* GenesReunited Hot Matches

[ChrisBowyer]

Having spent a week or so exchanging emails with GenesReunited support (who incidentally are extremely helpful) about how their hot-matches work (or don’t), I thought other FH users might be interested in the result. I’ll start with my conclusions and explain why afterwards. If you have a very small tree, and nothing but unusual names, don’t worry, it’ll probably work fine for you. Everyone else…

1. Remove everyone from your tree that you don’t want to get emails about.

2. Mark all the matching trees for people you’ve already contacted as ‘Don’t show me these again’.

3. Don’t mark any other trees as ‘Don’t show me these again’ even if they only contain a few non-interesting or spuriously matched names.

4. Make sure the box in your account details that says ‘Do you wish to be notified by a hot matches or new names email…’ is checked, even if you don’t want any emails.

You have to understand a bit about how it works… Once a fortnight they go through the names on your tree one by one and search for matching names on other people’s trees. They match by first name, last name, and approximate date of birth. A lot of these will be just coincidence of name and date, but the whole point about hot matches is that you can make an informed decision about whether you share relations with someone by the number of names matching between your trees… They display the tree with the most matching names first…

Sounds perfect doesn’t it… But, each fortnight they will only find 100 matches with your tree. So, when it gets to your John Smith born in 1900, you will get up to 100 other John Smiths born in 1900 from up to 100 other trees, and then it stops. You will have no idea without looking at each one to check the place of birth whether these are likely to be your John Smith, because having found 100 matching names it will not try to see whether those trees also contain matching children or parents. Next time (a fortnight later) you’ll get another 100 matches, and these may be added to the trees you’ve already matched with. Over time, the trees you really do share whole families or branches with will accumulate more matching names and drift up to the top of your list.

So, back to the advice:

1. Remove everyone from your tree that you don’t want to get emails about…

GenesReunited is, we’ve found, a great way of making contact with people you can share research with. Even if you’re terribly strict with yourself and check all the sources, it’s a lot easier when you know what you’re looking for. But it’s not a great way of publishing your research… find another site if that’s what you want (no one can see it on there anyway unless you explicitly allow them). So, if you have a big tree and you don’t want to be inundated with people who think they might be related to someone’s in-law’s cousin, leave them off. We used to upload everyone we know anything about (we all like to show off a bit), then we cut it down to all relations, spouses and immediate in-laws, but now its relations and spouses only. FH Split tree helper with a custom query is perfect for this. At least most of the messages we get from GR now are actually about people we’re interested in.

But apart from that, all these names that you’re not interested in are going to use up some of your 100 hot-matches per fortnight. This means it will take even longer for the names (or more to the point the families and branches) that you are interested in to appear as such in the hot match list.

2. Mark all the matching trees for people you’ve already contacted as ‘Don’t show me these again’…

You’ve already shared your tree with these people. You’ve probably copied names and dates from them, and they from you. There’s no point letting the next hot-match run waste its quota of 100 telling you about them. Let it find you something you didn’t already know instead. You can always see them again by clicking the ‘View ignored hot matches’ link.

3. Don’t mark any other trees as ‘Don’t show me these again’ even if they only contain a few non-interesting matched names…

Just because someone’s tree has a John Smith born in 1900, but on the other side of the country from yours, doesn’t mean they don’t share any other relations with you. Just leave them on the current list and ignore them if they have less that a few matched names. Next month (or next year) you may find they match a whole branch, in which case they’ll move up the list and you can have another look. If you say ‘Don’t show me these again’, it doesn’t just remove them from your current list, it also stops it adding any more matching names from that tree, so you would never know.

4. Make sure the box in your account details that says ‘Do you wish to be notified by a hot matches or new names email…’ is checked, even if you don’t want any e-mails…

If you say you don’t want to be notified by email it stops looking for matches for you. Maybe they’ll change the wording one day. Just another junk mail to delete once a fortnight I’m afraid (sigh).

And Finally…

Don’t mark all your hot matches as ‘Don’t show me these again’ even if that conflicts with the advice above. If you do, the next time you click the hot-matches link you get the page that says ‘No matches, don’t worry…’ but it doesn’t contain the link to view the ones you’ve hidden. You’ll have to wait until it finds some more before you can get them back.

[sheenash]

Chris,
I am surprised that no-one has seen fit to comment on your article as it is very clear and helpful.
I have been irritated too by these so-called matches which aren't. Now I can do something about it.
Despite this, I have had some useful 'hits' and been in touch with lost cousins.
Thanks for investigating this topic with GR.
Sheena

[Valjean]

This is excellent advice thanks. My only comment is that I read somewhere that I can only have up to 50 pages of hot matches? If this is true, and I keep all my hot matches that may one day relly match, I am soon going to run out of pages!!!

[ChrisBowyer]

I have 52 pages of 'current' hot matches at the moment, plus 8 pages in the ignored list, so I think the 50 pages limit must be a myth.

[Tombaston]

At the moment my tree on Genes Reunited consists of only one person, me. As my name is David Brown I have four pages of hot matches with more added every month. Your hints on how to cut these down to the real ones if I expand to my full tree are very welcome.

[RuthE]

Chris,
Thanks for explaining the process for us. I now understand why some of my contacts with whom I had lots of matches weren't coming up on the list for quite some time. My maternal grandparents are a Brown and Smith, so you can imagine my hot matches list is full of poor contacts. When I next do an upload, I shall be sure to cut down the number of Brown and Smith siblings to a minimum.

[hls]

Many thanks chris - I have been so overwhelmed with these matches (mostly wrong) that, other than wondering if i should delete them all, I hadn't considered an answer to the problem. As soon as i have come to grips with other than basic family historian functions (I am a new user)i shall definitely split my tree.

[vickyh]

Great articles but I find that the 'Don’t show me these again' option doesn't seem to work. I still get well established contacts popping up with each new batch.

[apttony]

An excellent article, Chris. I've now reset a lot of my family tree and match details. Waiting to see the result!

regards
Tony Thompson

[GladToBeGrey]

Excellent advice Chris. I've exchanged emails with GR about Hot Matches; I get very irritated by lists of supposedly Hot Matches when my person is born in (say) Devon and the so-called 'Hot Match' is born in America, or Australia, or Yorkshire ... etc etc.

I asked whether it would be possible to, perhaps optionally, also compare the PoB field, giving it some weighting based on the degree of correspondence, and then set a threshold value for genuinely Hot Matches versus the rest, which are mainly stone cold.

But GR say it's too difficult, despite there being several well documented techniques for doing this kind of comparison already in existence.