854-1932)
Faithia Futrell (1808-1887)
Green Berry McCormick (1848-1930)
Johann Peter Meuth (1809-1857)
Margaretha Meuth (1843-1910)
Levi William Pitts (1838-1880)
Martha Jane Prater (1837-1930)
Frank Seidel (1827-1893)
Rebecca Singleton (1850-1912)
James Storm (1786-1863)
Telitha Surginer (1842-1912)
James Vaughn (1804-1857)
James Vinson (1804-1884)
And the results? Eight of the NADs I have nearly 50% are not shared with either of my parents.
Not good for someone's test that has been supposedly phased with his parents AncestryDNA tests.
BTW I am also working up a list of false matches that I still have even after AncestryDNA's new phasing algorithm. That isn't looking so good either.
AncestryDNA is not having a good week in my eyes for sure.
Saturday, May 7, 2016
AncestryDNA's updated matching algorithms - Some before and after analysis
Found this interesting article on the Cruwys news blog. It has some analysis of their results from the AncestryDNA algorithm update at
http://cruwys.blogspot.co.uk/2016/05/ancestrydnas-updated-matching.html
I am about half way through updating my personal test results after the v3 update. I can definitely see some change and some improvement in IBD matching. I am one of those who probably benefited the most with this update thanks to have tested both my parents and the implementation of their new phasing algorithm instituted with this update.
Having said that I still have false matches including a close in 4th cousin, good confidence, 29.5 cM/2 segment match that neither of my parents has.
Right now I have two support tickets into AncestryDNA Tech Support for matching issues that defy all logic. Sometimes that computer most be drinking heavily! ;-)
In case you haven't seen the new matching criteria, here is what Ancestry has posted in their help files for this v3 update.
And here are the shared cM levels for various relationships now.
I will be doing some additional updates here on the FRB Blog as I complete the updates for my testing. But in the meantime, here are some initial statistics before and after AncestryDNA v3 application.
Larry Van Horn (this is me)
Before Update: 82 Leaf Matches, 124 4th Cousins, 18 DNA Circles, 5 NADs, 111 Pages
After Update: 93 Leaf matches (+11), 146 4th Cousins (+22), 16 DNA Circles (-2), 5 NADs, 138 Pages
My Father Warner Lee
Before Update: 138 Leaf Matches, 213 4th Cousins, 10 DNA Circles, 10 NADs, 160 Pages
After Update: 128 Leaf Matches (-10), 220 4th Cousins (+7), 12 DNA Circles (-2), 10 NADs, 170 Pages
My Mother Gloria Ann
Before Update: 24 Leaf Matches, 165 4th Cousins, 8 DNA Circles, 4 NADs, 107 Pages
After Update: 22 Leaf Matches (-2), 201 4th Cousins (+36), 4 DNA Circles (-4), 4 NADS, 134 Pages
My 2nd Cousin Jerilyn
Before Update: 76 Leaf Matches, 249 4th Cousins, 4 DNA Circles, 9 NADs, 205 Pages
After Update: 83 Leaf Matches (+7), 373 4th Cousins (+124), 4 DNA Circles (n/c), 9 NADs, 252 Pages
My Wife Gayle
Before Update: 269 Leaf Matches, 512 4th Cousins, 24 DNA Circles, 17 NADs, 217 Pages
After Update: 280 Leaf Matches (+11), 549 4th Cousins (+37), 21 DNA Circles (-3), 17 NADS, 238 Pages
Gayle's 2nd Cousin Bryan
Before Update: 166 Leaf Matches, 598 4th Cousins, 14 DNA Circles, 18 NADs, 232 Pages
After Update: 165 Leaf Matches (-1), 621 4th Cousins (+23), 15 DNA Circles (+1), 18 NADS, 258 Pages
Some of these numbers continue to fluctuate. Two days ago I watch the number of 4th cousins on my test page change several times during the day from 149 to 147 and back up to 149. It changed like this back and forth all day long. This morning it is back down to 147 for now.
I did lose five MRCA matches, but most were down there a bit in the weeds so to speak. Here are the casualties of the v3 update.
Moderate 5.1cM/1 8GGP Cornelius Dabney-Susannah Swann
Moderate 5.5cM/1 7GGF Heinrich Furrer-Russena Rosser
Moderate 5.7cM/1 7GGP John Chalfant-Dorothy Adams
Moderate 6cM/1 5GGM Margaret McCarter-1st husband William Motley
Good 8.8cM/1 7GGP William Hurt-Anne Stennard
Also if you have not done so be sure as soon as possible download your deleted v2 matches from Ancestry. Roberta Estes on her blog at https://dna-explained.com/2016/05/03/ancestry-update-downloading-v2-deleted-matches/ has all the instructions to perform this function so I won't repeat it here. Remember in order to have this v2 deleted file prior to the update if you starred or noted matches, and if those matches got deleted during the Ancestry update, Ancestry created that spreadsheet file for you to download. It’s located under your setting gear wheel, to the right of your name.
More to follow, the analysis continues . . .
http://cruwys.blogspot.co.uk/2016/05/ancestrydnas-updated-matching.html
I am about half way through updating my personal test results after the v3 update. I can definitely see some change and some improvement in IBD matching. I am one of those who probably benefited the most with this update thanks to have tested both my parents and the implementation of their new phasing algorithm instituted with this update.
Having said that I still have false matches including a close in 4th cousin, good confidence, 29.5 cM/2 segment match that neither of my parents has.
Right now I have two support tickets into AncestryDNA Tech Support for matching issues that defy all logic. Sometimes that computer most be drinking heavily! ;-)
In case you haven't seen the new matching criteria, here is what Ancestry has posted in their help files for this v3 update.
And here are the shared cM levels for various relationships now.
I will be doing some additional updates here on the FRB Blog as I complete the updates for my testing. But in the meantime, here are some initial statistics before and after AncestryDNA v3 application.
Larry Van Horn (this is me)
Before Update: 82 Leaf Matches, 124 4th Cousins, 18 DNA Circles, 5 NADs, 111 Pages
After Update: 93 Leaf matches (+11), 146 4th Cousins (+22), 16 DNA Circles (-2), 5 NADs, 138 Pages
My Father Warner Lee
Before Update: 138 Leaf Matches, 213 4th Cousins, 10 DNA Circles, 10 NADs, 160 Pages
After Update: 128 Leaf Matches (-10), 220 4th Cousins (+7), 12 DNA Circles (-2), 10 NADs, 170 Pages
My Mother Gloria Ann
Before Update: 24 Leaf Matches, 165 4th Cousins, 8 DNA Circles, 4 NADs, 107 Pages
After Update: 22 Leaf Matches (-2), 201 4th Cousins (+36), 4 DNA Circles (-4), 4 NADS, 134 Pages
My 2nd Cousin Jerilyn
Before Update: 76 Leaf Matches, 249 4th Cousins, 4 DNA Circles, 9 NADs, 205 Pages
After Update: 83 Leaf Matches (+7), 373 4th Cousins (+124), 4 DNA Circles (n/c), 9 NADs, 252 Pages
My Wife Gayle
Before Update: 269 Leaf Matches, 512 4th Cousins, 24 DNA Circles, 17 NADs, 217 Pages
After Update: 280 Leaf Matches (+11), 549 4th Cousins (+37), 21 DNA Circles (-3), 17 NADS, 238 Pages
Gayle's 2nd Cousin Bryan
Before Update: 166 Leaf Matches, 598 4th Cousins, 14 DNA Circles, 18 NADs, 232 Pages
After Update: 165 Leaf Matches (-1), 621 4th Cousins (+23), 15 DNA Circles (+1), 18 NADS, 258 Pages
Some of these numbers continue to fluctuate. Two days ago I watch the number of 4th cousins on my test page change several times during the day from 149 to 147 and back up to 149. It changed like this back and forth all day long. This morning it is back down to 147 for now.
I did lose five MRCA matches, but most were down there a bit in the weeds so to speak. Here are the casualties of the v3 update.
Moderate 5.1cM/1 8GGP Cornelius Dabney-Susannah Swann
Moderate 5.5cM/1 7GGF Heinrich Furrer-Russena Rosser
Moderate 5.7cM/1 7GGP John Chalfant-Dorothy Adams
Moderate 6cM/1 5GGM Margaret McCarter-1st husband William Motley
Good 8.8cM/1 7GGP William Hurt-Anne Stennard
Also if you have not done so be sure as soon as possible download your deleted v2 matches from Ancestry. Roberta Estes on her blog at https://dna-explained.com/2016/05/03/ancestry-update-downloading-v2-deleted-matches/ has all the instructions to perform this function so I won't repeat it here. Remember in order to have this v2 deleted file prior to the update if you starred or noted matches, and if those matches got deleted during the Ancestry update, Ancestry created that spreadsheet file for you to download. It’s located under your setting gear wheel, to the right of your name.
More to follow, the analysis continues . . .
Subscribe to:
Posts (Atom)