Don't they read emails?

[sunkeeper]

I recently sent an inquiry to ask why a person with a considerable match to me on Gedmatch didn't show as match at FTDNA. I clearly stated we shared 14.8 cm on chr 22 and a combined total of 22 cm on all matching chr. Today I receive a reply with the following:
Our algorithm requires a longest block of at least 9 cM and then at least 20 cM shared to make a match. Our requirements are different from GEDmatch.

Does 18.4 not exceed 9 any more or is it the total 22 is not high enough above 20?

How many other matches am I missing?

[The_Contemplator]

...we shared 14.8 cm ...
Does 18.4 not exceed 9 any more ...

Well both of those numbers (14.8 and 18.4) still exceed 9.

The customer reps might not read all the details and just send the prepared replies they are told to send. There could be other aspects of the algorithm that FTDNA takes into account that Gedmatch does not. For example when there are mismatched bits within the matching segment. With Gedmatch you can modify how strict it gets when it comes to those. No idea on how FTDNA does it.

You could simply reply to that email and clarify. Or private message Darren here in the forums to see if he can help. Click here to message him. Give him the kit number and details.

[thetick]

You are likely just getting the canned response.

It is also possible your long match crosses a Centromere which FTDNA will break up.

See https://en.wikipedia.org/wiki/Centromere

[sunkeeper]

Thanks for both replies. I figured it was a canned reply, I had thought before they spewed the egg all over their faces they might have at least read the numbers first.

Very poor customer service.

[John Strathern]

Do administrators read gripes

Sunkeeper: Never mind about them reading emails. Don't think they read posts. Take the money and run. I don't go on Ftdna often. Generally problems. Family tree is one of the wort out of any site. Anyway: Do they read posts? I envision a person in a dark room, with thick glasses reading and responding to various gripes. He or she works part-time.

[gtc]

private message Darren here in the forums to see if he can help. Click here to message him. Give him the kit number and details.

This.

[Ann Turner]

Our algorithm requires a longest block of at least 9 cM and then at least 20 cM shared to make a match. Our requirements are different from GEDmatch.

That 9 cM statement is news to me, and I'm not sure it's correct. I just checked my matches for the last couple of months, and I'm seeing some in the 7.7 cM range.

But back to your question: GEDmatch does have a higher tolerance for mismatches. If you display the graphic bar, do you see any red lines? If so, FTDNA might not count that section as a single segment.

[sunkeeper]

This is what the largest chr looks like, don't think there is any red.

[Ann Turner]

There might be a tad at the end of the segment that FTDNA wouldn't include, but it doesn't seem like it would be enough to disqualify it (unless the cM value changes drastically, which seems unlikely).

Maybe the small segments are the problem? In addition to the mismatch bunching limit at GEDmatch, the way FTDNA uses blocks of 100 SNPs might underestimate the "true" length of the segment.

Personally, I'd much prefer that FTDNA abandon the use of small segments entirely for calculating relationships. It seems to create a lot of false negatives.

[ech124]

FTDNA is not a unique example.

I have cases where my 23andme kit shows a good sized match to another 23andme kit on Gedmatch. Yet when we shared at 23andme, there was a zero match. In yet another example, there was a match, but at 5 cM on 23andme (far less than Gedmatch).

Every company out there including Gedmatch does their analysis in a slightly different way. In the majority of cases the results are consistent, but not always.

[sunkeeper]

FTDNA is not a unique example.

Every company out there including Gedmatch does their analysis in a slightly different way. In the majority of cases the results are consistent, but not always.

I understand they all have their way of doing things, I just find this one in particular to be a bit to big for it not to be showing. Over 14 cm of a match is a bunch.

My biggest question is why did I get the robo answer I did? They clearly do not read the emails or I wouldn't have been so insulted by the reply. I didn't understand why it was not showing as a match, obviously neither do they!

[gtc]

My biggest question is why did I get the robo answer I did?

That's a question best directed to the owner/managers, Bennett Greenspan and Max Blankfeld.

[The_Contemplator]

My biggest question is why did I get the robo answer I did? They clearly do not read the emails or I wouldn't have been so insulted by the reply.

So far they tend to read all my messages fine. Maybe they read it and skipped the numbers given that you gave two different ones. So instead of asking for clarity they figured giving you the basic info would be enough.

In the image you showed, there is clearly red lines over the area covered by the blue bar. So maybe FTDNA sees that as one too many mismatches to count the whole segment.

No need to feel insulted by them. They likely handle a lot of messages and I bet a lot of them are from people that don't understand the basics. So once in a while they might mix some people with those people asking the basic questions.