Does WikiTree validate members' GEDmatch kit numbers?

Question

One of this week's new featured profiles has a third cousin who is a DNA-tested WikiTree member.

EDIT: There was initially no DNA Connections section on the featured profile. My initial guess was that this was because the profile is "Private with Public Biography and Family Tree". The DNA Connections section subsequently appeared, at which point I realised that two profiles for the featured profile's father had only recently been merged and that the implications of the merge had not yet had time to work their way through the system.

I'm not sure that there is any point in omitting the DNA Connections section, as clicking the link to his father's Open profile more or less reveals what would be in the DNA Connections section of the featured profile.  Theoretically, the son could be expected to share either half as much or twice as much DNA as the father with the WikiTree member, depending on whether the WikiTree member was a distant cousin (as established by a second click to view the relationship) or a descendant (it is unlikely that a man born only 108 years ago could have a GGGGgrandchild old enough to be a WikiTree member).

The real problem is that the WikiTree member's GEDmatch kit number is missing a digit - it is in the format Axxxxx instead of Axxxxxx.  Shouldn't WikiTree be validating GEDmatch kit numbers when they are entered?  Or at least checking that they are in a valid GEDmatch format?

I can't see the WikiTree+ suggestions for the member's profile, which is also "Private with Public Biography and Family Tree", but perhaps the member, whose last contribution was in June 2019, would see such a suggestion if she logged in again.

Another problem is that the featured profile's mother was the subject of a merge within the last few hours, so that the Connection Finder is currently displaying nonsense in the first few degrees - it shows:

David Lynch
→
✖Edwina Sundholm
(his wife) →
✖Edwina Sundholm
(her wife) →
✖Donald Lynch
(her husband)

where it should show:

David Lynch
→
✖Donald Lynch
(his father)

Past experience has taught me that this problem will right itself within the next 24 hours or so, and that I (and many others) will then appear to be two degrees closer to the featured profile.

I have been advised myself in the past to avoid merges or changes of LNAB for currently featured profiles or their close relatives!

Answer 1

There is some simple validation when someone enters the number -- if you enter an incorrect number you will see the message "The GEDMatch ID should be one or two letters followed by five to seven digits, optionally followed by a letter and digit."

But it doesn't check to see if the number actually works. If you find someone has entered an incorrect number, a quick message to them might be helpful.

Comments

Thanks, Jamie. I don't recall coming across a GEDmatch kit number with only five digits, at least not since GEDmatch changed all the kit numbers comprising an "F" followed by an FTDNA kit number, which could have had as few as one digit. Are you certain that five-digit kit numbers are still possible?

---

If I'm reading it correctly, it's an Ancestry kit, and if I had to guess, the 5-digit kits were some of the earliest ones. So that DNA result has most likely been around for quite awhile. But I'm not an expert on the DNA end of things, so may be reading it incorrectly.

---

I am virtually certain that the Axxxxxx GEDmatch kit numbers always comprised the letter A followed by exactly six random digits, between 0 and 9 inclusive. As this would imply, about 10% of those in my spreadsheet of kits of interest begin with A0. It wouldn't make sense to have some kit identifiers including leading zeroes for numbers below 100000 and others without the leading zeroes.

This is completely different from the FTDNA kit numbers and the former GEDmatch kit numbers derived from those. The oldest kit number in my largest FTDNA project is 7198, dating from January 2003. I presume this was preceded by one-digit, two-digit and three-digit kit numbers, and the latest all-numeric kit numbers have seven digits, with 1011201 dating from December 2023.

A kit identifier consisting of an A followed by only five digits should certainly be automatically rejected by WikiTree as invalid.

Answer 2

There were multiple merges with his parents that caused this temporary duplication. If you check it again today, you'll notice that those have been resolved. Unfortunately, post-merge, there's about a 24-hour or less period where it takes an overnight cycle to clean up some of those inconsistencies.

Comments

I have edited my original post above to acknowledge that the overnight cycle after the merges also cleaned up the inconsistencies regarding the absence of the DNA Connections part of the featured profile

Answer 3

There is a WT+ suggestion checking the GEDmatch ID. See https://www.wikitree.com/wiki/Space:DBE_171

Comments

What an interesting variety of errors!

I spotted Y-DNA and mtDNA haplogroup labels entered in place of GEDmatch kit numbers; kit numbers with the digit 0 replaced by the letter O or o; old Fxxxxxx kit numbers from the days when the GEDmatch kit number was the FTDNA kit number preceded by an F; surnames, usernames, etc.

I hope that most of these are now being caught at the initial data-entry stage.

Answer 4

I left a message on Ancestry for the member with the typo in the GEDMatch number. Hopefully, she will see it there and fix the problem.

Ken