One million more connected, how far are you from them?

+22 votes
286 views

Since we started the 100 circles study four months ago, the number of connected profiles has grown by more than one million. I have compared for the original "Focus profile" Jean-Joseph Vatant, the population of circles before and after this extra million. The image below is showing the growth of circles 1 to 40 (click to enlarge).

The growth of the very first circles is not visible at this scale, and relatively very small (putting my steady stubborn work on them in perspective, just in case I'd think I'd achieved a lot).  The growth of circles 1 to 20 is roughly one thousand profiles, that is one thousandth of the total growth. 

At first sight the circles with the largest growth are roughly the ones which were already the most populated (around distance 30), but looking more closely, it appears that the greatest growth happens just a bit closer than the most populated circle.

The most populated circle was in November and is still in February the 30th, but the maximal growth has taken place in the 28th and 29th circles.

Meanwhile the mean distance to/from Jean-Joseph has "dropped" a little, from 33.15 to 33.01, still waiting for the magic shortcut which would make it drop drastically.

In interpreting such figures, I would not want to get carried away. The growth of circles is due both to addition of new profiles, and reconnections bringing back profiles from further circles. The part of each cause is difficult to assess, but what is observed for the 28th and 29th circles has probably something to do with reconnections.

Similar work for other Focus profiles is under way, and we'll of course continue to monitor those figures over time, and post updates as seen fit.

WikiTree profile: Space:100_Circles
in The Tree House by Bernard Vatant G2G6 Pilot (176k points)
edited by Bernard Vatant
Very interesting, Bernard.  Thank you for the update!

1 Answer

+9 votes

It is not all smooth and even. I have been collecting data for Olof Andersson at weekly intervals from 18 November and I agree that maximum growth is usually in circles slightly closer than the maximum. But graphing all of the samples together, one of them sticks out. Something happened in the tree between Dec 28 and Jan 4, drastically increasing the population in the circles nearest to his most populated circle and correspondingly shrinking the circles beyond. As if a large chunk of the tree had moved from one place to the other.

Similar things happened with other profiles at that time - I have just not been following them as closely, so it's harder to pinpoint. It does not seem to have been a new, drastically shorter connection for Olof individually. Nothing like when a new connection brought Werkelin-2 four steps closer to the rest of the tree.

by Eva Ekeblad G2G6 Pilot (583k points)
Eva, what do you think happened?

I'm not at all sure, of course. But I do suspect that something like this, that shows up in the growth data for profiles very differently situated in the Global Tree, without having any dramatic effect on the mean distance of the individual profile, is caused by an event at a node that is an important junction in the tree as a whole. Unfortunately the diagrams don't tell us where :-)

It could be a merge, a new connection or a disconnection.

I know Bernard noticed something similar with Marie Stuart at the same time - and for Pocahontas it was the opposite: a big drop in the circles closer than the "peak" and corresponding growth in the circles beyond. A research problem here is to have comparable "before" and "after" data for profiles situated at interesting spots in the tree.

I'm glad Bernard posted something from the four-months-into-this point; I keep thinking to myself that after a year we will know enough to start some real research ;-)

Thanks, Eva.  I can see how a merge or new connection could move circles closer, but it is harder to imagine how a disconnection would do that.

As for gathering data for profiles situated at interesting spots: Could such spots/profiles be identified now?

Related questions

+9 votes
1 answer
+39 votes
8 answers
+17 votes
4 answers
+13 votes
21 answers
+12 votes
2 answers
585 views asked Feb 18, 2022 in The Tree House by Bernard Vatant G2G6 Pilot (176k points)
+25 votes
1 answer
+24 votes
4 answers
+40 votes
1 answer

WikiTree  ~  About  ~  Help Help  ~  Search Person Search  ~  Surname:

disclaimer - terms - copyright

...