Lost so many matches

[royfarnol]

Should we store our old matches?

Maybe this new build is not the cat's whiskers after all?

I don't understand that people who were less close before are now closer and people who were closer before are now less close? That seems an odd sort of algorithm.

Could the next build resurrect the old matches?

[Lincoln]

It is a shame that I lost the one match from my dad's side. All the rest were either my mom's or unknown. One line of my family was from Lorraine, France but always considered themselves German, with German names and spoke German. The match was from further back in time, when they were in what is now present-day Germany itself. I didn't know how the match connected, but it did give me chills when I saw it. It was the only match here that I would have recognized before I had started on my genealogy (my ancestor's name was a slight variation of my match's name).

(The cousin was a 5th Cousin - Remote Cousin - 45.41 9.46)

[T E Peterman]

Genomic information is constantly being refined. I'm not sure who designs the new builds, but this is presumably coming from the scientific community. Who knows what Build 38, or Build 39 will bring. This may shift around.

One of the enhancements that I would like to see Family Tree DNA develop would work as follows. Many of us have had a field of siblings or other close relatives tested. All of my father's cousins, no matter how distant, are also cousins to his brothers. My father & his brothers each have match lists that are somewhat different.

My proposal (and I have recommended this in the suggestion box) is that if I pull up an "in common with" report for my father & one of his brothers, it should list, not only those who meet the Family Finder threshhold for both (total greater than 20 cms/ LB greater than 7.7 cms), but the match should be listed if it only meets this criteria for one of the brothers. One could even apply this approach to first or second cousins; if you have tested a field of these from both sides of the family & a match is shared on one, but not the other, you would know which side the match is likely related on.

Timothy Peterman

[Cats]

I think it is "responsible" to do this because then you have a historical record of who Family Finder had matches you to, even if they don't currently. There is no reason not to do this. If you want to rely on a 3rd party (ie, Family Tree DNA) to store all of this information, there are a number of risks, including the possibility that matches are removed by a future Build.

Timothy Peterman

I tested my brother with FTDNA years before Family Finder came on the market. I now manage 7 family finder kits. I do not have the science background nor am I as well read about DNA as a lot of you. I marvel at the conversations and am still struggling to figure out what half identical means. I've sent my DNA and the other kits I manage to every 3rd party site or research project for my benefit and often times for theirs (aren't a lot of Lithuanians tested) and half the time I haven't any idea how to interpret the results they post (K2 or K 14 what is the difference?) I upgraded my brothers YDNA purely for the benefit of the haplogroup project as he had no close matches at 25 so why should I test more except for the benefit of the group. I try to understand and participate, I really do. I have certain people I really count on for advice and they don't even know it or maybe they do. At times you have been one of them. I even have a disputed result and FTDNA customer service still refuses to tell me the results of conversations with the "experts". (they don't call me back) I've had to rely on Matt, JOlson, ATurner and even draw conclusions myself, not very scientific ( I drew the conclusion I could accept easiest). Therefore when you say that a responsible FF participant should download their data I want to know why or what you mean by that. I think that I am a responsible participant. It would be great if you shared this with us BEFORE the date was wiped out, now I feel like I lost out on an opportunity then again not sure I really did, let me explain.

When FF switched from Affy to Illumina I downloaded all of the names of matches and saved them as suggested. Truthfully no one that I was I communication with before wanted to communicate with me after they were shown not to be related by the new and improved test. Truthfully I didn't want to continue to find a match with them either however, I did lose a few people that I enjoyed connecting with. These were people I spent time, sometimes a lot of time, trying to find connections to etc. Have to say that experience jaded me a bit. Since the switch to Illumina I haven't really tried hard to find connections to anyone that is considered remote. SO GLAD I DIDN'T. I lost a lot of those matches again with this new platform. I did not download the data except my close and distant matches, no other kits I manage. I do want to look at one match my mother lost but other than that I do not know what I would do with the old matches just like I never knew what to do with the old Affy matches I downloaded. However, I did not expect this to happen again as it did before. I really didn't. I thought there would be some kind of refinement but not a huge number of pages lost. With the new tests and tranfers I expected LOTS of new matches. Now I just have questions, how good was Illumina anyway? Are we still using Illumina or is Build 37 something different? It would have been great to have a heads up as I would have saved some of my kits matches. Yes I would have. I just did not expect this to happen again.

I remember how I felt when the Affy to Illumina switch occurred and now it has happened again. There are a lot of people who are newer and they do not understand or know what happened. Truthfully in a way neither do I. There are many testers who go into this thinking that you do not need to have a lot of knowledge to test and find matches. Sometimes the search is for birth family and the search is even more meaningful. A lot of people haven't any idea what just happened or why.

Could someone take the time to explain what just happened? Can you explain how the matches that were before us and are gone weren't really matches? Remember that some people are new and never thought they had to have an real understanding of DNA science in order to participate. My best match and truly the biggest find for me on FF is in his 70's. I couldn't even walk him through downloading his data to Gedmatch. I want people like him to feel comfortable testing and be given an explanation when things change. We should not need a degree in science to test and understand.
Thanks

[T E Peterman]

My only point is that we, as Family Finder participants, have two choices:

1. download your matches, so that you have a record of them.

2. don't download your matches & count on a 3rd party (ie, Family Tree DNA) to maintain them.

Most of the matches that people "lost" are really still there; their shared centimorgans are lurking just below the matching threshhold set by Family Tree DNA, which is 20 cms total & 7.7 longest block.

This happened because the overall half-identical centimorgan count was lowered from about 3387 to 3352.11. My father & I use to share 3380.69 cms; now we only share a paltry 3350.19 cms. I am no less related to him than I was last Thursday. I manage FF results for about 30 kits now, all close relatives, including some that are second cousins to my parents. In most cases, Build 37 lowered the shared cm count a bit, like from 190 to 185. A few actually increased, but most went down a little bit.

The same thing happened with everyone's matches, from the closest to the most distant. Simply because people have a lot more remote relatives than they have closes relatives, many matches for each of us, with Build 36, hovered just above the threshhold (20 cms total & 7.7 cms LB). Many of those lost just a bit, maybe only 1 cm or less; but when that happened, they slipped below the threshhold. They are still there; just not reported as matches.

Timothy Peterman

[CAM]

Cats this may help with some of what you are wondering about.

I understand what you are talking about I have 8 kits including my own which was a waste since both of my parents know have had FF. a lot of my matches neither parent still has. My FF I really do not look at anymore unless I get an e-mail for someone who only has me. Then I get to tell them sorry can't help you do not have either of my parents unless the case they have my mothers brother and not my mother. Yes I did have a few like this need to re check that.

This is FTDNA'S explanations

"Welcome to myFTDNA + Page Help Interactive Tour


February 22, 2013. As you can see, your Family Finder results are completed. The upgrade is complete. With that upgrade, however, you may notice some differences with your matches. The purpose of the build was to refine the data, so the number of matches in the speculative range may have been reduced by eliminating false positives. Matches in the speculative range may be legitimate distant matches, where distant could mean hundreds of years, or irrelevant matches created by “noise”—small coincidental matching segments that create the illusion of matching. The industry-standard build allows us to more accurately distinguish between real matches and "noise." Even some of your closer matches may move somewhat to more accurately reflect their genetic relationship to you. We will continue to work on improving our matching algorithm to give you more accurate results. Again, we thank you for your patience during this process.

Bennett Greenspan
Max Blankfeld

[vinnie]

Hi Cats,

You seem to be frustrated and discouraged, so I'd like to encourage you to hang in there through these changes.

My mother, father, and I all have the following message on our accounts:

"Friday, February 22, 2013 - Your kit is part of a small group that had some minor data compatibility issues so we're having to rerun your results through the matching program. This means that the matching portion of your results will be delayed until approximately Tuesday, Feb. 26th. You should have the Population Finder portion of the test, which you can access through the drop down menu from the navigation bar near the top of this page. We apologize for the inconvenience and appreciate your patience while we ensure you receive accurate, complete results."

I do not show up on my parents' accounts, and no one at all shows up on mine, whether under the matching or the chromosome browser tab. While the parent-child issue may be different from any "general" matching issues, obviously, our kits may affect everyone to whom we may match. So I think everyone should wait until at least Tuesday for these issues to be fixed for everyone else who's in the same situation as my family.

[kml2]

Hi Tim,
Thank you for the most coherent explanation I've seen of what happened.

I did think to download the data from my list before the change. Luckily. Over half my list disappeared, and the few new adds and some of the changes are very strange if the idea was to focus on matches closer in genealogical time.

I'm surprised by the radical nature of the decision. From what I've been reading, the issue of IBD vs. IBS is very far from decided.

I would rather have seen FT take FF in the direction of what they have done with MTdna results, that is, let you see the historical trends by including very distantly related (out to "distance 4) matches. Why do what for MT data is in effect to let us see thousands more years of separation and give up in FF a few more hundred years?

For those of us from ancestries with a dearth of records, the names and geographical locations of even very distant matches were useful in establishing the overall picture. We knew that we were most likely never going to find the common ancestor, but we learned a lot about the population movements of our ancestors from seeing the geographical and ethnic scope of those more distant matches. To me, this new list is a step back.

I'd rather have more information than less. I'd rather have the choice be mine than made by people who have no idea what might be helpful to me.

Kathy

[MFWare]

With all due respect, the problem here is more serious than changing criteria for matching levels. If you match someone in the database, then you are free to change to change the label on the match to whatever you want.

I now have a one-way match with my brother. On his list of matches, I am his closest match. On my list of matches, my brother is nowhere to be seen.

Family Finder in its current state is a massive foul-up. Hopefully, FT-DNA will straighten-out its mess soon. In the meantime, I recommend that FT-DNA's angry customers to keep their fingers crossed.

[JuanCarlos]

All I know is I lost most of my matches, but received no new ones. My brother lost even more matches, almost all, and received two new ones. My niece also lost most matches, but did not get any new matches.

[T E Peterman]

If you spot an anomaly like the one MFWare mentioned, use the Contact Us Suggestion Box & politely report the problem to Family Tree DNA.

I'm not about to download my new (reduced) match list until I hear reports that this type of thing has been resolved.

Timothy Peterman

[mollyblum]

I am totally confused. How can you not match your own brother?

I understand lowering the threshold but how can that affect a first or second degree relative? Does FtDNA know? If so do they think this is ok?

Do they know how that could affect adoptees looking for siblings or parents? What if their sibling is on here but-oops missed their threshold by 3 centograms or whatever they use..

Something is very wrong.

With all due respect, the problem here is more serious than changing criteria for matching levels. If you match someone in the database, then you are free to change to change the label on the match to whatever you want.

I now have a one-way match with my brother. On his list of matches, I am his closest match. On my list of matches, my brother is nowhere to be seen.

Family Finder in its current state is a massive foul-up. Hopefully, FT-DNA will straighten-out its mess soon. In the meantime, I recommend that FT-DNA's angry customers to keep their fingers crossed.

[Táltos]

All I know is I lost most of my matches, but received no new ones. My brother lost even more matches, almost all, and received two new ones. My niece also lost most matches, but did not get any new matches.

And yes. This is what has a lot of people wild (amongst other issues with Build 37). The hype, so many people thought they would have lots of matches with this Build/conversion. Instead we all got a bad haircut.

[Lincoln]

Did everyone get this posted on their page? I assume so, but who knows?

February 24, 2013. Thank you for your patience as we work out the last kinks in the Family Finder build 36 to build 37 upgrade process. We have noted problems with some kits in the testing process. These kits will be re-run and added back into matching in the next week. Meanwhile, you may notice that some of your matches even in the immediate category are missing from your match list. If the data indicates they are accurate matches, they will reappear on your list once all the final adjustments have been made. ...

[JuanCarlos]

And yes. This is what has a lot of people wild (amongst other issues with Build 37). The hype, so many people thought they would have lots of matches with this Build/conversion. Instead we all got a bad haircut.

That's right! I was expecting to see those matches that were held over during this transition, instead, like you said, I got a bad haircut too.