Please treat the name WRIGHT in the same way as is applicable to SMITH

LegacyUser

Paul said: I am being overwhelmed by having to dismiss suggestions against WRIGHT and WRIGHTSON relatives, both relating to record hints for sources and possible duplicates.

Further, I am having to send messages to a number of inexperienced users who are merging the most unlikely of duplicates imaginable, just because "FamilySearch" is suggesting them as a match.

Some years ago, Robert Kehrer advised me that the most common names ARE treated differently when constructing the algorithms, and I'm sure this must apply to SMITH and possibly a few other English surnames.

So please save me from being distracted from important, productive work in Family Tree, by "doing to WRIGHT" whatever is currently applied to SMITH, thereby stopping these countless, totally inapplicable suggestions that are appearing on the affected person pages.

Please see https://www.surnamemap.eu/unitedkingd... . I do not wish to sound selfish, so am sure if exceptions could be applied to some of these names, too, other users would also be spared much wated time!

LegacyUser

Paul said: Well, I didn't expect my issue to be resolved too quickly, but neither did I expect the situation should get worse so quickly, either.

See https://www.familysearch.org/tree/mer....





The only thing these two individuals appear to have had in common were a wife and daughter called Ann(e). One from Yorkshire, one (apparently) from London. It is bad enough to be constantly receiving "Stephen Wright" suggestions from all over England for different time periods, but Joseph Wright might be Stephen Wrightson???

LegacyUser

Adrian Bruce said: Face palm again....

LegacyUser

Adrian Bruce said: I wonder if there is some sort of knock-on from the way that FS treats patronymic family names? Another thread elsewhere, which I have been reading just for interest, suggests that there is coding in to (very roughly) equate Magnusson and Magnusdottir (say). I wonder if Wright is treated as having the same root as Wrightson and is therefore equated now? Would the same problem exist with Wright and Wrightley (were there such a name)?

LegacyUser

Paul said: Although in this case I was specifically highlighting the forename difference, there did used to be some strange anomalies when it came to surnames (and probably still are).

For example, I got WRIGHT but not WRIGHTON (no "s") when making a WRIGHTSON search, or didn't get RIPON when looking for RIPPON or RICKABY for RICKERBY.

We probably also have the problem to contend with whereby things work slightly differently in Family Tree than on the "main" FamilySearch site.

The point of the original topic relates to FS manager Robert Kehrer's comments when I raised a similar topic some years ago. He assured me the most common names (say Smith, Jones, etc.) are treated differently, to save possibly thousands of "possible duplicates" appearing, say against a "John Smith".

Generally, I would like to see that extended further down the list of "most common surnames", but specifically to NOT include WRIGHT as an alternative for WRIGHTSON. As mentioned in another thread, I have only come across WRIGHTSON being recorded as WRIGHT on literally half a dozen occasions in my 35 years of research, so would be delighted if I could be left to find these exceptions without FamilySearch's help. This would save me many hours of unnecessary work in dismissing all those WRIGHT suggestions that appear against the many hundreds of WRIGHTSONS I have added (or am working on) in Family Tree.

LegacyUser

Paul said: Maybe I should be grateful my name is not SMITHSON. To get some idea of how the current algorithm might be working, I went to FIND and searched for John Wrightson instances in Durham, England.

Checking the exact name box for JOHN (but NOT Wrightson) I got 727 results - 50 of which were for John Wrightson, the rest for John Wright.

Doing the "same" exercise for a John Smithson, I received 1920 results - 30 for John Smithson, the rest for John Smith.

After performing other searches, using various criteria, I have come to the conclusion that the factor is NOT necessarily the surname WRIGHT having "looser" search criteria than SMITH, especially when it comes to offering possible duplicates. Rather, names like Wrightson and Smithson (and probably Taylorson) do need to be treated as being unrelated to the names from which they were derived.

So just as an "exact name" search on WRIGHTSON & SMITHSON solved my problem when using FIND, applying this factor to the "possible dupes" algorithm would be the thing to save me so much wasted time.

LegacyUser

Paul said: Just another example of a totally crazy suggestion of a possible duplicate:
(1) Birthplace already inputted and does not match.
(2) Herbert just one of middle names of suggested dupe.
(3) Wrong surname.



However, I know from previous experience that without my recording this as "Not a Match" another user would almost certainly merge the two IDs!

LegacyUser

Paul said: Working on the above (Herbert Wrightson) today. Went into his Sources section and what did I find? Yes, a source for Ralph Herbert Alan Wright HAD been added by "FamilySearch" in 2014. (Though I believe Nov 7, 2014 was one of those dates that there were thousands of imports to Family Tree, which all had FamilySearch as the username.)

(BTW - middle name Henry added by me to Herbert Wrightson since earlier image posted.)



(Source now detached.)

LegacyUser

Adrian Bruce said: Paul - I really hate to say this but it looks like you hit the wrong button - or FamilySearch thinks you did. I was looking at this just in case HH Wrightson had a long involved history with the India Births source attached years ago that I could explain. But...

If you go into All Changes for GQ1T-85C ​ Herbert Henry Wrightson, you will find the All Changes history contains TWO "Merge Completed". The last was for Herbert Henry Wright, 14 October 1870 – Deceased • MBCY-TYQ

and it says
Reason This Merge Is Correct
Different surname / place of birth

Prior to that is the other Merge Completed for
Ralph Herbert Allan Wright, 7 March 1873 – Deceased • 9JNR-29Z
and it says
Reason This Merge Is Correct
Wrong birthplace, first name & surname

The last merge is why the India Births source was on your HH Wrightson chap.

So you've written the two reasons in terms of rejecting the suggestion but appear to have selected the wrong button. Clearly I have no idea why FS thinks you selected the wrong button - but one thing to be wary of is that if FS had not finished drawing the screen when you hit the (as you thought) right button, the click might actually have ended up on the wrong button.

LegacyUser

Paul said: Adrian

Thank you for pointing out my glaring error! Just reversed the merge.



Glad it was you that found my error - I would have been well shot-down if it had been someone else (no names!). Proves even I'm not perfect!

LegacyUser

Adrian Bruce said: I'd blame the screen repainting if I were you... :-)

(That actually regularly happens to me on Ancestry with their hints - I go to click on one button but by the time I really click, the print icon has appeared from nowhere and shifted the buttons along by one... FFS this is what egg-timer cursors were invented for, to tell the user, "Hang on, I'm not done painting the screen yet...")

LegacyUser

Paul said: I must have been having a funny half-hour earlier. Anyhow, both the Herbert Henry Wright (illustrated) and Ralph Herbert Allan Wright IDs have now been restored. Thanks again.

LegacyUser

Juli said: India?? Wow. Just wow.

LegacyUser

Adrian Bruce said: As we have said, relationships seem to trump any geography in dupe suggestions.