GEDMATCH Generational Matrix

So interestingly I was shown a report from GEDMATCH the other day by a distant cousin. Looking at it I saw some very telling details that I had not considered before.

The report is a generational matrix. Think of it like a mileage chart for cities you used to find in AAA books. The columns and rows are GEDMATCH members instead of cities. The boxes are filled with the estimated generations to a most recent common ancestor rather than the number of miles. For example my sister says 1. My first cousin says 2. My first cousin, once removed says 2.3. The more distant the relationship, the higher the number. 

What this particular chart showed me was that there are some people who I am related to through different routes some more closely than others. For example, I am related to a person named NEF59. GEDMATCH estimates our distance to be 4.7 generations. Now the distance to my cousin Jennie, who is related though their mothers to my father, is only 2.3 generations. She is also at my dads generation so her distance to our shared ancestors should be less than mine by around 1 point. This works out pretty well except for a few examples like NEF. GEDMATCH thinks that Jennie and NEF are actually at a distance of 7.5 generations. That is over 4 generations different than what I would expect.  Since NEF is reported at 4.7 for me, it should be 3.7 for Jennie.  I am thinking this means that NEF actually has someone in their tree that matches me more closely than whoever NEF is matching in Jennie's tree. It also means that just because Jennie and NEF match, I cannot assume all of NEF's segments are coming from my paternal side through Jennie and my father's mothers. It seems likely that I am actually more closely related to NEF through a completely different ancestor that is not related to Jenny at all. It is even possible that Jenny is related to NEF through her father's line and I am not related to the person that Jenny is related to either. 

I have now decided to go through and look at key individuals and do a chart for each of them to determine exactly what the generational projection is.  My matches with Vicki, the cousin with most recent common ancestors who are both brick walls GUIDER ROBERTSON/ROBINSON turned up nothing really unusual, however, it did reveal a couple of people who were cousins themselves. This helped in narrowing their tree so I could work with them and know that we were talking about her mother's mother's family rather than having to research her entire tree looking for leads. 

Bradley County Swaffords / Lee's

So I am going through all my matches again.

This time I notice another LEE family in Bradley County. This time the same city as the Millers. It wouldn't even need to be a travelling minister.

I wonder where the big breakthrough is going to come?

Anyway, this Lee family is also descended of the Swafford Lee family. This guy's name is Burrell Lee as well, though on this census record it is spelled Lea.

Name:	Burrell Lee
Age:	38
Birth Year:	abt 1822
Gender:	Male
Birth Place:	Tennessee
Home in 1860:	District 10, Bradley, Tennessee
Post Office:	Cleveland
Family Number:	984
Value of real estate:	View image
Household Members:	Name Age Burrell Lee 38 Sarah E Lee 28 Robert A Lee 3 Manerva J Lee 2 Nevin Owens 17

And my Miller family

Name:	Mary A McMiller [Mary A Miller]
Age:	25
Birth Year:	abt 1835
Gender:	Female
Birth Place:	Tennessee
Home in 1860:	District 10, Bradley, Tennessee
Post Office:	Cleveland
Family Number:	938
Value of real estate:	View image
Household Members:	Name Age John McMiller 26 Mary A McMiller 25 Wm H McMiller 7 John S McMiller 5 Julius M McMiller 3 Mary McMiller 1

It makes me wonder if some of the researchers at Ancestry might have linked the wrong Burrell Lee to their tree....

Sites that bring researchers together

Ancestry is a great site for hints, online searches and documentation. However, as you will find out if you do much research there, there are a lot of newbies out there. And, as I was guilty of when I first started working on my tree, I would click and add without giving it much thought. Later, those clicks and adds would get copied to someone else's tree and then in effect be given much more weight than either of us had intended.

On Ancestry it is not uncommon to find glaring mistakes of parents being 90 years older than their kids. I have found census records of african americans with the same name linked to white folks. All in all,  the trees there are not dependable.

On the other side of the spectrum are Wikitree, Findagrave and Family Search. All three are great places to work with others on your trees. The least popular so far is Wikitree. It is a cumbersome interface and they have a lot of rules about posting information on the site. However, what all three sites share, is that there should be only one record for each person in history. Findagrave uses as its source point, the burial record or headstone.  This generally gives one a great meeting point. Findagrave can attract descendants from all over, looking at the site. There is one grave, hence one record. I have yet to run across someone posting incorrect father's mother's siblings etc.. on this site though I guess it could happen. I have found a couple duplicates for information, but all in all, I have used it to find some excellent researchers who are managing pages for my ancestors and have gotten some great history from it.

Familysearch is another awesome site. In the past year or so they have instituted a system where you can actually tag a name in a census document and attach it to your family tree. If someone later comes along and tries to tag that same record, the person in your tree will be suggested. This is great because theoretically on each census there should be one and only one record for each individual. When there is a conflict you can discuss right there on the record who's research is better.  Once you find your ancestor and you clean everything on their page up to your liking, you can also tag their record. Anytime anyone visits or makes changes to it, you will get a notification. That way, you  will see whatever additional information they have added. Also if you believe that information to be false, you will have a place to note that, change it back and/or contact the person that made the change.

These are all built-in features that work much better than Ancestry's limited tracking system.

Martinucci DNA saga

So some excellent results from my brother Todd's DNA test. He matches someone with ancestors from a town called Villa De Chiavenna, Italy. I had already identified this town as a possible origination point for my Martin(ucci) surname.

I also revisited an older site http://photohistory-sussex.co.uk/BTN-Lombardi.htm which points to a photographer named Eugenio Lombardi Martinucci (1849-1920). He is also for Chiavenna and bears a striking resemblance to my father.

I contacted the owner of that site, and he said he would try and reach the Martinucci descendants for me.

I was able to contact the Maraffio from Chiavenna.  He did not know much about genealogy, but his sister in law is a Mormon and is supposedly working on the family tree. I am also trying to conatct her.

I also tried running a search on my matches for people born in Chiavenna, and I was able to pull up one family from there Panzeri. This match was also a match to my cousin, who is descended from my Italian Grandfather. This is a good lead.

It sounds like a trip to Chiavenna is in order in the coming years. It is nearby to Andermatt and my mother's grandmother's home so I can do research there as well on the REGLI family.

This adds more surnames to my search including PANZERI, GINI, NYE, DOLLIMORE,  and TAYLOR, MARAFFIO,

I have been busy entering this information into familysearch and wikitree.org and I am actually going to write a little about those sites and findagrave right now.

Searching for the Swaffords

 So on my DNA test when I look at my list for Swafford relatives, the first one on the list is Jameswoolsey49. James E. Woolsey runs the account but the test is from a woman though, and with a little googling I came up with his wife, Sandra (or Samdra) Carolyn Fisher. Now Sandra is great grand daughter to Ida Mae Swafford. Her parents were Jay Fisher and Laverne Priest.  Nobody has found any records for Ida Mae except the marriage record, but I was able to find and Idea M Swofford 4 years before her marriage, in the same county Arkansas. I am 99% sure this would be the same woman.

Following the line back, even though the grandparent was born in Tennessee, I cannot connect them to the Swaffords and Lees I am looking for. I did notice a HOWARD surname as well.

I have tried to contact the Woolsey family with no luck.

UPDATE:

I did do some more work and traced these Swafford's back to our most recent common ancestor. This ancestor makes me and Sandra Fisher 7th cousins.  Ancestry tags Sandra as an EXTREMELY HIGH match at 4th-6th cousins.  Ancestry's description of the relationship is 4th cousin but could possibly vary to 5th or sixth cousin.

This means that either Ancestry is doing something very wrong in their analysis since they are citing this with their most definitive match category (the same one used for my sister), or we are somehow related more closely.

Unfortunately, until I can get the Woolsey family to upload the DNA results to GEDCOM I will not be able to look into exactly how we match. Perhaps we share more than one relative. Who knows?