At what shared DNA threshold are the DNA matches noise?

[bigsnake49]

10cM, 20cM? Anybody knows?

 

[TardisBlue]

It depends, IMO. I personally pay more attention to the length of the segment than to the sum of shared segments. The longer the segment, the more interesting. That's why, personally, I'd take notice of a 14 or 15 cM shared segment, but would probably discard 2 shared segments of 7 cM each… unless there's genealogical evidence of shared ancestry, like ancestral names in common, or ancestors sharing the same birthplace. From what I've read, a shared segment >15 cM generally means the match is real (IBD), or rather, there's like 90% chance that it is. But of course, it doesn't mean that if you share a smaller segment, it's necessarily a false match. The only way to find out is for you and your potential match to both work on your genealogies to find a most recent common ancestor.

 

[Jaime__]

The longest the cM the nearest in time to your ancestor´s line (most probable).

There is a common mistake when people think about the length of segments. A 4cM segment could be totally valid and plenty of segments can pass generations unaltered.

 

[Pajonk]

I think that it is not correct answear...

 

[Jaime__]

I think that it is not correct answear...

And your reasoning?