Gedmatch no longer accepting FTDNA kits

[Frederator]

I agree that the "privacy" issue seems . . .

I concur entirely with your analysis except to say that it would surely be very expensive to make the change to existing FTDNA kits at Gedmatch, both in terms of time and money. The pattern of dealings here demonstrates that even if a genuine "privacy issue" exists, it is entirely FTDNA's fault by reason of its own precedent in practice. FTDNA should pony up the $$$ needed.

For my money, the most outrageous aspect of all this is FTDNA's attempt to assert de facto ownership rights over the kits lawfully purchased by Gedmatch users.

FTDNA: You can't have it both ways, you can't have both our money AND a lock on our information.

[Germanica]

I'm not happy at because I have an important new match at Ancestry and I was going to ask him to upload to gedmatch and compare him with my cousins.

I don't understand - you can still upload Ancestry results to Gedmatch, there's no reason you can't ask this person to upload to Gedmatch. Unless you mean that your cousins tested with FTDNA? But if they have already uploaded to Gedmatch, they should still be there for the time being. Gedmatch says they MAY have to remove all FTDNA match results, but they haven't yet.

And if your cousins are at FTDNA, you and the other person on Ancestry can always do the transfer to FTDNA ($39 for the full results) to compare the results there. Of course, that means putting money in FTDNA's pocket and I suspect some people won't want to do that anymore, but it is an option.

[Armando]

I agree that the "privacy" issue seems to be related to the use of the FTDNA kit number in the Gedmatch ID.

....Gedmatch should change their program and assign a random ID to new uploads. In addition, they should replace the upload FTDNA menu option, to something like "Other companies upload". What will FTDNA have to argue about then?

The point is that the FTDNA kit number becomes associated with a specific person, through an e-mail address. That means that those FTDNA customers who upload to GEDmatch and are also in public projects at FTDNA have lost their privacy

I concur entirely with your analysis except to say that it would surely be very expensive to make the change to existing FTDNA kits at Gedmatch, both in terms of time and money.

I wouldn't think that a change in kit numbers would be hard or expensive. Are there any database programmers here with experience with large databases and ID changes?

[phx2128]

All this speculation that the Gedmatch requirement for an email address creates an issue doesn't make any sense to me. Agree that your kit is now "publicly" associated with an email and I use "publicly" loosely as you have to become a member at Gedmatch to see it. So what, the user still choose to upload and provide the email. My DNA, my choice!

This is a public relations nightmare for FTDNA. Communicate effectively with your customers! FTDNA has underestimated the support and value for Gedmatch in DNA analysis.

Definitely referring everyone to Ancestry WITH subsequent upload to Gedmatch until FTDNA works to facilitate a solution to this.

[hansonrf]

It is FTDNA,not GedMatch, who correlates the FTDNA kit number to the raw data file. Seems, IF this is really the problem, that FTDNA just needs to change/randomize the file name on the raw data download going forward so there is no predictable association. GedMatch then wouldn't know then, either...

Ancestry and 23andMe don't use identifiable 'kit numbers' on their sites, although 23andMe [at least used to...] embeds the user's name in the file name. I've found nothing identifiable in Ancestry raw data.

[telly441]

FTDNA has added a "Gedmatch' category to their contact us form. I suggest that people contact them using method to see if they will give any additional information.

It is obvious that they plan to continue the normal business practice of keeping their customers in the dark and not inform them of anything.

I wonder if ABC's 20/20 would be interested in this since they did a big story on the family finder product earlier this year?

[ech124]

Until FTDNA makes a public statement, I have to side with Gedmatch. The Gedmatch statement says that the issue only applies to FTDNA, not 23andme and not Ancestry. The only difference I can think of for FTDNA is the use of the ID number. It is extremely unfortunate that Gedmatch designed the database using that ID, but it is what it is. Going forward, it is probably easy to have new users assigned a random ID (they already do it for 23andme and ancestry). To change the existing users could be a massive job and it is probably why Gedmatch is resisting and there is an impasse. It is impossible to know without knowing the software. I would think even for Gedmatch to delete the existing kits would require an extended downtime to regenerate all the match tables.

[hansonrf]

Until FTDNA makes a public statement, I have to side with Gedmatch. The Gedmatch statement says that the issue only applies to FTDNA, not 23andme and not Ancestry. The only difference I can think of for FTDNA is the use of the ID number. It is extremely unfortunate that Gedmatch designed the database using that ID, but it is what it is. Going forward, it is probably easy to have new users assigned a random ID (they already do it for 23andme and ancestry). To change the existing users could be a massive job and it is probably why Gedmatch is resisting and there is an impasse. It is impossible to know without knowing the software. I would think even for Gedmatch to delete the existing kits would require an extended downtime to regenerate all the match tables.

23andMe 'kits' are not randomized by GedMatch; they are also extracted from the downloaded filename [which includes the user's name, too].

[telly441]

I have already downloaded all my raw data files from FTDNA and printed copies of everything in case they decide to restrict anything else from their paying customers.

Over kill, maybe but a cautious move given their recent behavior.

[Osiris]

1. It seems hypocritical in that FTDNA allows you to load data from other companies to their database but does not allow you to load their data to other databases. Would they also try and sue Ancestry.com if they allowed imports?

2. The raw data is just a file. As is 23andme and Ancestry.com. If you have any word processing skills you should be able to take your FTDNA raw data and make it look like another company and continue to load anywhere whenever you feel like. With this happening I'm sure we'll start to see developers making this process as easy as clicking a couple buttons. Kind of makes all their effort in this manner useless.

[gilbertdh]

Gedmatch, the last time I tried to find it, has no privacy policy at all.

FT DNA, 23andMe, and DNA Land do have privacy policies.

I believe DNA Land markets their privacy policy as one of the advantages of using their service.

Although privacy policies can change any time, it is at least some agreement and acknowledgement about what they intend to do with the data.

Is it possible that Gedmatch is doing something with our DNA data that FT DNA objects to?

I’m not a lawyer, but since FT DNA does generate the data from our DNA, they may have some right to tell another company not to do some things with it if it violates their patents or procedures.

David

[ech124]

It is smart to save whatever data you can now before the situation gets worse.

I am most worried about Gedmatch (since they seem to be the target of the aggression). Start to take screenshots of your data if you can.

[vinnie]

What if Gedmatch were to hide the contact e-mail addresses but still provide a contact link so that matches could still contact each other, but kits wouldn't have visible e-mails directly associated with kit numbers?

[Armando]

It is FTDNA,not GedMatch, who correlates the FTDNA kit number to the raw data file. Seems, IF this is really the problem, that FTDNA just needs to change/randomize the file name on the raw data download going forward so there is no predictable association. GedMatch then wouldn't know then, either...

The only place that the file has the kit number is in the file name itself which could easily be changed by the customer once they download the file. Gedmatch would ask for the kit number prior to an upload but that shouldn't have been necessary because it isn't necessary for a 23andme upload.

Ancestry and 23andMe don't use identifiable 'kit numbers' on their sites, although 23andMe [at least used to...] embeds the user's name in the file name. I've found nothing identifiable in Ancestry raw data.

23andme used a person's name as part of the file name as recently as January. That can also be changed easily.

[MMaddi]

What if Gedmatch were to hide the contact e-mail addresses but still provide a contact link so that matches could still contact each other, but kits wouldn't have visible e-mails directly associated with kit numbers?

That would remove one of the most attractive features of GEDmatch - the ability to directly contact matches, as we can at FTDNA. At AncestryDNA and 23andMe, they use an internal messaging system to contact matches, not direct e-mail addresses. We all know how that creates low response rates from matches.

Also, I don't know if the change you're suggesting would be burdensome time-wise and money-wise for a volunteer website.