Paranoid I shouldn't be matching them - just in case... thanks.
Sometimes records can seem a match to people in the tree but only later after finding more documents can it be 'proven' one way or the other. I see nothing wrong with attaching a record you suspect belongs to the person in the tree - unless there are 'obvious' discrepancies with names, dates, locations, or relationships.
Just keep it in mind there is no black and white rule. You try to use common sense and analytical thought. If you think there is a high probability it is the same person person You link it. Otherwise no. Different people may be using different Thresholds for their analysis and it is a subjective process . But seriously consider what the possibilities are that it is not a true match. Factors can surely vary from one person to another.
To stay out of trouble, consider this: FamilySearch never gives any matches. The programming algorithms present suggestions, possibilities, and occasional random nonsense.
It is totally up to you to evaluate what the program presents and decide on your own if the information presented, whether a source record or a Family Tree hint or possible duplicate represents the individual you are researching or not. If you can conclude it does, then attach the information to that person. If you can't, then don't.
Sometimes this is easy because there is so much information that it's obvious. Other times, there may be so little information in a record, it can be impossible. Then there are the times when the name is off a bit, a date is close, places are not quite what you would have expected and you have to spend a few days, or longer, first learning about the record then developing a rational line of reasoning for why any discrepancies are there. That's the fun of all this.
If you're talking about merging, my advice is this:
Look for (sourced) conflicts rather than similarities, and resolve those before merging anything.
Way too many people merge because one child belongs to a different couple, rather than just moving that person to the right family. Or they'll merge because the dates and vital info match, without checking whether that info is sourced and correct for both profiles it appears on.
Likewise, I've seen way too many profiles marked as Not a Match because the unsourced, estimated dates or guessed birth location were different, ignoring the fact that all relationships are identical, the duplicate family was added recently (usually by GEDCOM import), and absolutely none of it is sourced. Situations where they are duplicates, just with mistakes.
If you are talking about when adding a person, spouse, etc., you type in the name and date, and a possible match comes up that you can select or ignore (often with a spouse attached). Is that what you are talking about? When you see that existing record suggested as the person you are trying to create, copy the ID (all you have to do is click it and it is copied), open another tab and go to that person. Then start studying the facts, and searching for new facts on that person. Make sure everything that you find is consistent. Now, here is the fun part where there is widespread disagreement on how to proceed. What to do if the evidence is inconclusive? This person may be the right choice, or he / she may not be. Now what? My take is this: the FamilySearch tools are wonderfully adaptable. The nice thing about FamilySearch is you can fix what you break. What is done can be undone. What would I do? I would add the match, go to notes and explain why I added. Yes, someone may come along and later disconnect the relationship - it may even be YOU if you discover it wrong! But that is ok. This is all part of genealogical research. HAVE FUN!
Perhaps you could elaborate just a little, to illustrate why you are raising this question. For example, are you talking about FamilySearch offering a "possible duplicate" suggestion and are asking on what criteria is this being based?
From the evidence available, users are left to make the final decision on whether an individual is a match for another. Or, perhaps, whether a record is a match for the individual to whom it is being suggested it refers.
Your question is too vague in informing us whether you are referring to "record to individual" match criteria, or "possible duplicate individual" match criteria.
Sorry, yes... I'm new to this. It is not involving a possible duplicate that has been flagged or that i am flagging...but when I am conducting a possible match from a source to an individual in my family tree, (the attach historical records to family tree section) hopefully that makes more sense? Thanks
"occasional random nonsense" - thanks for the chuckle. Nodding in agreement over here.
Thanks for taking the time everyone! I appreciate it. I think all of the answers together have painted a good picture for me as I begin to dig deeper into family history and in particular merging and attaching sources - initially two of my biggest concerns. 🙂