Drinking Through a Fire Hose!

Posted on by

I have incorporated all of my AncestryDNA ThruLines and MyHeritage Theory of Family Relativity Matches into my Common Ancestor spreadsheet (see Chapter 7 of the free book, Segmentology Fundamentals, at ISOGG). Here is a tabulation of Matches with Common Ancestors (CAs) at all companies:

23andMe                 167

Ancestry           10,435

FTDNA                      239

GEDmatch              170  [I’ve not been able to find these at the other companies]

MyHeritage            261  [this includes 113 from Theory of Family Relativity]

Total                    11,272

Clearly AncestryDNA leads the pack; but note that at the other companies, all the Matches have known shared DNA segments in specific Triangulated Groups (TGs).

Here is a breakdown of Ancestry by category:

ThruLines         8,516  [includes 144 wrong but fixed;  plus 104 which are now gone]

No Tree                   175  [determined by Pro Tools]

Private Tree            27  [determined by Pro Tools]

Unlisted Tree      531  [Note the large number of Matches; not found by ThruLInes]

Found in Tree   1,186 [Just searching]

Total                    10,435

Of the above 10,435 Matches at Ancestry

  • There are 1,078 (roughly 10%) that I have tagged as incorrect – and I move those out of the active part of the spreadsheet
  • There are 741 Matches with known shared DNA segment in TGs [plus 328 additional Matches for whom I don’t know the CA]
  • There are 6,885 from 5xG grandparents or closer (nominal 6Cs) – this reflects the power of ThruLines to find them.
  • There are 2,077 Matches from 6xG grandparents; 1,232 Matches from 7xG grandparents; and 181 Matches from more distant Ancestors.

Note – some of the Matches are listed more than once when they are related to me multiple ways; some because they tested at multiple companies.

So what’s the point here?

#1 is that I’m drinking through a fire hose!  Granted that I’m retired and can spend time on genetic genealogy…

#2 is that the data is out there – or rather, the data is here, within the reach of the major DNA companies…

My morning routine includes seeing if there are any new ThruLines at Ancestry or any new Matches at GEDmatch (particularly Ancestry ones). Often I cannot get through that chore before I have to break for other responsibilities.

If I have time, my next task is working down my Ancestry Matches in my Common Ancestor spreadsheet and evaluating their shared Matches with Pro Tools. I didn’t have Pro Tools when I first developed the CA spreadsheet, so there is a lot of catching up to do.

It’s a two pronged approach – enter the Match in my Tree [tagging them: “DNA Match” AND using a special Dot in their profile when I do so] and entering their path to our CA [tagging each person: “DNA Connection”] in my Tree; and then evaluating the shared DNA Matches using Pro Tools [sorted on the Match’s relationship]. I’ve done this through my 4xG grandparents and now have 3,368 Tagged Matches. Still thousands to go, plus all of the new Matches with CAs I find with Pro Tools.  Drinking through a fire hose.

The point is that I’m building a large family of DNA-linked descendants for each Ancestor – easy to review in the CA spreadsheet and in my Tree. AND, as I find more and more Matches with segment information, the consensus builds for Chromosome Mapping info.

[22DK] Segment-ology: Drinking Through a Fire Hose; by Jim Bartlett 20260616

Free: Segmentology Fundamentals eBook available for download at ISOGG/Wiki

13 thoughts on “Drinking Through a Fire Hose!”

  1. Thanks Jim. Just a quick process question. For your daily thrulines routine, are you just filtering on Common Ancestor then sorting on New? Or is there more of a trick to it.

    Thanks

    Pat

    Liked by 1 person

    Reply ↓

    • Pat – I filter on Common Ancestor AND Unviewed Matches – this picks up New Matches as well as Old Matches which I haven’t looked at and were recently discovered by ThruLines. There is actually a high percentage of these Old Matches which ThruLines finds for a variety of reasons – not the least of which is my own Tree which broadens with each new ThruLines path I agree with and add to my Tree. And other genealogists are also adding new info that helps ThruLines. This method does not pickup Old Matches that I have viewed for some reason (maybe as a Match in a Cluster, but I just lost patience trying to find our CA, or many other reasons). So, about every six months, I filter just on ThruLines alone and scroll through my list (now over 1,000) checking to ensure that I’ve dotted each one – either with a CA Dot or a ThruLines WRONG Dot. I’m also sensitive to the fact that ThruLines does not look at Unlinked Trees or Private and unsearchable Trees, so I have to look for those among Clusters and Shared Matches lists using Pro Tools to find close relationships to *known* Matches.
      I appreciate your question, as we should all be sharing ways to “sharpen the saw” as Steven Covey used to say. Jim

      Like

      Reply ↓

    • Pat, Just after posting my reply to you, I used my double filter and got a new ThruLines: a Match who had been on Ancestry for over a year – 8cM/1Seg Tree with 5 people… TL showed a 5C1R with 4 of the Match’s Ancestors back to the CA already in my Tree – the grandfather and mother were easy to confirm and add… I can’t help but think that the fact that I had most of his ancestral line already in my Tree, helped ThruLines… Jim

      Like

      Reply ↓

      • Nice! Cool on the process thing. I do basically the same. Your reply prompted another question regarding ”floating trees”. Have you ever toyed with the idea of connecting a floating tree to your actual tree (via some shadow person) to see if it prompts a Thrulines? Risky, I know. But otherwise updates to other people’s trees that hook to your floating tree would go unnoticed (I think).

        Liked by 1 person

      • Pat, I have linked a few Floating Branches to my Tree for a very short time – just enough, mid-week, to get some ThruLines (or not), and then I disconnect them. Well, one hit a goldmine and I’ve kept it, and got help from some cousins and some more records/info. But, in general, I’m reluctant to “foul” the Ancestry database – I’ve seen too many cases of Ancestry finding an unproven relationship, and sending everyone ThruLines hints. In this circumstance, I wish Ancestry would adopt the MyHeritage concept of confirming or rejection Theory of Family Relatives – where I’ve seen some of those “hints” then go away. Jim

        Like

  2. Jim,

    I have followed your blog with interest for a few years now, but if you’re drinking out of a fire hose I’m the guy in the desert who is squeezing a few drops of liquid from a cactus !

    I’m based in England with mostly Scottish ancestry so I have many fewer matches to start with (27,882). The number has crept up from 15,000 or so when I tested. I have 247 Common Ancestors, of which 19 are wrong. At least half of them are matches I placed myself which the Common Ancestor system then claimed, I hadn’t realised at first that it did this or I would have marked them somehow but there’s no point now. So a realistic number of useful Common Ancestors for me would be about 110 that Ancestry has found in seven years since I tested, for MyHeritage the number is about 10. I get one new suggested Common Ancestor every few months so I do keep a vague eye on it.

    Anecdotally from discussions on British forums these numbers aren’t at all out of line with other dna testers over here. We’re all in the same desert !

    Steve

    Liked by 1 person

    Reply ↓

    • Steve – I can relate, a little bit. In 2010 I did both 23andMe and FTDNA. It was new and we had few Matches. I guess the good news then was that I really worked hard on each one (and over the next 2 years, I didn’t have a clue how to use segments – it was all genealogy). In your case in the UK, the numbers probably won’t surge at any point. So… get the most out of what you have. I’d be Triangulating segments at MH (no genealogy required) – it’s a strong grouping method. And experimenting a lot with Clusster – the other powerful grouping method. AND I’d upload to GEDmatch – for additional Matches from FTDNA and 23andMe – which can also be incorporated into TGs and Clusters. Good luck. Jim

      Like

      Reply ↓

  3. Two quick questions — by “unlisted tree” are you referring to unlinked trees? And how do you handle what Blaine Bettinger calls the “unlinked family cluster”– more specifically, when do you believe it’s legit if you have some of those yourself? E.g. My mother’s matches include about 125 Hemphill-Mackie descendants (thru 9 different children) that points to an upstream link with one of her 3G great-grandmothers (ahnentafel 61). For my own tracking purposes, in Ancestry, I use a dot and notation as if they are a CA. (In addition, matches descending from this “CA” show up on the other DNA sites, and she also has matches for whom the “CA” would be Hemphill’s parents or Mackie’s parents.)

    Liked by 1 person

    Reply ↓

    • Cathy,
      You are right – ULT is UnLinked Tree – my typo. A critical takeaway is that Ancestry never looks at those Trees – some are 1-person duds; but some are goldmines… I don’t know until I open them.
      I call Blain’es UFCs a Floating Branch, and I have several – some with over 100 Matches all descending from a CA. I put them at the bottom of my CA spreadsheet (and assign a FG name instead of an Ahnentafel #, but they look the same as each of my Ancestor Groups (like Family Group Sheets). I’m convinced that each one is probably an NPE, often a missing spouse line. I study the geography of the CA, and try to pair it with my Brick Walls… Next step will be to see if some of the Matches will appear in known Clusters (or on a Shared Match list)…
      Your idea of a special DOT is an excellent one. Another chore: backtrack on several hundreds of these Matches and Dot them.
      Brick Walls are hard for a reason – usually: virtually no info about that ancestor. I have one that is a tic mark in an 1840 census – so I know the parents, but have never found a name (and he had to have my Y-DNA!). So a blank box in my Tree, but his line goes back for several more, known, generations.
      Let’s keep talking about how to tie in these FG/UFCs – geography, ethnicity, religion, migration, 1790-1840 census tics …

      Like

      Reply ↓

  4. Good gathering of thoughts on this one Jim … “Drinking Through a Firehose” is correct. I have pulled together (at the individual segment level) a single spreadsheet of chromosome segments by company as well …. an amazing dataset but one that is so “firehose like” that I have to return and re-return to it just to grasp portions of the story that it is “trying to tell me” …. I sometimes awaken with numbers and segments and color-coding and vendors “stuck in my head while visions of sugar … ” …. oops that’s yet another story line. 🙂

    Liked by 1 person

    Reply ↓

    • John, Yes both CAs and TGs make for similar, but different, spreadsheets – both of mine over 10,000 rows of firehoses. In my eBook: Segmentology Fundamentals, I devote a full chapter to each one. About 5 years ago I was caught up with all of my Shared DNA segments from FTDNA, 23andMe, MyHeritage and GEDmatch included – each one was in one of my 372 TGs. I then shifted my attention to the genealogy side of this hobby, and focused on Shared DNA Matches (and Clusters) – I’d pretty much caught up with that when Pro Tools came along – so it’s now back to “drinking” from each Shared Match to find our how *their* closest Matches fit in. It does keep me busy. I’m looking at the efficacy of putting all of the “DNA Connections” into a spreadsheet (many different names and generations) as a lookup for new Matches. If a Match had an ancestral “hit” with anyone on such a list, I’d have a CA *BINGO*.
      As a side note on the shared DNA segment spreadsheet – your TG segments are based on your DNA segments from Ancestors and your crossover points. This should be mirrored by each of the “segment” companies (FTDNA, MH, 23, GEDmatch). So I have them all in the same spreadsheet, but with a column for company so I can focus on one company at a time when adding/Triangulating new shared segments. But most of the time that spreadsheet is sorted by TGs, so I can focus on a CA consensus for each TG.

      Like

      Reply ↓

  5. Ciao. Si Jim ricevo ancora corrispondenze dna su mhyeritage che si aggiungono a quel gruppo che ti ho parlato molto alcuni sono rom alcuni sono misti rom e alcuni balcani. Loro sono 6kosovati 4 bosniaci 2 ungheresi 1polacca lituana rumeno greco e bulgaro.la triangolazione va sempre tra 7,1cM poi cambia io e due di loro ha 8,7fino ha 9cM.

    Liked by 1 person

    Reply ↓

    • Kevin – this is good news and shows you are on the right track. The additions from MyHeritage are compatible with what you had before. Hopefully, some of the new Matches will have good Trees that go back far enough to build a branch of your Tree. Good luck. Jim

      Like

      Reply ↓

Leave a comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.