Quick question regarding a DNA match

[brh2001]

Hi there!
So I was recently reviewing my AncestryDNA matches and came across something a bit odd. To cut it short, I have several DNA matches who are the grandchildren of one of my great great great uncles (the brother of my great great grandfather, making the matches my second cousins twice removed,) each of them within a 20-30cM relation to me. However, there is one DNA match that I have who is also (ostensibly) the grandchildren of one of my great great great uncles who is a whopping 100cM. In Ancestry's estimates of likely relationships, a half 1st cousin twice removed is at a 51% likelihood and a 2nd cousin twice removed is only at 22%. Does this make it likely that his grandfather is actually my great great grandfather?

Cheers!

[Gadget]

Hi and Welcome to Rootschat 

Have you looked at

https://dnapainter.com/tools/sharedcmv4

If you enter 100 in the top box, you'll find that you have a 30% chance of being 2c2r with the 100cM Also, if you enter 20 and 30 you have an 8-9% chance of being 2c2r.

I think it might be best to check the 20-30cM matches as, the calculator suggests they are more likely to be more distant or half cousins.

Gadget

[Gadget]

The chart images below show

1 full range of relationships with shared 100cM

2 full range with 30cM

[brh2001]

Hi there!
I haven't actually come across this tool before. What makes the extremely low likelihood of my other matches being half cousins incredibly puzzling is that we all have shared DNA matches from our common ancestors' maternal AND paternal lines. You are seemingly correct in that this is the more likely culprit, however I am just baffled as to how it would be the case. Perhaps it's just an unusually low amount of shared DNA with those cousins?

[Gadget]

Have you contacted any of the lower cM group and, if so,  do they match each other?  Have you contacted the 100cM person, does he/she match any of the others? 

I've found contacting just a few matches can give important clues. If some of you could upload your results to sites that have chromosome browsers and examine how/where you match, you might be able to solve some parts of the jigsaw.