I think the hierarchical order you have placed them makes perfect sense.
The order is completely by chance. It was just the order in which certain aspects were noticed by me. I think that Peter's notes on the matching process are very good but it seems that some people might approach this topic from a slightly different angle and end up with the wrong expectation. To me, this whole matter is fundamental to the purpose of LC and so I stayed up late to try and give an alternative description which might give some enlightenment rather than leave it and forget to comment later. I hope that my comments when taken with others will help to clear up any misunderstandings. After all, what is the point of using LC and expending a lot of time and effort if the matches cannot occur because some of us are entering slight variations?
By the way, I do not claim to be perfect and always know what needs to be done in all circumstances so welcome corrections from Peter or anyone else if I have said anything that is unclear or incorrect.
Or is there a heavier weighting on the census ref, and then fuzzy logic matching for the remaining fields?
Obviously the census reference must match exactly for LC to even realize that there may be a potential match. After that, the other fields identify individuals within the subject household, some of which may be very similar if not identical. Hence the benefit of Peter's fuzzy logic matching which suggests that a further manual check may be beneficial. Unfortunately, some members may approach any second check with preconceived ideas and not notice subtle differences in their entries from the original census/transcription. I also can be guilty of that so try to be very careful when re-doing such checks and do not go in with an expectation that I must be correct. If only there could be some way of resolving such issues where all parties are convinced that they are correct, rather than just ignoring that there might have been a possible match.
one relative of mine is pretty messed up at FMP compared to the transcription I used from Ancestry. Does this mean I'm not going to match if another member has used FMP?
Almost certainly, because each member will say that their entry is correct. There is no means for negotiation and resolution.
Unfortunately, FMP and Ancestry take different approaches to adjustments in their transcriptions.
FMP make changes but do not alert anyone that they have done so, presumably assuming that everyone is only interested in the latest, correct information whereas LC members may have relied on the value from years previously and may not realize that a correction has been made.
Ancestry do not change the record but do add a comment indicating an alternative suggested value, or more, at least for the name of the person.
Neither attempt to alert earlier readers about such changes.
When I add an entry at LC from Ancestry and then notice that the grey arrow check reveals a different value from FMP, I usually alter my submission to match that shown by FMP on the assumption that other members will probably do likewise, or have used FMP originally. That does not make a lot of sense in the example quoted. However, it is essential that differences noticed as a result of the grey arrow check are not allowed to persist as that would inevitably lead to matches being missed. This is especially true when Ancestry has given the wrong value of Folio. As Peter has often mentioned, do not rely on the transcription, except in the case of the UK 1881 census.
