When the Internet Was Wrong: How DNA, Census Math, and Bad Trees Helped Me Rebuild a Lost Family

Genealogy will humble you fast.

You start with one innocent little question like, “Who were Matilda Weasner’s parents?” and suddenly you are neck-deep in 1810 census tick marks, misspelled surnames, Canadian militia records, cousin marriages, Ancestry ThruLines hallucinations, dead ends, family legends, bad FindAGrave merges, and DNA matches that are somehow both screaming the answer and refusing to speak in complete sentences.

So, you know. A normal hobby.

But over the last stretch of research, I finally had one of those moments genealogists live for: the records, the DNA, the geography, and the social history all started clicking together.

Not perfectly. Not neatly. Never neatly.

But enough that the fog began to lift.

This post is not a final proof argument. It is a research narrative and working methodology: how I used DNA triangulation, timeline math, FAN-club research, and even deeply flawed online family trees to reconstruct several tangled family lines connected to my McDonald, Weasner/Wisner, Ensign, Sturges, and related Niagara/Upper Canada ancestors.

And honestly? The biggest lesson was this:

The internet is not the tree.

The tree is the evidence.

I Stopped Letting Online Trees Lead

For years, I made the mistake most of us make at some point. I looked at online trees and assumed that if enough people repeated the same parents, the same spouse, the same birthplace, the same “everybody seems to have this,” then maybe there was something behind it.

Sometimes there is.

Sometimes there is absolutely not.

Sometimes the online tree is a flaming pile of copied errors, merged identities, half-read documents, and algorithmic nonsense.

But here’s the trick: bad trees are not useless.

They are compost.

I know that sounds dramatic, but I mean it. Bad trees often contain little scraps of truth buried under five layers of genealogical garbage. A wrong parent may have a correct will attached. A mangled family may include a land deed nobody else bothered to read. A completely impossible timeline may still point toward the right neighborhood, the right surname cluster, or the right migration pattern.

So I stopped asking, “Is this tree right?”

I started asking, “What source did this tree accidentally preserve?”

That shift changed everything.

DNA Became the Compass

The breakthrough came when I stopped waiting for the paper trail to politely name everyone in a clean, modern way.

Because frontier families did not leave clean records.

The Weasner/Wisner family certainly did not.

Their surname was spelled every way imaginable: Weasner, Wisner, Wesner, Wysner, Wiesner, Misner, and probably several other versions I have not even found yet. They moved across the Niagara frontier, lived in poor or working-class communities, crossed between New York and Upper Canada, and left behind records that were fragmented, misspelled, or missing entirely.

So instead of letting the paper trail dictate the family structure, I let the DNA point me toward the paper trail.

That meant manually clustering DNA matches across descendant lines.

For example, in the McDonald/Ensign research, I noticed that descendants of Abel, Samuel, and James McDonald all shared DNA with descendants of Temperance Mary Ensign and the Gilbert family. That was not just a vague hint. That was a biological pattern.

Even when Ancestry’s ThruLines could not help because of its generational limits, the match matrix still could.

When multiple descendants of different suspected siblings match descendants of the same older family network, that is not random. It does not automatically prove every relationship by itself, but it tells you where to dig.

DNA became the compass.

The records became the map.

My Direct Maternal Line: Matilda Weasner McDonald

The emotional center of this research, for me, is Matilda Weasner.

Matilda married Abraham McDonald and later Daniel Brown. She is part of my direct maternal line, which makes this more than an abstract surname puzzle. I can trace my mother to my grandmother, to Viola, to Eliza, to Matilda, and then, if this reconstruction is correct, back to Matilda’s mother: the woman recorded in later documents as Clara, but who may actually have been Sarah, Sara, Eliza, or another phonetically similar name.

That woman is the next frontier.

And I mean that in both the genealogical and spiritual sense.

Because when you are following a direct maternal line, you are not just collecting names. You are following the unbroken chain of women whose bodies made yours possible.

No pressure, right?

Reconstructing the Lost Weasner Sibling Group

No single document has appeared so far naming all the children of Jacob Weasner/Wisner and his wife.

So I had to reconstruct the sibling group through a combination of DNA, census placement, migration patterns, marriage networks, and repeated neighbor associations.

The working sibling group currently includes:

• Daniel Weasner, born around 1812 to 1815, later connected to Crowland, Ontario and Niagara County, New York.
• Margaret Weasner, who married Enoch Stewart.
• Jacob Wesner Jr., born around 1817, who appears near related family in the 1851 census.
• Jane Wisener, who married Warren Goff/Gough.
• Andrew Wysner/Wisner, who married Patience Sloat and later was killed by his brother-in-law in Sanilac County, Michigan.
• Matilda Wisner, who married Abraham McDonald and later Daniel Brown.

The key is that descendants of these lines do not just vaguely appear in the same online trees. They share DNA with one another in patterns consistent with descent from the same ancestral couple.

That matters.

A lot.

Especially because this family also has pedigree collapse, which is the polite genealogical term for “the frontier was small, everybody knew everybody, and yes, cousins married cousins.”

Pedigree Collapse: Annoying, Confusing, and Weirdly Useful

This Weasner/Goff/Stewart/McDonald network contains several cousin marriages.

For example, Matilda Wisener, daughter of Jacob Jr., married Joseph Warren Goff, son of Jane Wisner Goff. That appears to be a first-cousin marriage.

John Henry Weasner, son of Daniel, married Esther Ann Goff, Jane’s granddaughter, creating another close loop (first cousins once removed).

Later, descendants of Jane and Margaret’s lines also intermarried (but at the fourth cousin level- not sure they were even aware of it at the time).

Andrew’s son married the daughter of his shooter (first cousin marriage).

To a genealogy algorithm, this is a nightmare.

To a human researcher, it is both a warning and a gift.

The warning is that DNA matches may look closer than they really are. ThruLines may confidently attach someone to the wrong ancestor. Shared centimorgans may be inflated. A match who appears to descend from one sibling may actually be inheriting DNA from multiple branches of the same ancestral couple.

The gift is that this same endogamy can preserve an ancestral genetic signature more strongly than expected.

In other words: the cousin marriages made the software messy, but they also amplified the family signal.

That is how DNA from Jacob and his wife may still be visible across multiple descendant lines more than 200 years later.

The “Clara Hicks” Problem

Many online trees identify Jacob Weasner’s wife as Clara Hicks.

I do not believe that is correct.

The “Clara” name appears to come primarily through a later marriage record for Matilda, where the clerk also mangled Matilda’s maiden name. If “Weasner” could become “Misner,” then I am not going to treat “Clara” as gospel without supporting evidence.

The online attribution to John Hicks and Elizabeth Eady appears to conflate Jacob’s wife with a woman born in England in 1814, which is biologically impossible if Jacob’s oldest children were born around 1812 and 1813.

That is not a small mistake.

That is a chainsaw-to-the-branch mistake.

The naming patterns also give me pause. In the next generation, I see names like Sarah, Elizabeth, Mary, and Ann recurring. I do not see the name Clara echoed in the way I would expect if Clara were truly the honored matriarch of this family.

So for now, I am treating “Clara” as a possible transcription or phonetic error, not a proven identity.

My current working hypothesis is that Jacob’s wife may have been a Sarah/Sara or possibly an Eliza/Elizabeth-type name, and that she may belong to the Fairfield County, Connecticut Sturges/Morehouse family network.

The Sturges Hypothesis

This is where the research gets especially interesting.

Autosomal DNA appears to connect the Weasner sibling group not only to the Wisner line of Orange County, New York, but also to the Sturges and Morehouse families of Fairfield County, Connecticut.

My current working hypothesis is that Jacob Weasner’s wife may have been a daughter of Ezekiel Sturges and Hannah Morehouse of the Weston/Norwalk, Connecticut area.

The evidence is not “one shaky online tree says so.”

The evidence is a pattern.

A DNA match on a documented Sturges/Lockwood line. Additional matches to descendants of Canadian Daniel Sturges. Shared matches across multiple Weasner sibling descendants. Census math in Connecticut that leaves room for an unidentified daughter of the right age. Migration patterns from Fairfield County into the Niagara frontier. And then the later clustering of Weasners, Sturgeses, Stewarts, and Goffs in Upper Canada.

Again: not a final proof argument yet.

But it is no longer random.

It is a working hypothesis with biological, historical, and geographic support.

Canadian Daniel Sturges: Probably Not Who the Internet Says He Is

Another major issue is the long-standing claim that Canadian Daniel Sturges of Virgil, Ontario was the son of Daniel “Freezing Daniel” Sturges.

I do not think the DNA supports that.

If Canadian Daniel were the son of “Freezing Daniel,” then his descendants would be much more distantly related to the Weasner/Sturges-descended matches I am seeing. The amount of shared DNA showing up across multiple Weasner sibling lines is too strong for that distant relationship to make sense as the primary explanation.

The census math also raises questions.

Canadian Daniel, born in 1781, fits surprisingly well in the household of Ezekiel Sturges in the 1790 and 1800 censuses. There is also a suggestive timeline: Ezekiel married Hannah Morehouse shortly after Daniel’s birth.

That does not prove the case by itself.

But combined with the DNA, it makes Canadian Daniel look much more likely to be closely related to the suspected Weasner matriarch — possibly even her brother — than a distant cousin through “Freezing Daniel.”

This is exactly why genealogy requires both biology and math.

The story may be colorful.

The timeline still has to work.

Timeline Math Is Ruthless

One of the most useful tools in this whole process has been basic timeline math.

Not fancy.

Not glamorous.

Just asking: could this person actually be the parent, spouse, sibling, or child that the tree claims?

If someone born in 1816 is listed as the father of someone born in 1826, no.

No, thank you.

That is a ten-year-old father.

We are not doing that.

If a “Peter McDonald” looks geographically tempting but his tombstone says he was a native of Drymen, Scotland, then he is not my Peter if my evidence points elsewhere.

If twins are born in 1891 and appear in the 1892 New York State Census under circumstances that point to one biological father, then later family storytelling or FindAGrave assumptions need to be challenged.

Timeline math is not optional.

It is the bouncer at the door.

The FAN Club Was Not Background Noise

The other major lesson: neighbors matter.

A lot.

Our ancestors did not move through the world as isolated individuals. They migrated in clusters. They married neighbors. They witnessed deeds. They appeared beside in-laws in census pages. They borrowed, bought, worshiped, worked, fought, and survived with the same recurring families.

That is the FAN club: Friends, Associates, and Neighbors.

In this research, families like the Dickhouts/Deckhouts, Sensabaughs, Pyles, Furrys, Minors, Stewarts, Goffs, Sturgeses, and others were not background characters.

They were evidence.

When I found Hiram and Ensign McDonald living near the Dickhout family in later censuses, that opened the door to the land and marriage network that confirmed two McDonald half-brothers married two Dickhout sisters.

That is not trivia.

That is structure.

The neighborhood was the record.

Mining Bad Trees Without Letting Them Infect Mine

Some of the most important clues came from trees that were, structurally speaking, a mess.

One tree had the wrong Peter McDonald and a scrambled timeline, but it contained a manuscript that helped clarify Amy Johnson’s parentage and pointed toward the Niagara District Loyalist network.

Another online tangle involving multiple Jonathan Ensigns had people mashed together incorrectly, but the attached documents mattered. A will. A land deed. A “sole heir” clause.

That is the whole trick.

Do not copy the tree.

Read the documents.

The person who built the tree may have misunderstood what they found. The algorithm may have merged three men with the same name into one genealogical Frankenstein. But a will is still a will. A deed is still a deed. A census page is still a census page.

The garbage pile may contain gold.

Just wear gloves.

Building Down Instead of Only Back

Another strategy that helped was building downward.

Most people build their trees backward: parent, grandparent, great-grandparent, and so on.

That is necessary, of course.

But with DNA, you often have to build down from suspected ancestors through all their children and grandchildren into the 20th century. That creates what I think of as a wide-net trap.

If I trace collateral lines forward — siblings, half-siblings, daughters who married and disappeared into new surnames, estranged branches, messy second marriages — then my tree is ready when a modern descendant tests.

The algorithm may not know what to do with a fifth-great-grandparent.

But if I have built enough descendant pathways, a new match can suddenly illuminate the whole structure.

That is how you beat the limitations of ThruLines.

You do not wait for the algorithm to hand you the answer.

You build the net.

The Maternal Line and mtDNA: The Next Frontier

The next stage of this research is Clara/Sarah/possibly-Eliza, the wife of Jacob Weasner and mother of this reconstructed sibling group.

This is where mtDNA may become especially useful.

Because I descend through an unbroken female line from Matilda, I carry the mitochondrial DNA of that maternal line. My current mtDNA haplogroup is V3b, which appears to be relatively uncommon.

If I can identify documented direct female-line descendants of Matilda’s suspected sisters — especially Margaret or Jane — and if they also carry V3b, that would be powerful supporting evidence that these women shared the same mother.

It would not automatically identify that mother.

It would not, by itself, prove whether she was a Sturges, Morehouse, Misener, or someone else.

But it would give us an independent biological test of the reconstructed sibling group.

And honestly, that feels huge.

Because Clara/Sarah is not just a missing name.

She is the woman standing at the edge of the paper trail, holding the next door closed.

What This Research Has Taught Me

This process has changed how I do genealogy.

I no longer treat online trees as answers.

I treat them as leads.

I no longer treat DNA matches as vague cousin confetti.

I cluster them.

I no longer accept a record because the name looks right.

I do the timeline math.

I no longer ignore neighbors.

I map them.

I no longer expect one perfect document to solve the whole thing.

I build the case from fragments: DNA, deeds, censuses, wills, naming patterns, migration routes, land records, marriages, and the quiet persistence of families living beside each other decade after decade.

And maybe most importantly, I have learned to be comfortable with the phrase:

“This is my current working hypothesis.”

That is not weakness.

That is honest genealogy.

Current Working Conclusions

Based on the evidence gathered so far, I believe the following are strong working conclusions:

Jacob Weasner/Wisner and his wife were likely the parents of Daniel, Margaret, Jacob Jr., Jane, Andrew, and Matilda Weasner/Wisner.

Matilda Weasner McDonald belongs biologically within that sibling group, even though no single known paper record names her parents clearly.

The “Clara Hicks” identity repeated in many online trees is likely incorrect or at least unsupported.

Jacob Weasner may be connected to the Orange County, New York Wisner family, possibly through David Wisner and Sarah “Sally” Blain, though this still needs more deed, tax, and DNA work.

Jacob’s wife may belong to the Fairfield County Sturges/Morehouse network, possibly as a daughter of Ezekiel Sturges and Hannah Morehouse, but this remains an active research hypothesis.

Canadian Daniel Sturges may be much more closely related to this suspected matriarch than commonly stated online, and may not belong under Daniel “Freezing Daniel” Sturges as many trees claim.

The DNA, FAN-club evidence, and migration patterns all point toward a tightly connected Niagara/Upper Canada kinship network that has been badly distorted by online tree copying, surname variation, and algorithmic overconfidence.

Collaboration Welcome

If you descend from any of these families — Weasner, Wisner, Wesner, Wysner, McDonald, Goff/Gough, Stewart/Steward, Sturges/Sturgis, Morehouse, Lockwood, Ensign, Gilbert, Dickhout/Deckhout, or related Niagara/Haldimand/Lincoln County families — I would love to compare notes.

I am especially interested in:

• DNA matches connected to the Weasner/Wisner sibling group
• Female-line descendants of Matilda’s suspected sisters for mtDNA comparison
• Primary records connected to Jacob Weasner/Wisner in Niagara or Upper Canada
• Land, tax, probate, or militia records connected to the Niagara frontier families
• Sturges/Morehouse/Lockwood migration evidence from Connecticut into Canada

And if you think I am wrong, please tell me.

Seriously.

I am not trying to protect a pretty family story. I am trying to find the truth.

Even if the truth is messier.

Especially if the truth is messier.

Because that is where the real ancestors usually are.

Find my family tree on:

MyHeritage Family Tree

Ancestry Family Tree