Start My Family Tree Welcome to Geni, home of the world's largest family tree.
Join Geni to explore your genealogy and family history in the World's Largest Family Tree.

How to best use Geni to link trees - a personal project shared to all

« Back to Projects Dashboard

Project Tags

Shared to everyone who want to build trees fast

Current Version 1.0 (rough version)

0. Health warning / work in progress

This is written in the context of South African genealogy, for genealogists working in other countries, you should extrapolate the ideas to specific aspects related to the country of research. Please contact me as I would gladly assist, and I would likely learn just as much as you can learn from me. Also contact me if you have better or additional suggestions and would like to contribute! Or start your own project, it is not hard!

1. How to translate knowledge into Geni Trees

Good place to start. Number one goal is to find if there is a profile which is exactly the same on Geni. With exactly the same there comes a health warning (see above) - not all people are work-horses, and although we all share the same DNA (going back far enough), we all do not share the same sense of accuracy or details. Just be aware of this. Exactly the same is not exactly the same, it is perspective, it could be that a name is sufficient for one, but the full names, DOB, POB, DOD, POD, DOM, order of children, all the children, solid proof, etc, is important for another.

This guide is written with the belief that one should always first use the "Search People" box located at the top line of Geni, to try and find the profile. The following sections deal with how to go from there.

2. Building a tree from the bottom up

If you read the last section you would gather that this is how I propose one should use Geni most of the time, if you are not a true Historian. I agree that there are some details that may be lost over time, and that the history is also important. But to link to the correct history, it is important that the present is OK, and the present therefore is more important.

Say you have the profile of a friend, then you are equipped to start. I mentioned this is South African right, so here are free and non-free resources.

OK, search the profile and you have then three options (only):

2.1 The profile exists on GENI

Great, so see how far DOWN the profile goes, and add the living persons information if it is missing, This you do in Tree View (View Tree) of the profile you just found. You need to set the options of that view to view 1 Generation, and Deselect 'Direct Ancestors Only' under Preferences, which is located at the bottom of the Tree View.

This enables you to also add the correct spouse, if it is missing, and is the only way in which you can, in one screen, view if someone already added the specific spouse.

2.2 The profile does not exist on GENI - but you can find the parents who do exist

OK, work for you and at this moment only you.

The first step is to find the parents of those profiles. You might have this available, then you first 'Search People' for the parents, if you find them then you repeat step 2.1 above from that level, otherwise you continue until you find the first parent available, you then add your details on the level below (and continue below) from there.

2.3 You cannot find anything.

A pity.

But there is still hope...

2.3.1 Modify your search

You should only search 'relevant' information, with that I mean that you search broadly. If there is more than one name, search only for the most common one.

Modify the spelling! Geni search is spell sensitive, so try alternative spellings. First only search on names. You get a lot of hits. Next steps is to search on DOB, again, only enter the year. Next step is to search on the name of spouse, here you only enter the surname of the spouse. Try different spellings too! Else try the most unique name of the spouse. Different spellings. No luck, then try the most common names. Try combinations.

2.3.2 No luck no matter what

OK, so you resort to add from the bottom, but you do so carefully as follows:

You start to add the bottom profile first. Add the person and the parents of that person. Add as much details that you can... DOB, DOB, POD, POB, DOM, Full Names. Place of Burial, Dates. Then you stop. Wait 5 minutes. Then open the Bottom profile, and view the bottom profile with 'Direct Ancestors only' (under Preferences) in Tree View. Does it pick up any matches? There are two 'circle icons' in tree view that are indicative of matches?

If nothing, perhaps if you add a sibling of the person you are considering. Add their spouses too, perhaps you can tie this from the female line. Continue adding information at the bottom. Wait 5 minutes, refresh the tree view with all the information you added and look for any matches. Only if none, with everything you got, then you should move to Section 3 below.

If you find something, you should still proceed with care:

2.3.3 Matches found

Now, in South Africa, a name and surname match only does not mean a lot. There are thousands of JHM Botha's walking around, which were born in the same year and married ME Van der Merwes. You need something more than just names. Exact DOB or POB can be helpful, as it limits the likelihood that we have different people. You probably, even if you find a likely match, still want to proceed to Section 3 and try and find more information.

2.4 Now build upwards

In Tree view, you now need to select the bottom profile, and then change the settings in preferences to only show direct ancestors and start with 3 generations. Zoom in and out with the middle mouse bottom so see all of them in one screen. Are there 1+2+ 4+8 profiles on your screen? If not, then your research (and restart the steps 2.1 to 2.3 above) area is clear.

Repeat the above for 5 generations, then 8 generations. The last one in particular does not fit in a screen, but you can modify your view by clicking on the Navigate button on the bottom of the screen. That will show you any missing branches. Well except if the profiles are already in your tree.

You can also do this for up to 20 generations, and there is also an add-in available on Geni to view the information in a circle format.


3. What you need to understand about country specific information

Unfortunately one has to have some knowledge about the history of the country in which you are doing genealogy. In South Africa, there is perhaps more examples than most, you need to know about the main events, where records were kept, traditions in marriage, names, farms. Every family has a unique history.

For South Africa again (Note too: the abbreviations used in this project can easily be searched on google - or ask me directly).

The Great Trek was 1835-1846, and during that, perhaps up to the 2nd Anglo Boer War up to the end of 1903, if a family partook, records may be gone, or extremely difficult to find. There were no archives and also very limited church records. In my opinion we only have strong traditional naming that was employed by these brave families, which in most cases, because the names of the children also honoured the wife's parents and grand parents, one is able to deduce the tree with great likelihood. This comes with a bit of experience, but a trained eye can be acquired quickly.

Before 1835 records can be found easily in the Cape and Graaff-Reinet (EC). Not complete, but most available electronically on NAAIRS, EggSA and Also, as people had many children those days, and parents died when there was still young children who could not fend for themselves, the MOOC archives is also valuable, although not complete. For all of these, not always with an easy index, but if you search long enough, have someone that can collect it if you are not in the area (archives, once you used the index). One has to account sometimes for different spelling in those eras, to name one obstacle.

Going back even further, although not 100% accurate, but close to that, are two sources, Kaapse Families and another (tbc), but in my opinion the information is already on Geni and is correctly revised, so the moment you start adding new profiles with DOB before 1835, warning lights should go off. At least check first using "Search People".

One should work smart, and there are many area related studies that can be purchased. Overberg and Noord-Kaap areas, family specific details (sometimes found at libraries, always available at our archives, as well as the research by GISA) can prevent duplicate research, if you can believe in the accuracy of other researchers.

4. Using the Research regularly

New duplicate information is added daily. The "Merge Center" under Research on the top line of Geni, gives a summary of profiles that have close matches. I use this daily, normally spotting people who create entire duplicate trees, going back to the 1600s.

Now it is important to note the best ways to use this function. The moment you add profiles, it will be a profile that you manage as well as a profile that you follow. If you did not create the first profile, but merged a profile you manage to one that existed before, it will be under profiles that you manage.

I definitely do not suggest that you create duplicate profiles (which are a lot of effort, wasted), but rather that you navigate upwards from profiles of interest and select the 'Follow' button located at the right top(ish) of the profile considered. This will enable you to use the Research screen to also find matches/duplicates of these profiles easily. How I do this, is to right click and to open both the father and the mother in separate windows or tabs, so that I can follow these profiles, and also from those windows repeat the process for their parents, and then for those parents. (Note: I follow thousands more profiles than those I manage.) As each profile has 2^n parents for the nth generation (see no 5 below why this is much less if n>5), it means that you quickly get to a situation where you will struggle to see all the tabs clearly, so use open in new window for some of the females (for example).

5. Overview of what makes it simple

I have studied 1000s of 100% South African trees for individuals, and there is not a single one where all the forefathers and -mothers were all unique. It is silly to deny this, it is just a fact when you have 500 males (?) and 300(?) females to start with. Do not be ashamed of it, and please do not delete information or break links or build your own 'clean tree'.

So to proof this I went and did the calculations. In the best scenario we can assume that partners were independent from each other (how they were chosen), so the probabilities below are best case scenarios. (for those interested, it is calculated based on a modification from the birthday date problem). Also note that the probabilities of having an unique tree which are stated below, are based from the year 1650, and assumes no new entries (for new entries a modification can be deduced based on the year of immigration).

3rd generation: 98.8% (2nd gen assume 100%)

4th generation: 86.0%

5th generation: 52.2%

6th generation: 6.5%

7th generation: 0.001% (1 in a 95 432 trees!)

After that is it millions, but the above would explain why I cannot find that 1 in 95 432 trees.

Finding the same parents actually reduces the amount of research required, exponentially.

Furthermore, I have not spotted one 100% tree in 1000s where a family has no ties to slaves, or only European ancestry, and even in that case the probabilities above are particularly worse, as the female pool becomes less than 100.

So to summarise: Link to the earliest known ancestor on Geni, check that the tree is complete (i.e. that it ties back to the first SV/SM that arrived), and focus your research on profiles which do not tie back to a SV/SM.

x. Motivation for this project

I put this last as it is probably not important. But I want to share how I feel about his - first read my profile. I need to add that I was away from genealogy for a few months, and when I returned nothing really happened on Geni. Well, except the same of the old problems, duplicate profiles and trees. Only bad things happened!

That makes me feel lonely. I need others to also work actively to clear up the tree, not really, but say a bus runs me over, then this project will at least capture the knowledge which, in the same collaboration effort of genealogy, must be shared. Every living moment we are in the position to share unique details which will no longer be there if we are not.

I am not unique in the sense of how I use Geni, but even the top users on Geni seem to be more involved in politics than actively contributing to linking profiles of the living. That is where our focus must be, as it is the quickest way to improve the accuracy of the tree! I motivate this paragraph by stating that the number of living is a large number...

Let's also be clear - at the end! Our common goal on Geni is to build a single tree that everyone can use. When your information (i.e. your tree, although not strictly yours) is accurate, then it is accurate for everyone (and now strictly!). Unfortunately the converse is also true, it is not the case of half of one....

Hopefully someone, not necessarily those that harm the accuracy the most, will read this, and be able to sort out some of the problems that continues to creep up. I therefore hope this is appreciated as both selfish and un-selfish, whatever that may mean, or is interpreted as I don't know what I mean... hard to explain, but those who understand I hope will contribute! :)