Connie, FTDNA has set the threshold for a match at 7cM - every match will have at least one segment that is 7cM long (and they'll have some other smaller, shared segments) The whole business of what's a good shared segment is a sliding scale of probabilities. There is NO fixed number for which all larger segments are from Common Ancestors and all smaller segments are not. FTDNA chose 7cM as a happy median - most such segments will be from Common Ancestors, but not all. Some people choose to use 10cM as a threshold, and in fact a higher percentage (but not 100%) of those will be from a Common Ancestor. Others use 5cM as a threshold (at GEDmatch, where you can set the threshold; or as a 2nd or 3rd segment for a Match at FTDNA). A higher percentage of these will be false or IBS (not from a Common Ancestor), but some more Matches and segments that are from Common Ancestors will be found. A lot depends on your goals... 1. If you want to map your genome - find specific segments from Ancestors to cover each of your 44 chromosomes - you should look for all the shared segments you can find (at 3 companies and GEDmatch) and set the threshold at 5cM. Then using Triangulation (and/or phasing), weed out as many of the IBS/bad segments as you can. 2. If you want to confirm much of your genealogy research and family lines, you might want to set the threshold at 10cM (or higher) and deal with fewer Matches. 3. If you are an adoptee, you'll probably only want to deal with much larger segments, hopefully from closer cousins, so you can focus on their families. 4. If you are focused on a particular brick wall, you might want a combination - using larger segments to work on a broad brush picture of your chromosomes and then perhaps use even 3cM shared segments that are on chromosomes and in areas that are probably where your brick wall is. You'd want to look at a lot of smaller segments in this are trying to find matches with ancestry that is beyond your brick wall (in place and time), so you can study those. 5. If you are looking for ancestry of a particular ethnicity, you would use the admixture tools (including several at GEDmatch), to find the general area(s) of your chromosome(s) that show that admixture, and again, lower your threshold to get all the shared segments and Matches that you can from that area(s). So as you can see there is not a single answer to your question that fits all situations. Same start or stop points can occur for several reasons: 1. These are the start/stop points of the segment you got from an ancestor (the start/stop points are recombination crossover points, and are fixed on each of your 44 chromosomes (randomly different on the chromosomes from Mom and Dad) - this is what you are hoping for... 2. These points may also be points on some chromosomes where the testing companies don't test (see grayed areas on the Browser) 3. These points may reflect the boundaries of blocks of SNPs FTNDA uses in their algorithms - they actually compare blocks of SNPs and the end points of shared segments are not really precise (think of sharpening a marshmallow) 4. And there are known crossover "hot spots" where breaks tend to occur - but that would be covered under #1 above. I don't understand "breaks in the segments". A shared segment is a contiguous string of SNPs - no breaks. If you mean gaps where the shared segments from several matches overlap, but don't exactly match up - well... the rule I use is they should overlap by 7cM, the same criteria as for a match. If there are several (say 5 in the chromosome browser, and they all look like lasagna - get the image?), then if they all share say 5cM it's a good bet they are a Triangulated Group (if the Matches match each other on that segment). If there isn't a clear 7cM overlap for all, it becomes a judgment call. If you have a bunch of segments in the same area of a chromosome for one person, they should wind up in two, well 3, groups - the person's Mom, the person's Dad, and IBS (neither). As with everything in random atDNA, it's possible to have something fall outside of these guidelines (say a 20cM IBS segment). The above works, but there are no absolute, 100% guarantees. Jim Bartlett On 11/21/13, Rvsailor@aol.com wrote: I am a project administrator of two atDNA projects. When dealing with a spread sheet of several ICW cousin matches downloaded from the GAP, should I discard all segments less than 10 cMs or should that number be closer to 7 cMs when the paperwork trail supports relationships? I initially went through the list and deleted everything below 10 but was concerned about discarding some of the 7, 8, and 9 matches as some had the same start or end points with others. Also, how should I handle matches on the same chromosome with breaks in the segments? Thanks. Connie Bradshaw John Carson of WNC Bradshaw & Harris