Large trees - is this a record?

[Helen7]

While searching through my DNA matches on Ancestry (searching by birth location for ancestors from a village in Bedfordshire), I came across one with a public tree containing over 274,500 people! I was astounded at this number, over 10x more than I’d seen in any individual on-line tree before, and it got me wondering if anyone has seen an individual tree bigger than this on Ancestry or any other site? (By individual I mean just created/managed by one person, so excluding the combined ‘One World’ type tree advocated by familysearch).

Perhaps surprisingly, I was able to locate my (distant) relatives quite easily in this huge tree (despite a common surname, the village location pinpointed them very well), but when I looked at this branch in detail, I discovered a reason for the large number of people. According to this tree, my 5x great-grandfather William married 4 women, all called Jane, 3 of whom were born after William died, and had 40 children over a 130-year period, born in various parts of Canada, Scotland and Bedfordshire. Despite William dying in 1784 (at least that date is correct), he is listed as living (with one Jane or other) in various places in Scotland and Canada in 1851-1901 (sometimes in different places at the same time). There are also at least 4 clones of the same William in the family. This looks like an extreme example of the tree owner accepting every hint offered by Ancestry without checking if they make sense, and ending up with a nonsensical collection of duplicate names and date/location errors. I haven’t looked at other branches, but if this branch is typical of the whole tree, it is easy to see how such a huge number of ‘people’ can result!

 

[peter]

As I've pointed out many times, because name collectors have the largest trees they turn up more frequently in search results. For example, if the average Ancestry tree has 1000 relatives, that enormous 'tree' is probably going to show up in search results about 275 times as often.

The only time I've ever corresponded with someone who had a very large was on Genes Reunited (many years ago) where the other person refused to answer any of my questions until I provided the name of my ancestor's mother and father.

Since the ancestor concerned was illegitimate, and the father's name unknown at that time (it still is) I could never meet the requirements, and she cut me off. She later disappeared from GR - maybe she joined Ancestry instead!

 

[Helen7]

As I've pointed out many times, because name collectors have the largest trees they turn up more frequently in search results. For example, if the average Ancestry tree has 1000 relatives, that enormous 'tree' is probably going to show up in search results about 275 times as often.

I agree about large trees turning up more often in general searches for relatives (and certainly in the Ancestry 'hints' from public trees), but I don't believe DNA matches would show this bias, as the DNA match itself does not rely on having a tree at all. Most of my DNA matches have no trees, or minimal ones, so this one stood out!

The tree owner has been an Ancestry member since 2010 and is from USA (and logged in yesterday so a current user), so unlikely to have been on GR as I thought that was a largely UK site?

I did wonder about contacting her (yes it is a woman, using what I assume is her real name) and offering to help her tidy up her tree but a) it would be a mammoth task which I don't have time for, and b) she would probably ignore me, as it would mean deleting thousands of people from her potentially record-breaking tree!

 

[peter]

I don't believe DNA matches would show this bias

They will if you're following the strategies in the Masterclass.

I did wonder about contacting her (yes it is a woman, using what I assume is her real name) and offering to help her tidy up her tree but a) it would be a mammoth task which I don't have time for, and b) she would probably ignore me, as it would mean deleting thousands of people from her potentially record-breaking tree!

Rather you than me!

 

[Helen7]

They will if you're following the strategies in the Masterclass.

Ah yes, I see what you mean. I was actually following your strategies in the Masterclass, and having got too many spurious surname hits, I switched to specific location. Of course the 'no tree' brigade are eliminated from the results list, and I can see that the bigger the tree the more likely they are to produce a hit for name or location. I'd still be interested to hear if anyone has come across a larger tree (whether DNA match or not), albeit with many duplicates and false information!

One thing I did notice was that if I specified both surname and location simultaneously in the search, I got 'No matches found', even though this big tree contains both so should have come up. In fact, I've found this to be a general problem for me in searches of my DNA matches: a search for both surname and location (selected from Ancestry's drop-down list) invariably produces no matches, even when I know there should be some (found when searching for either name or location but not both). Any idea why this should be?

Rather you than me!

I agree, not worth the bother.

 

[peter]

One thing I did notice was that if I specified both surname and location simultaneously in the search, I got 'No matches found', even though this big tree contains both so should have come up. In fact, I've found this to be a general problem for me in searches of my DNA matches: a search for both surname and location (selected from Ancestry's drop-down list) invariably produces no matches, even when I know there should be some (found when searching for either name or location but not both). Any idea why this should be?

There's a problem with locations generally - Ancestry only recognise place names that are in their list. When someone uploads a GEDCOM there will probably be numerous birthplaces that Ancestry don't recognise, or identify incorrectly. For example, my ancestor born in Cornwall St, London, is regarded as having been born in Cornwall. Others get relocated to the US.

 

[Helen7]

There's a problem with locations generally - Ancestry only recognise place names that are in their list. When someone uploads a GEDCOM there will probably be numerous birthplaces that Ancestry don;t recognise, or identify incorrectly.

Yes I'm aware of that situation, and it is frustrating, but I don't think it accounts for all the 'no matches' I am finding. I've tested it with several cousins with an ancestor in their tree with the correct surname and exactly the right birth location (as featured in Ancestry's list). If I search for both I get no matches, whereas if I search for either surname or birth location independently, the relevant cousin comes up as a match in each case. I'm probably being thick here, but I don't see why putting the two search terms together should negate the hits you get from applying each search term separately. It would be useful to search by name and county (especially for common surnames), but I can't seem to get any sensible results that way.

Perhaps I should post this in the 'Search problems at Ancestry' thread and see if my experience is unusual. I know people (including myself) have had problems when specifying 'exact' matches in searching databases. Perhaps this is another example of that.

 

[peter]

I don't see why putting the two search terms together should negate the hits you get

Because BOTH need to match.

 

[Pauline]

I don't see why putting the two search terms together should negate the hits you get

Because when you do so the search doesn’t seem to work properly!

 

[Helen7]

Because BOTH need to match.

But in the examples I mentioned, both did match, e.g. one cousin has an ancestor named Lambert born in Drayton, Leicestershire, England. I've double-checked that the exact name and birth location for this person is in the tree, but searching on these fails to find them. However, I have noticed that ticking the 'search for similar surnames' box seems to find surnames which are in fact identical (that are not found if this box is not ticked). And if the tree entry has UK after England the match is not found on a combined name/location search but is on a location search only. Very strange!

 

[Alexander Bisset]

I did wonder about contacting her (yes it is a woman, using what I assume is her real name) and offering to help her tidy up her tree but a) it would be a mammoth task which I don't have time for, and b) she would probably ignore me, as it would mean deleting thousands of people from her potentially record-breaking tree!

I’d be fascinated to see how many errors Family Tree Analyzer threw up with this tree.

 

[Helen7]

Because when you do so the search doesn’t seem to work properly!

I agree, there is something very strange going on with the searches at Ancestry (as you have reported on another thread), and anomalies in searching DNA matches seems to be just another example of this.

 

[Helen7]

I’d be fascinated to see how many errors Family Tree Analyzer threw up with this tree.

Many thousands, I would think. Could FTAnalyzer cope with such an enormous and error-strewn tree?

 

[jorghes]

Perhaps surprisingly, I was able to locate my (distant) relatives quite easily in this huge tree (despite a common surname, the village location pinpointed them very well), but when I looked at this branch in detail, I discovered a reason for the large number of people. According to this tree, my 5x great-grandfather William married 4 women, all called Jane, 3 of whom were born after William died, and had 40 children over a 130-year period, born in various parts of Canada, Scotland and Bedfordshire. Despite William dying in 1784 (at least that date is correct), he is listed as living (with one Jane or other) in various places in Scotland and Canada in 1851-1901 (sometimes in different places at the same time). There are also at least 4 clones of the same William in the family. This looks like an extreme example of the tree owner accepting every hint offered by Ancestry without checking if they make sense, and ending up with a nonsensical collection of duplicate names and date/location errors. I haven’t looked at other branches, but if this branch is typical of the whole tree, it is easy to see how such a huge number of ‘people’ can result!

I've seen similar errors on smaller trees - one of my 3x great grandfathers is in a few trees with four or five wives (often just the same wife added multiple times) and with census records from the UK and children born in Australia in the same year.

I once attempted to correct someone... needless to say I received no reply.

 

[Bob Spiers]

But in the examples I mentioned, both did match, e.g. one cousin has an ancestor named Lambert born in Drayton, Leicestershire, England. I've double-checked that the exact name and birth location for this person is in the tree, but searching on these fails to find them.

I long ago accepted that checking under both name and location parameters would not work, so I stopped combining them. I also recognised that 'searching for similar' often found names that were identical but only appeared when the box was ticked. You are trying to apply human logic to a binary situation which sees things as off and on. As Peter says both need to match, and if they don't in the way the algorithm is programmed to understand then it delivers a NUL return.

This does not just apply to Ancestry, FMP also stumbles when asked to double search - and again as Peter often says, less is more.

 

[Alexander Bisset]

Many thousands, I would think. Could FTAnalyzer cope with such an enormous and error-strewn tree?

I should put in an option if such a user was to load that sort of error strewn tree that it popped up a message with please delete this tree it’s beyond redemption. Then took them to web links to learning crochet or some other less taxing pastime.