Generating Fake Family Trees Automatically

A new scam is on the Internet: a piece of software that will create bogus family trees to be uploaded to your web site. Why would anyone want to do that? The complete answer is a bit convoluted, but the short answer is to boost your web site higher in the search engine ratings so that gullible people will see the ads on your site. In fact, the program's advertising boasts that you can "create unique, non-duplicate content that millions of people search for, AND that neither humans nor search engines can tell is 'real' or not."

FakeFamily.com is very open about its purpose. After purchasing this software for $75, you are advised to create a bogus genealogy site and put affiliate ads on it to defray the expenses. Then you upload fantasy family trees created by the FakeFamily.com software. As the web site says:

Fake Family will generate content that is 100% unique.  The content created is seen by people as being REAL.  There is no dispute about that.  And, SO WILL THE SEARCH ENGINES.

The same web page also says:

Yeah, this is an unusual take on content.  But you need to THINK unusual to beat the engines.  Let this soak in and you'll find that we're speaking the truth.

On another page, the web site states:

Fake Family creates gedcom files that fit the standard gedcom format. This is the file format that allows interfacing with any standard "gedcom to html" program. The thing about a family tree created with Fake Family is that the data all fits. For example, names are "era specific" - meaning, you will get more "Orville" and "Bertha" names in the 1880s than the 1980s. Infant mortality, marriage rates, migration data is also encoded (and more).

You import the GEDCOM file created by the FakeFamily software into almost any genealogy program and tell that program to create HTML files. You then upload those HTML files to your web site that contains advertisements. Google, Yahoo, and the other search engines will eventually discover the site and decide that it is unique content. The site's bogus content will be added to the search engines.

If someone in the future searches for "John Brown" and you have a person on your fake genealogy web site named John Brown, the search engine will point the gullible searcher to your site, where he or she will see your ads. You then hope that they will click on an ad and buy something, and then you get paid. Not a bad scam: using fake data to entice people to buy something!

The owner of FakeFamily.com obviously feels no remorse and even brags:

Which is more important: Verifying that the "mesothelioma" information on your scraped page is "valid" or that some person is really your great great grandpa?  Come on.  I've NEVER seen anyone (certainly not "black hat") have an issue with the validity of any other content.

When I first read the information on this web site, I thought it was someone's idea of a practical joke. After re-reading several times, I have changed my mind. There seems to be no humor and no punch line. Apparently this fast buck artist really wants to flood the Internet with bogus genealogy material, all for the purpose of making easy money.

I tried to see who registered the domain name of FakeFamily.com but found it registered to Domains by Proxy, Inc., a company that specializes in registering domain names by proxy in order to hide the names and addresses of the true owner(s).

I normally take new programs and web sites for a "test drive" and use them myself before writing about them. However, I am not going to do that this time. I'll save my $75.00, the price to obtain the software that generates bogus family trees. If you decide to waste your money on this scheme, please let me know how you make out.

To get a feel for the purpose of this FakeFamily.com site, read this page first: http://fakefamily.com/reply.htm

Then read these pages:

http://www.fakefamily.com/members/fake.php

http://www.fakefamily.com/members/familytree/index.htm

http://www.fakefamily.com/members/

So, let me ask a couple of questions:

1. Are you willing to be known as someone who places bogus information on the Internet?

2. Are you willing to be known as someone who is gullible enough to pay $75 for this software?

Serious genealogists have always preached, "Verify your sources!" If this software gets loose on the Internet, the need to verify every piece of information will be more important than ever.

As always, let the buyer - and the genealogist - beware!

UPDATE: A few hours after posting this article on the daily newsletter web site, I was contacted by one of the developers of FakeFamily.com. As you might imagine, he was not happy with my article. He disputes several of my assumptions and is adamant that this program is no scam. I suggested that he post his explanation for all to read. After an exchange of e-mails, he has done so on his own web site.

There is a new and rather detailed explanation of the purpose of fakefamily.com available now at http://www.FakeFamily.com. I would invite you to read that information now before making any assumptions about this program.

He also pointed out that there is a larger reply written to someone else at: http://www.fakefamily.com/reply2.htm. I would invite you to read that page as well.

Please read both of those pages and then make up your own mind as to the intent of this program.

Thank you.

 

Comments

[Louis Kessler]

I am awestruck.

The explanation on the website and screenshots of the product and the website itself indicate that whoever wrote this program is actually someone with excellent programming skills and a knowledge of internet marketing.

It's too bad that he used his talents to produce this garbage when he could instead have produced something innovative and useful.

I can't imagine anything more tasteless to genealogists than this product.

[Dick Eastman]

UPDATE: A few hours after posting this article on the daily newsletter web site, I was contacted by one of the developers of FakeFamily.com. As you might imagine, he was not happy with my article. He disputes several of my assumptions and is adamant that this program is no scam. I suggested that he post his explanation for all to read. After an exchange of e-mails, he has done so on his own web site.

There is a new and rather detailed explanation of the purpose of fakefamily.com available now at http://www.FakeFamily.com. I would invite you to read that information now before making any assumptions about this program.

He also pointed out that there is a larger reply written to someone else at: http://www.fakefamily.com/reply2.htm. I would invite you to read that page as well.

Please read both of those pages and then make up your own mind as to the intent of this program.

Thank you.

[Chris Dunham]

The Louisiana Secretary of State website shows that Red Belt One, LLC, of Shreveport, LA, was chartered last April:
http://www.sos.louisiana.gov/cgibin?rqstyp=crpdtlC&rqsdta=35913503K

The corporation's officers were listed as Chester T. Kelley and Donald R. Harrold.

The man who wrote those immortal words "Geneology is my main hobby," was almost certainly Don Harrold, who maintains a family website at Harrold.org.

I don't know why Don doesn't offer a link to his own homepage from FakeFamily.com. Could it be that he's trying to hide his honorable business activities from the genealogical community?

If my research is ever slowed by a search engine clogged with fake families, I'll be sure to drop Don a line and thank him.

[fakefamily]

Chris Dunham,

If you spent as much time on finding out the truth about Fake Family as you did researching our corporate entity, you'd see you've got nothing to worry about. But, I don't suppose that's the way these things go.

I wonder how many people are going to take a moment to actually email us or look into the truth?

Sincerely,
Don Harrold

[Jim Terry]

I have sent the particulars of Red Belt One's scam to PayPal asking them to suspend Red Belt One's account.

Jim Terry
Webmaster
Millennia Corporation
www.LegacyFamilyTree.com

[fakefamily]

Jim,

You should be aware that your use of the word "scam" is not an opinion. You are making a legal claim. As such, be prepared to hear from our attorneys if you do not retract or amend your statement.

Sincerely,
Don Harrold

[Chris Dunham]

Mr. Harrold:
It took me about 15 minutes to track your identity. It took me about 5 minutes to learn the truth about FakeFamily.com. There is no excuse for hampering the research of others. (I'm sorry, is "hampering" a legal claim too?)

Mr. Terry:
Thank you for sticking up for decency in genealogy.

[fakefamily]

Chris Dunham,

It should not have taken THAT long. We don't hide who we are. We have a message forum you can post in. Our email address is posted at the site, too. Of course you didn't bother with that and you are obviously ignoring the facts about what our software does.

Fake Family does not "hamper" anyone's research: I'll post a project created with Fake Family. 1000 people in it. You find even one couple in there that's got a counter part in the REAL WORLD (Where first and last names match, and the birth states of both match, and where the birth dates are correct) and I'll retract every word I've written and make our home page read, "Chris Dunham Was Right", and stop any sales of Fake Family.

If you can't find a match, you must post here a retraction and admission that YOU are wrong about our software. That is does NOT hamper research and that you could find no evidence of a real-world connection with anything created with the software. I mean, I'm giving you a golden opportunity to prove how diabolical this software is.

Are you willing to take this challenge?

My guess is, probably not.

[Andrew Millard]

The producers of this software claim astronomical odds against it matching any real person's details exactly. My genealogy data isn't like that. It's full of holes, ambiguous dates (and names), and in some cases data collected that looked like it might be relevant so ought not be ignored.

Let's take one example from the sample fake tree provided: Susan Linda Edwards b.6 Dec 1949, Lewellen, NE. SSDI tells me that there was a real Susan L Edwards b.14 Jun 1948, with a Social Security number issued in Ohio. Now imagine that I have a tale from an elderly relative of such a cousin Susan born "a few years after the war, but the family moved around and I can't remember where they lived at the time". How am I to distinguish the real from the fake one? From this side of the Atlantic I have to employ a researcher in both places to look at local records, and the misinformation from Fake Family has doubled my research costs.

Because the fake data produced are realistic, humans will believe this data and spend money trying to verify it. The odds are also not as high as claimed for a match because Fake Family is not selecting names at random, but using the frequencies of the period in question. It is claimed that selecting from 1000 surnames and 5000 surnames yields odds of 1 in 5 million of matching the name within the decade. Not only is Susan Linda Edwards matched, but John William Williams b 1940 in the sample fake tree, there are 6 possible matches *to the year* in SSDI for John W Williams. I've not checked all the names in the fake tree, but 2 out of 104 matching forename and middle initial within the decade is looking a long way from 1 in 5 million to me.

[fakefamily]

Andrew,

The example you give proves my point: The ONLY connection between the two people (the two Susans) is their NAME. The dates don't match. The places don't match. You've not addressed the issue of spouses, children, parents or how anyone in YOUR family could connect to these names.

Regarding John William Williams, again, the only thing you'll find is a match of a name. And, you've only got a middle initial with the SSDI record.

Your imaginary tale of an elderly relative passing information to you is precisely why you do NOT take ANY data at face-value. It also neglects the main facet of your research: Why are you looking for a Susan L. Edwards in the first place?

If you know ANYTHING more than a name you will find that the information in the Fake Family data is NOT what you want. If you do NOT know anything more than a name, ANY data that's "close" you could claim a match from. But, that still does not tie the person to YOUR TREE.

All you've got are some names that show a similarity to some names. Names to NAMES. And, you don't even have the middle names from the SSDI records. JUST MIDDLE INITIALS.

You are reaching for a reason to argue this point. You'll get quite a few people who reply, "yeah!" or, "thanks for shining the light of truth on this evil software". Meanwhile, you've only - in truth - illustrated the point that Fake Family does NOT create REAL family trees. It creates fictitious name groupings from a random data set.

Your conclusion that, "2 out of 104" is not correct. What you've got are 2 created NAMES that are CLOSE to some names in the "real world". On every other count there is NO correlation. Besides all that is the fact that you would have SOME reason to connect a person to your tree. Absent other information all you've got are names that look similar.

I do not dispute that at some time in the last 120 years someone named John W. Williams probably lived. But there are also hundreds of people named George Bush in the world, too - both past and present. But they are not all Presidents. You need a little more than a name and a decade to prove someone is a part of your family tree.

My challenge remains.

[fakefamily]

Andrew,

I also challenge your assertion that we claim "astronomical odds" for matching "exact" details.

The odds are 5,000,000 to one of getting a name to match. That's mitigated by the fact that you'll have many names in your tree from which to draw, obviously. So, let's say you've got 5000 names and the odds of any one Fake Family tree project creating a match of First and Last names becomes 1 in 1000. But, that is ONLY A NAME. And does not include middle names. If you throw in middle names you are back to 1 in a million.

But, to be conservative, I'll stick to the VERY conservative 1 in 1000 of matching only a first and last name of one person. Now, you have to have SOME other data from which to make your conclusion that this person is related to you somehow. Take your pick: By birth date? By birth place? By spouse? By parents? By siblings? By death date? By marriage date and place?

Take any of these. Let's say, by spouse. Let's even say all you know is a spouses first name. That's still 1000 times 1000 which is 1 in a million that you'll get a match with a spouse.

I'll even cut that number in half. Let's call it one in 500,000.

So, what you COULD get is one spouse with first and last name and one spouse with only a first name. No dates. No places. No children. No parents. No marriage information.

That's called: NOTHING. I'm being super conservative here and 1 in 500,000 still leaves you with NOTHING if you don't have other data on your end to corroborate.

Let's just throw in the date of marriage. Fake Family will choose from 120 years of data (by default). So, take your 1 in 500,000 and multiply by 120. You are up to 1 in 60,000,000 now that you get two spouses (one with only a first name) and a YEAR of marriage.

1 in 60 MILLION.

I've given very generous odds. And we still have not connected these fake people to ANYONE in your tree. If you begin to ask how these folks are even connected to you, you find that if you go to parents or children or siblings of either spouse you get odds that now reach the stratosphere.

Look, I'm not saying you will never find a common name in a Fake Family project. What I am saying is that the leap from "Hey, I found a name" to "This person is in my family" is a HUGE gulf that Fake Family (by design) just cannot cross.

[fakefamily]

NOTE: All this is truly a moot point since we do not actively promote Fake Family, nor do we even sell it to the general public. There are so few people that even know about it that this is all kind of a - well - moot point. I would venture to say that we've gotten more publicity from this blog than anything we could have (or would have) done.

[Andrew Millard]

Fakefamily wrote "The example you give proves my point: The ONLY connection between the two people (the two Susans) is their NAME. The dates don't match. The places don't match. You've not addressed the issue of spouses, children, parents or how anyone in YOUR family could connect to these names."

My point is that I am initially unlikely to have all that information. All I know about some of my relatives is that they emigrated to the USA. I have a name and birth year from UK records but that is all. The fake information you generate is plausible and given the way it is constructed is likely to generate matches to vague information such as I have. If I find a match to vague information in a line I am pursuing, then I will spend money to verify it. How much is it going to cost me to hire a researcher to check state birth records for someone who doesn't exist? If there's no birth registration to match your fake data, but it all looks plausible (after all there's a whole tree that looks like someone researched it thoroughly, so perhaps it was just a mistake in the birth information) perhaps I try searching elsewhere for that person, marriage and death records might be searched at more cost. I would have to spend a lot of money to convince myself that *one* person on a Fake Family tree was entirely fictional, and whilst I have two possible matches to my initial vague information my research is held up. There's no point pursuing the family of the Susan L Edwards born in Ohio if there's a distinct possibility I should be chasing one in Nebraska to find my relatives. So Fake Family may well cost me money to invalidate false information. As you say 'the leap from "Hey, I found a name" to "This person is in my family" is a HUGE gulf', but it will cost me to attempt to bridge that gulf. The odds of matching a single name and details in my tree may be low, but given the number of genealogists out there the chances that a good number of people get vague matches and spend money pursuing them must be reasonably high.

As to what this type of fake data does for those of us pursuing one-name studies I hate to think. Then it is only the surname that need match for the data to be relevant to what we are researching.

Fake family also wrote 'I also challenge your assertion that we claim "astronomical odds" for matching "exact" details.' and followed that with an argument that concluded 'you get odds that now reach the stratosphere.' What's the difference between astronomical and stratrospheric odds?

You may not promote or sell Fake Family to the general public, but it looks like a fairly sophisticated piece of software, and at $75 apiece you must need to sell quite a few to recoup costs. Users are each likely to create a number of trees to attract traffic. Perhaps there will soon be 100000 people on Fake Family trees. Each person generated increases the chances of a match to my tree.

[Jim Terry]

Don,

I used "scam" in the classic dictionary sense "A fraudulent business scheme." You admittedly generate fake genealogies for making a quick buck.

By your own words "The content created is seen by people as being REAL. There is no dispute about that. And, SO WILL THE SEARCH ENGINES." You also misused our software to help promote your questionable commercial venture without the permission of Millennia Corporation.

But all we hear from you are rationalizations and threats: "As such, be prepared to hear from our attorneys if you do not retract or amend your statement."

Jim Terry
Webmaster
Millennia Corporation
www.LegacyFamilyTree.com

[fakefamily]

Andrew,

You say, "My point is that I am initially unlikely to have all that information. All I know about some of my relatives is that they emigrated to the USA. I have a name and birth year from UK records but that is all. The fake information you generate is plausible and given the way it is constructed is likely to generate matches to vague information such as I have."

Again, thank you for your considerate replies and questions.

However, I grant you might not have all that information. But you must have SOMETHING besides a name. Once you move to something besides a name the data created with Fake Family distances itself from your tree in remarkable fashion.

The part about this that I'm still hoping you'll see is that just a name is not enough. You need other data to corroborate that name. And, I don't mean from Fake Family. I mean from YOUR research.

You are not researching names in a vacuum. You why you are researching the people in your tree. Those names will be associated with SOME other data. If not you may as well research some one else's tree for which you have no knowledge.

That "other data" from your tree is what separates your family from anything created with Fake Family.

In summary, I want you to know that I appreciate your comments and your tone. I do not enjoy being attacked, flamed, and made-fun-of. Your posts are engaging without rancor and I thank you for that.

One more thing, regarding our cost: I can promise you we have lost big time money on the development of it. With no plans to market it to the masses, those losses show no sign of being recouped anytime soon. That may sound like an odd business model, but it's just the truth. And, let me be even more clear: By "any time soon" I mean, more literally, that we have no plans to release it other than to the very small number of folks who have access to it now. A point I made earlier is still true: We don't advertise it and the traffic we've gained from this blog is more than we would have achieved under any circumstances with our own marketing efforts (which are basically non-existant).

Whoever made this information available to Dick Easton (and, Mr. Easton himself, actually) has had the opposite effect from what they desired. Instead of less attention and less people interested we've got MORE of both.

However, neither of those will weigh in on my decision to keep Fake Family a limited-release software. And, that's if we sell licenses to anyone else ANYHOW.

On a side note: You would probably find it fun to see the way it works. It's pretty interesting watching generations of fictitious folks migrate across the US...

[fakefamily]

Mr. Terry,

A) You chose to air something publicly which was best suited in an email. We immediately removed the images (although under the "Fair Use" clause of US copyright law we are not compelled to do so, by the way). A simple email would have sufficed.

B) There you go again: Talking about stuff you don't have a clue about. We are not a fraudulent business scheme. Well, unless you mean we are "fraudulent business scheme" like Mormons are a "cult". I don't suppose Mormons like being referred to as a "cult". Wait, check that, I KNOW they don't like it. Well, lots of people have that OPINION but if you ask the Mormons I know they will tell you it's a FALSE opinion. Well, if you'd spend less time making libelous comments and more time looking at the facts you'd see that our corporations are certainly not "fraudulent" nor is our software a "scheme".

We DON'T EVEN ADVERTISE IT. This blog brought to your attention something that is not an issue. Certainly not at the time of this blog posting.

Regarding this supposed "misuse" of your software: I certainly did not know that it was a "misuse" to show folks how to use Legacy. We profit ZERO from telling anyone about Legacy. If you don't want us to show images of Legacy, we will absolutely honor your request. But that is not due to any law it is due to our NON-FRAUDULENT business model.

I must say that it would be nice if - every now and then - people would check with folks before they start airing publicly what is a business matter between two corporations. I can't imagine posting on a public blog a disagreement (whether real or imagined) I had with another corporate entity.

Finally, it's also sad to see you boil down my comments as "rationalizations". I guess one man's "rationalizations" are another's "reasoned replies to baseless attacks".

I choose the latter.

[fakefamily]

Mr. Terry,

Wait, one more thing: We do not claim to "admittedly generate fake genealogies for making a quick buck."

Those are your words that come from your world view. However, they do not represent what we do or believe.

Your take on it may be that we are hucksters of some sort or another, but your take is not "truth". We neither admit to what you posit nor do we ever look to make a "quick buck".

You see, to make a "quick buck" implies that we are actively either using the software or selling it in any number. We do neither with any degree of urgency.

Your opinion, notwithstanding.

Sincerely,
Don Harrold

[fakefamily]

Andrew,

I just noticed something. You show that a real Susan L. Edwards shows up in the SSDI, right? In fact, the Susan Linda Edwards in the Fake Family project would have died as "Susan Linda Morales". The SSDI record you would want to find would be Susan L. Morales. However, the Fake Family Susan Linda MORALES died in 1974 in Nebraska. SSDI records do NOT show a Susan Morales who died in either 1974, nor in Nebraska.

So, what you've got in the case of Susan L. Edwards MORALES is a complete miss in SSDI.

Sincerely,
Don Harrold

[Dino (All Dino, All the Time)]

Much Ado About Nothing ...

Big deal, someone actually admits to creating 'fake' family trees. These actions are no worse than Family Tree Maker's World Family Tree CD's (and let's remember they SELL those) which are loaded with unverified, totallly false info.

Or no worse than companies that sell websites and CD's that feature genealogies with links to fictional kings and queens.

Or the ones that sell lineages that go back to Adam and Eve.

Any competent genealogist will insist on verification of the facts before adding these people to their family trees. They won't be able to do that, so they'll discard this info. Name collectors, who just don't care, will add them.

Although I'm not in favor of what this form is doing, it may be a good thing for genealogy in the long run if it encourages people to start posting their sources (which fakefamily won't have.)

Come to think of it, I seem to recall an entire family tree for the Wheeler family from the show Into The West posted on the TNN site with no objections from the genealogy community. This was a fictional family, much like fakefamily's families.

There's so much (expletive deleted) on genealogy family websites now that fakefamily's efforts probably won't even raise the level of background noise.

[Mark Roy]

The fact that the guy is generating bogus trees is just part of the problem. The purpose of the product is produce information (which will *never* be useful) in order to lure people to a web site in hopes they'll click on an ad.

Anyone who creates a business or product based on duping people is a slimeball, IN MY OPINION (so don't sue me!)

[Chris Dunham]

Here are just two ways this product could hamper my research. If I were starting a single-name surname project, I might Google the surname to find new leads. FakeFamily could produce an entire family line using that surname, the existence of which I would then have to disprove. The fake family websites might even crowd authentic websites out of the first few Google SERPs.

Or suppose I'm writing a town history, and FakeFamily happens to assign my hometown to a fictional family. Again, I would have to waste my time proving it false.

However much Mr. Harrold tries to rationalize his rationalizations, the end-result of his little enterprise is to make it harder for me to do my job.

I think I'll go over to GenForum, and start posting made-up answers to people's queries. The good genealogists should be able to prove them false.

[Bobbi Deo]

I contacted my State and Federal legislators. If this is not a crime it should be made one. I suggest that other readers do the same.

Bobbi

[Dino (All Dino, All the Time)]

You're right Chris ... the good (or should I say real) genealogists would be able to separate the wheat from the chaff.

That's part of being a genealogist. No genealogist worth their salt would ever put anything into their database without confirmation of the facts or without a big, fat warning in the notes.

If someone is starting a single name project, I would certainly hope that they are verifying information they are putting into the project. If they aren't, they're just name collectors offering nothing of value to the genealogical community.

And if sonone is doing a town history, what are the chances that they would be fooled into thinking that someone that appears in no other record is worthy of inclusion.

The Genealogical Proof Standard has been codified for this very reason. As a reminder, the GPS consists of five elements:
1. a reasonably exhaustive search;
2. complete and accurate source citations;
3. analysis and correlation of the collected information;
4. resolution of any conflicting evidence; and
5. a soundly reasoned, coherently written conclusion.

Any of the fakefamily data would not pass point 1. Lack of source citations (point 2) should be another red flag. Steps 3 and 4 would quickly eliminate and fakefamily data from consideration and inclusion.

As for the authentic websites being crowded out of the first few Google SERPs (whatever those are), isn't that wat GPS point 1 (a reasonably exhaustive search) is all about?

I do not want it to look like I am defending Fakefamily. I'm not. But I also don't think that this data will fool many genealogists. They're just playing the search engine game and may have come up with a tool that will work for a few months until the search engine programmers leapfrog back over them. Then they'll be on to some other idea.

[Dino (All Dino, All the Time)]

Bobbi,

Are you suggesting that creating works of fiction should be a crime? If so, you'll have to go after everyone that has submitted and unresearched, undocumented family tree to Family Tree Make and had it published and sold on one of their WFT CD's.

[Chris Dunham]

Dino, my point is not that good genealogists will be fooled into incorporating phony info into their work, but that they will be slowed.

I get plenty of leads from lousy sources. Some pan out, and many don't. My objection to the FakeFamily scheme is that it's a nuisance, not that it's the end of genealogy as we know it.

(Note that FakeFamily uses dates only since 1880, and that leads from family-history websites are generally more reliable the more recent the event.)

You're absolutely right that, if it catches on, the Google engineers will take action. Unfortunately, many good family websites will be caught in the crossfire.