Merges - Maybe Sometimes It's Just Not Worth The Trouble

What is your definition of "private tree?"

Geni's is here

http://www.geni.com/worldfamilytree

and here

http://www.geni.com/company/privacy

Excuse me . . .I should have said "stand alone tree", or, "new tree being created by a new user", or, "old tree that has been rediscovered". The issue of Privacy has nothing to do with this question.

I asked because there are technical differences so we're driving to a common, well defined terminology.

A "standalone tree" is one that is not connected to the World Family Tree. We all start as "standalones" unless invited by a family member to join their already connected tree.

So for me, it was a moment of glory when I was able to join, by merge into a "connected" tree, and discover new family branches. It was the reason I joined Geni. Of course it resulted in duplicates that needed to be merged to existing profiles. I have had a lot of technical difficulties "deleting" profiles - it kicks me back to the home page - so I learned to merge, then resolve data conflicts to the "main" profile (furthest to left) right after the merge.

I understand exactly what you are saying, and why you are saying it, and I agree with you, although this is not at all what I am trying to discuss.
Perhaps I have not made my question clear, which although unlikely, is plausible. I regret starting this discussion and will no longer be following it.

After further consideration, and some influential input, my decision to abandon this discussion has been reversed, and I will attempt to continue to share this train of thought.

I thought that this was a valid and a good question.

Those of us who care for shared parts of the tree spend time merging in duplicate trees -- very often with less or even incorrect information -- and here I mean trees that are not standalone.

These trees are created often by new members who either do not yet understand the collaborative, shared tree aspect of Geni or who have not found the profile(s) they added when using Geni search.

That's my view, that if we could educate new members better, they could start building their trees bottom up FROM THE MOST RECENT FAMILY MEMBERS and that when they went to add a profile that was in the World Family Tree, Geni would let them know that. It does in fact, but perhaps not quickly enough, not reliably enough (because of the complexities of name matching) and not in a way that catches their attention.

I for one do not like the fact that I spend time merging in duplicates but here's why I do it: (1) others have already merged one or more profiles in the tree (2) when I tried to get people to delete duplicate trees in the past, they couldn't or wouldn't and (3) if I don't merge, someone else is sure to and they will do it more automatically with less thought and research than I will perhaps, with more problems ensuing.

I'd love to hear others' approaches and views.

It really would be useful if Geni coud grab a new user right away and offer to connect them to the Big Tree BEFORE they add a hundred unnecessary duplicates. I would bet that almost everyone could be connected before they add duplicates more than a generation or two beyond their private tree.
And then, there could maybe even be a system that does not allow future merges to be suggested for those who have opted out of the original offer of connection. Part of what inspired the original question/consternation was the crunching of numbers. 19,919 profiles were added yesterday, while only 1,936 merges were completed. How many of those 19,919 were duplicates?

Hatte, well said - I thought it a good question, too.

We've seen tools evolve & improve as Geni has grown from the some 7 million connected profiles (when I connected to the World Family Tree) to over 71 million today. We've seen process improve also.

We're faced with scaling issues, perhaps.

For trees such as the ones W describes -- standalone with no added value, etc --- I suggest - send a message to manager(s) -- if get a response indicating desire to merge, then work with to accomplish it [thus folks in Erica's position before connecting will be happy]; if no response or not interested in merging -- follow W's suggestion to leave alone.

If messaging managers is too much bother for you, then suggest just follow W's suggestion to leave alone.

W - you realize folks can only export 5 times the number of profiles they added -- so if they do not first build a decent sized tree of profiles they added, they cannot get a decent Gedcom? Given that situation, it makes total good sense on their part to first add the folks they have info on, then do a Gedcom export, then allow a merge.

It is also easier to cleanly add the folks you have info on, etc -- vs attempting to untangle a mess when you have not first established a clear picture. Another reason for folks to first establish a tree of their own before choosing to merge.

See, there is a lot to be said on this highly relevant topic!

I had not thought about the export angle. I hate the idea that the limit on export means that people are incentivized to duplicate.

And the point about building a clear "sketch" tree before merging is well taken. In fact it's a method I use.

If it is true that only 5 times the amount of profiles added can be exported, which then inspires the addition of who knows how many duplicates by tree-builders wishing to create a personal Gedcom file, then it seems that the policies currently in place are working against the creation of a one-profile-per-person consolidated Big Tree instead of working towards that stated goal.
Currently, there will never be enough merging of existing duplicate profiles completed to overcome the influx of new duplicates. In the last 7 days 131,098 profiles were added and 12,535 merges were completed. That leaves 118,563 profiles unmerged, a very high percentage of which MUST logically be duplicates because of the relative (pun intended) completeness of the existing branches.
If all of my close genealogist cousins decided to start their own new tree with the intent of exporting a Gedcom, and made it as complete as my tree is before being either willingly or unwillingly connected to the Big Tree, what number of new duplicates would that amount to? And, what's more, what does that say about the true number of profiles in the big tree if most of them are duplicates stuck in merge-limbo?

Couldn't a script be written that could more automate the merge process?

Adding duplicates to become a manager does not help on your GEDCOM export.
It is the Added By marker that counts, and the Added By of the oldest profile in a merge wins that honor. Fair enough,.

Erica Howton
Perhaps your question provides its own answer . . .maybe merges are not really wanted, because unmerged duplicates make the tree look bigger than it really is, which increases the "curb appeal" of the website.
If 100,000 people add duplicates of the same 100 people, the tree suddenly contains ten million profiles! Currently the Big Tree connects me to 75,000,000 "people" . . .so . . .
Don't get me wrong. I LOVE the website and would LIKE to see it working at optimum efficiency and growth, creating a "one for one" tree of over (at least) a thousand million (one billion) real people, but given the current state of affairs I just don't see how that's going to be possible. With over 120 million births worldwide this year so far, how will we ever make a profile for everyone? B^]

I am feeling that "scaling" word again.:)

One profile for every person, without duplication, is the goal; "everyone is connected" is the motto.

I think we're (just?) faced with "how" - not vision.

Private User

I feel as if you are answering a question that was not asked, so I'm a little confused by your statement and am not quite understanding what you mean to say.

Please explain how "adding duplicates to become a manager" and the honor of holding the oldest "Added By" marker relates to our discussion about the idea that the current limit on export may mean that new users are prompted to add duplicates of branches of trees that already exist, which works against us getting all of the duplicates merged.

W - I think the title of this discussion, says it all. Your point becomes deviated in discussions - meantime, one is simply confronted with long lists of accumulating matches (sometimes duplicated with same managers' names). Historical matches can be interesting and well-documented, but to confirm each, one by one...partly in the hope of clearing what seems like a backlog?? I have written to request advice or a solution from Geni,but no response. It appears the only way to deal with this problem which you (gratefully) compare,, is, to make own selections, leave the rest, and hope for the best, while not swamping the original tree with unknowns, if at all possible. (extra technology aside

Re: a comment I made on Nov 12 -- limit on Gedcom Exports is now only 4 times number you added!!! And that is up to a max of 5000 for basic, 100,000 for pro.

based on updated by date in FAQ in Help - apparently changed Nov 15 2013!! http://help.geni.com/entries/466639-How-can-I-export-my-GEDCOM-

Maybe they changed it to 4 after you posted it was 5?
I wouldn't doubt it.
B^]