Did you think we currently have 11.5 million people on WikiTree? Nope, it's 10.5 million.

Question

Hi WikiTreers,

As announced here yesterday, Special:Unconnected now enables you to browse all the people on WikiTree who aren't connected to the "big tree."

To enable this feature we had to create a special database of everyone who's unconnected. Just like the database of everyone who is connected, it's rebuilt every day.

Since we're building these databases, it's easy to get counts of how many people are connected and how many people are unconnected.

As shown on Special:Unconnected there are 2.75 million unconnected.

As shown on Special:Connection there are 7.84 million connected.

Hmmm, I got to thinking. 2.75 + 7.84 = 10.6 million. According to our stats, as published on the home page, there are 11.5 million people on WikiTree.

What happened to the other million?

A little investigation turned up the answer: They got merged-away. They were duplicates that got merged and redirected.

I thought we were excluding these redirects from our count, but apparently not. Ugh.

As soon as I realized this we updated the home page from saying "Our tree includes 11,519,313 people contributed by 297,572 genealogists" to "297,575 genealogists from around the world have contributed 11,519,394 people to our shared tree." At least now it's technically correct, if still misleading.

We'll update the actual stats to ignore redirects ASAP.

On the bright side, as some Leaders have already noted, this means WikiTreers have merged almost a million duplicates! That's pretty darn impressive.

And, of course, we're not supposed to be about quantity here. We care about quality. Getting millions of new profiles would be as easy as removing all the restrictions on GEDCOMs. Most genealogists are happy to add their trees. It's merging and collaborating that's hard.

Still, I know it's nice to have a metric of progress, and we've been using the wrong one. I apologize for that.

Chris

Comments

Umm, Chris, I'm sorry to be so dense, but I don't see any difference in meaning - or even nuance - between the old and new statements.

Would it be possible to use the sum of the count of the two searches for Connected and Unconnected as the source for the number of profiles to actually report the correct number?

Answer 1

If we're honest, though, quantity is important, too, because quantity is an indicator of coverage and coverage is one of our goals.

Answer 2

That's actually fascinating in and of itself! I'd wondered just what the rate of duplication has been — especially after having a cousin upload his GEDCOM files resulting in numerous complete duplicate profiles.

That it's only about 10% shows just how well the site works, in particular the tools to check for matches when entering new profiles — in addition to the careful work of those engaged in merging.

It's also important to say this: I also appreciate the integrity that you demonstrate in ensuring that the numbers aren't overstated. That commitment to building and maintaining trust is one of the underlying foundations that makes this site great, and which keeps people involved here.

As a suggestion, perhaps it's best to keep the dual metric posted (maybe not on the front page, but somewhere)? It might give some of the folks trying to connect others to the tree a bit of encouragement.

Thanks for sharing this, Chris!

Answer 3

It's good that people have corrected mistakes. We don't want misleading information. But 10.6 million people is still pretty good!

-JN