Saturday, May 3, 2014

What's Your Ancestor Score and How Does it Relate to Your DNA Test?

I usually judge the quality of any genealogy by how well it is sourced and researched, and not the number of people in a given genealogy. It is not about the number of people in a genealogy, but the quality of the research, what is known about those family members, and how well the genealogy has been documented.

Yesterday, I came across an article by Randy Seaver that had an interesting methodology I thought I would try. And when I did the little exercise a little light bulb went on. I thought about it in the context of the autosomal DNA test I took at two years ago. The author of this exercise didn't have what I have used it for in mind when he wrote about it at Geneanet, but I have discovered another use for it when you think about it in terms of your AncestryDNA autosomal match results.

First, here is what Randy Seaver posted up at the Geneanet Blog on 28 April (What's Your Ancestor Score?)

"How to know how complete your family tree is?  This question may sound strange. What should we talk about? Quantity or quality?  Randy Seaver gives us a simple solution in a blog post entitled "What's Your Ancestor Score?".

"Genealogy is a long term hobby and we need to collect as many sources as we can to build our family tree. The following is not an encouragement to accept mistakes and poor quality work but, from time to time, it can be interesting to evaluate the job already done and to motivate ourselves.

"How to:  This is simple and efficient.  Compare the number of possible ancestors with the  number of identified ancestors on a 10-generation report.  The percentage is your Ancestor Score.  Here is an example with my own family tree."

The reason why I find this interesting is because the AncestryDNA test results presented to you on your matches page actually looks at up to the first ten ancestor generations when it compares your tree with the tree of your autosomal DNA match.

So I've added a bit more information to Randy's original table so that I can get a better feel generation by generation of where my tree actually stands through each of the 10-generations that would be compared to my matches 10-gens if they have them.

As I have pointed out in my DNA classes, you may have a lot of quality matches that do not show up as a leaf at Ancestry because either your tree or your matches tree has not been built out fully to at least the 10-generation level. Think about all those "low" and "very low" confidence matches you have. Those lower quality matches will normally equate to more distant common ancestors with your DNA matches and they are further back in time on your family tree.

Another way to think about this is do you know the names and have information on all 1023 of your ancestors that would appear in your 10-generation tree? Does your DNA match at Ancestry or FTDNA know all about their 1023 ancestors in their 10-gen chart?

As you can see from my chart below I do not know who over 25% of my ancestors are starting at generation seven. By the time I get back to gen nine, I do not know over half my ancestors that I would have at that generation level. I only know 320 of the 1023 or 31.2% of the ancestors that I would have at 10-generations.

Larry's 10-Generation Ancestor Numbers

Generation  Possible     Identified  Percentage  Total          Total          Total
                    Ancestors  Ancestors                     Ancestors   Identified  Percentage
1                  1                 1               100%          1                 1                100%   
2                  2                 2               100%          3                 3                100%
3                  4                 4               100%          7                 7                100%
4                  8                 8               100%          15               15              100%
5                  16               16             100%          31               31              100%
6                  32               24             75.0%         63               55              87.3%
7                  64               37             57.8%         127             92              72.4%
8                  128             53             41.4%         255             145            56.8%
9                  256             76             29.6%         511             221            43.2%
10                512             99             19.3%         1023           320            31.2%

I have added Gayle's info and it is comparable to what I found in my chart above.

Gayle's 10-Generation Ancestor Numbers
Generation  Possible     Identified  Percentage  Total          Total          Total
                    Ancestors  Ancestors                     Ancestors   Identified  Percentage
 1                  1                 1               100%          1                 1                100%   
 2                  2                 2               100%          3                 3                100%
 3                  4                 4               100%          7                 7                100%
 4                  8                 8               100%          15               15              100%
 5                  16               16             100%          31               31              100%
 6                  32               26             81.2%         63               57              90.4%
 7                  64               41             64.0%         127             98              77.1%
 8                  128             61             47.6%         255             159            62.3%
 9                  256             54             21.0%         511             213            41.6%
10                 512             45               8.7%         1023           258            25.2%

Add into the mix what my autosomal match at Ancestry, FTDNA or 23andMe may or may not know about their tree, and now you see why many matches (the names of the "Most Recent Common Ancestors" or MRCAs we share with our autosomal matches) remain a mystery to us.

Quite a few of my fellow genealogists that I have talked to equate the lack of matches in their test to flawed DNA tests, lack of cousins testing, or just the whole thing being bunk science, etc, and consider DNA testing to be a waste of time.

Really? Have you really sat down and taken a hard look at your ancestral lines in your genealogy? Are you convinced that your paper trail is accurate and complete in every respect? Are you absolutely sure that there is not a "not parent expected" (NPE) event in your tree at some point? Remember my old saying in class, "mom's baby, daddy maybe." Do you have all 1023 ancestors identified in your 10-generation ancestor chart?

I would submit that the more you dig, the more useful that DNA test you took really is. I like the thought that an autosomal DNA test can help me prove or possibly disprove a lineage paper trail.

I look at my test, first as a genealogical record that has no equal anywhere else in the my mountains of research that I have down over the last 35 plus years. I use it not only to verify my paper trails, but as a research aide to open up possible new lines of ancestor research. Everyday as I spend more time working with my results I discover important new ways to utilize the pages of DNA matches I have at Ancestry, FTDNA and GEDMatch.

As I have said many times over the last semester in my classes at Tri-County Community College, build out that ancestor tree as accurately as you can and also build it back towards you as well. That will go a long way to helping you spot potential matches that may not be obvious between your tree and your DNA match's tree.

Bottom line: That $99 test at Ancestry is the best money I have ever spent in my genealogy research and glad that I tested not only me, but my parents as well. IMHO my AncestryDNA test has been the most important genealogical tool I have used in my 35 years of genealogical research. As V. Utt said during DNA Day 2012 in Estonia, "The best book that has ever been written is in us and nowadays we have (the) opportunity to read it."

Have you read your book yet? Do I have your attention now? Go create your genealogical record today by ordering an autosomal DNA test.

DNA Cousin cM levels for relationships
Cousin Rel     Ave cM     Relationship                     Generation     Total AncestorsYou                                                                              1st Gen
Sibling           2350 cM     Parents                               2nd Gen       2
Half Sibling   1600 cM     Parents                               2nd Gen       2
Parents           3580-
                       3585 cM     Parents                               2nd Gen      2 (2)
First Cousin     800 cM     Grand Parents                     3rd Gen      4 (6)
Second Cousin 200 cM    1st Great Grand Parents      4th Gen      8 (14)
Third Cousin    100 cM   2nd Great Grand Parents      5th Gen    16 (30)
Fourth Cousin    50 cM    3rd Great Grand Parents      6th Gen    32 (62)
Fifth Cousin       25 cM    4th Great Grand Parents      7th Gen    64 (126)
Sixth Cousin      12.5 cM  5th Great Grand Parents     8th Gen   128 (254)
Seventh Cousin   6.25 cM  6th Great Grand Parents    9th Gen   256 (510)
Eight Cousin       3.125 cM 7th Great Grand Parents  10th Gen  512 (1022)