We're getting close to 22 million profiles

+26 votes
550 views

As of 9:57 Eastern Standard time on November 25, 2019, "Our tree includes 21,936,991 profiles". At the usual rate of growth, we should pass 22,000,000 about the first week in December.

in The Tree House by Greg Slade G2G6 Pilot (691k points)
I would be curious to know if among recently created profiles, what is the percentage of sourced vs unsourced? Or the percentage of humanly created or gedcom created? Maybe that's already been answered? Does this mean we're getting millions more junky profiles?

22 million would have truly meant something if the base standard wasn't this.

I'm not sure that anybody has gone through and checked (and it would take some time to do so), but based on the profiles I've seen, I would say that new profiles being added now are much more complete than profiles added years ago. I noticed in the Timeline page that WikiTree didn't even have fields for the places of birth and death until February 2009, I couldn't find when the {{Unsourced}} template was implemented, but most of the profiles that I have come across that are unsourced and unconnected date from before 2012.

Greg, that's encouraging. Thanks.

22,008,951 profiles right now.  When did we hit 22 even?

5 Answers

+16 votes
I noticed that earlier this morning myself. It motivated me to go ahead and add a profile for my wife's great-great-grandfather.

More than 300 profiles have been added since you posted 20 minutes ago.

Growing quickly!
by Thomas Fuller G2G6 Mach 9 (94.5k points)
Whoot! Whoot!  Keep them coming!
+12 votes
This is an interesting contrast with another recent post:

https://www.wikitree.com/g2g/946368/same-sorry-situation-with-sources
by Living Kelts G2G6 Pilot (554k points)

That's true. There are people who argue (strongly) that speed is the enemy of quality. I disagree. While everybody has a speed which is too fast for them to be able to do quality work, that speed is different for each person. 

Many years ago, I used to work through a temp agency, and one kind of gig that I would get on a regular basis was helping companies do stock inventories. The person supervising us temps would always caution us that accuracy was more important than speed. But after I had done those kinds of gigs enough times, I got to be pretty fast at it. I used to worry that that would upset the client, but one client told me that I had the lowest rate of errors out of all the batch of temps they had in on that occasion. So I stopped worrying about it and just worked at the speed that worked for me.

I suspect that, if somebody studied these kinds of tasks, they'd end up with what's called a "bathtub" curve: at the slowest rate of working, the accuracy is probably lower, because the worker gets bored at having to go so slow, and isn't paying enough attention. Then, the accuracy would rise as the worker reaches the optimum speed range for them, and then drop again as the rate gets too fast for them to keep up. (I'm trying to banish the image of Lucille Ball trying to work on an assembly line in a chocolate factory from my mind, but that's something like what I'm talking about.)

The trouble is that some people assume that the slower you go, the more accurate you're going to be. And some people assume that anybody who is working faster than they are "must not be doing it right". But life isn't that way. At many things, pretty much anybody could work both faster and better than I can. At some things, I'm actually both faster and better than other people. (It would be nice if the tasks that I can do faster and better actually paid well...)

The main point that I keep coming back to again and again is that people are different. I mean, really different. Not just that people can differ wildly from one another in all kinds of ways (height, weight, speed, agility, ability to work out the square roots of large numbers in their heads, etc.), but that the number of ways in which people can differ from one another is immense. 

Then, too, they way people like to work is yet another way in which people differ. Some people prefer to work on one profile until it's "done" (whatever their definition of "done" is), and not move on to another profile until then. Other people like to whip through a whole list of people, adding one thing, then going back and adding another thing, then going back a third time and adding still more, and so on. 

The information gets added to the tree either way, so it's all good. The problem comes when people try to impose their preferences on other people. If we were talking about a work situation, where people were being paid to contribute to WikiTree, that might be acceptable. (Well, for some people. Personally, I refuse to work for people who try to micromanage like that. What matters is -- or at least should be -- that the job gets done, not whether I paint the wall from left to right or right to left.) But with a project like WikiTree, which depends entirely on work contributed by volunteers, getting all picky about how people work is a really good way to drive people away. (And, since we're ad-supported, driving people away reduces the ad income on the site, thereby threatening the continued existence of this sandbox that I love playing in so much, so it really annoys me.)

Greg, you make some good points.  Once long ago I was a telephone operator (in the days of very primitive technology), and although my statistics were among the best for speed and error rate, I got fired anyway.  For not following rules that didn't help me.  Oh well, nothing to do with the question at hand, I suppose.  However, are you suggesting we have no rules at all, because they can't be applied to every case?  

The downside of just letting people do whatever they want, with no serious standards, is that when people visit WikiTree, they are likely to find a lot of junky profiles, and that is not helpful in attracting the kind of new member who could really contribute.  Many of my genealogy contacts from Ancestry and elsewhere don't want to come here because of all the wrong information already here.

Oh, we have all kinds of standards here. Not just the Honor Code (although it might help if everybody re-read that at least once a year), but the Help pages are full of directions on the proper way to do things. And beyond the standards, there are recommendations: things that WikiTree doesn't require, but encourages people to work towards as their skills improve. At its best, WikiTree is a wonderful example of people encouraging one another, sharing tips, and all of us progressing. Fortunately for me, nobody came around passing judgement on my early profiles. I slowly learned, went back and improved what I had done before, and will probably learn more in the future and have to go back and fix still more. 

The problem that I see is one of tone. There's an old saying that, "You catch more flies with honey than with vinegar." I think that it's particularly appropriate to WikiTree. We desperately need more people working on WikiTree: adding profiles, sourcing, connecting, adding categories, writing biographies, adding photos, etc. You name it, and it needs to be done, not just to one or a few profiles, but millions. 

So when somebody new comes in, doing things "wrong", what we need to do is welcome them, point them to the appropriate help pages that deal with the stuff they're doing "wrong" (and remembering that there can be a huge difference between "the way I prefer to do things" and "wrong"), and encourage them to up their game.

What we should not be doing is jumping all over them over every little detail, because the most likely result of that is that we'll drive them away. So instead of ending up with somebody else helping to get stuff done around here, we end up with somebody who's going to hang out on all the other genealogy boards talking what a bunch of jerks we are.

 As for people on Ancestry complaining about junky profiles here, there are plenty of junky profiles (and whole trees) on Ancestry, just like there are on FamilySearch, Geni, MyHeritage, FindMyPast, and every other genealogy site. (There's another old saying about the pot calling the kettle black.) If the presence of data errors means that you shouldn't use the site that those errors are on, then there is no genealogy site that anybody should use, because there are errors everywhere. But my attitude is that the correct response to finding an error is finding the sources, and using them to correct that error. But people can't fix the errors on WikiTree unless they're here on WikiTree. Whining about it from a distance is just a waste of time.

Of course there are also crappy profiles on Ancestry (but when I find garbage on Ancestry, it's not on my tree, so who cares?).  The people who make them are not the ones to whom I was referring.  And I'm not whining from a distance.  I am an active WikiTree member.  My contacts are not whining either.  They've just chosen not to participate here, much to my regret for reasons including wanting their expertise, and wanting their DNA information posted here.

It isn't just a harmless little problem when people come to WikiTree and ignorantly post new profiles willy-nilly.  Nearly every time I do a search in advance of adding a profile to WikiTree, I find (on average) five to ten empty profiles for the name I'm searching, most so incomplete I have no idea of whether I'm about to create a duplicate or not.  Other times I find obvious duplicates, some that have existed for years such as two wives of a particular person, both attached to him, with the same name.  Yet no one has done anything!  Yes, there are people working diligently to clean up WikiTree profiles, but it seems the number of problem profiles is growing, not shrinking.

P.S.  Who wants flies?
+9 votes
I suspect that we will reach the 22m still in November. My experience is that when we approach a new million the profiles count is rising faster than usual.
by Jelena Eckstädt G2G Astronaut (1.5m points)
+10 votes

Thanks for the reminder, Greg! 

As for the prediction, I tend to agree: Sometime on December 2nd. 

There are currently 4 days remaining in November (4 d, 4 h in UTC time when I last updated my WikiTree spreadsheet). Approximately 52000 profiles remain between the present total and 22 million.  That would require 12700 profiles per day. If we want to be generous that could be 4 d, 16 h until it is no longer November anywhere: That threshold would require 11320 profiles per day. 

Typically, WikiTree is growing by about 8500 profiles per day. While it is true that there are brief periods when this growth is significantly higher, these, on fairly cursory examination the data, do not appear to line up with million markers (as suggested by Jelena Eckstädt) or even 10, 15, and 20 million markers:

Milestone Rate in lead-up
10 M 9160 p/d
15 M 9970 p/d
20 M 8830 p/d

These are higher than the 8500, however there is some variability, depending on exactly how and when the rate is measured (explained with the 2nd graph). Overall, the data seems to be fairly consistent across larger stretches of time, as suggested by the graph of WikiTree's cumulative profile count:

Graph of WikiTree's Growth: Cumulative number of profiles by date.

However this does hide some of the variability in the data that I have. This hasn't been a thing that I check every day, so my data isn't perfectly uniform. But I can look at how consistent the rate of growth has been if the changes in profile count is plotted against the sampling interval (à la the sideways funnel plot):

The data does appear centred around 8500 profiles / day.

I also checked the growth rate from around US Thanksgiving week from last year, and that was ~ 8500 profiles / day also. So if we were to go from when I wrote this (the initial composition), at 8500 profiles / hour. Continuing at that rate, WT should be expected to hit 22 million around 16:30 Pacific Standard Time (GMT-8) or 19:30 Eastern Standard Time (GMT-5) on December 2nd.

Still,given the data that is available, it would be possible to hit 22 million within either of the pre-December windows suggested above, however it would be very unlikely... probably in the sub-5% bin. I'd be happy to be proven wrong, but I wouldn't want any of my American cousins and neighbours to spend less time with their family over the holiday to do so! 


I keep meaning to do a tracking experiment with greater detail, but to date I've only been loosely tracking the overall trend with timestamped data. Weekends are probably when most of the activity is happening. Something for later.

by anonymous G2G6 Pilot (140k points)
Wow, JN, thank you for the detailed stats!

How close did we get? 22,008,951 profiles right now.

Given the rate you suggest, it had to be gosh darn close!

22,000,148 profiles (6,083,649 with DNA ) : 659,563 . Wed 4/Dec/2019  00:44 GMT )  08:44  Perth

https://www.wikitree.com/g2g/950349/blink-again-in-advent-for-22-000-000-profiles-folks

+10 votes
I’ve added about 300+ profiles this month (all sourced). Onward and upward!
by Pip Sheppard G2G Astronaut (2.7m points)

Related questions

+32 votes
20 answers
1.7k views asked Mar 4, 2019 in The Tree House by Pip Sheppard G2G Astronaut (2.7m points)
+41 votes
5 answers
1.1k views asked Mar 23, 2022 in The Tree House by Pip Sheppard G2G Astronaut (2.7m points)

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

disclaimer - terms - copyright

...