DNA cM across segments

I really don't understand how DNA works in relation to cousins and relationship position.
How can I share only 32 cM across 2 segments with A my 3rd cousin and then more, 67 cM across 3 segments with B a more distant cousin. I know exactly/all which ancestors they come through.
Looking at Peters coloured relationship chart, A is my 3rd cousin and B is my 4th cousin. In my logic surely I should share more DNA with a closer cousin. I really don't understand. Can someone please help.
To make it even more confusing FTM has cousin B, as my 3rd cousin 1 x removed?

 

How much DNA we share with any particular relation will depend on how much DNA we have each inherited from our common ancestors. So we may well share a lot more - or conversely, a lot less - DNA with a cousin than, on average, we might expect to.

 

So does this mean that the 3x great grandparents, two sons on my dads side passed more DNA down the line to my 4th cousin, than my 2x great grandparents, two daughters on my mothers side to my 3rd cousin?

 

Yes, I agree - I didn’t express that clearly.

 

The coloured chart below is the one I have been looking at.

 

That is the chart I'd assumed you were referring to but I'm not sure exactly what it is you don't understand about it. I guess I may have confused you more by not expressing more carefully what I was trying to say in #2.

You can see from the chart that, while on average you might expect to share 74cM with a 3rd cousin and only 35cM with a 4th cousin, the range (which is given below the average figure) shows that there would be nothing untoward in sharing more DNA with a 4th or 5th cousin than with a 3rd.

DNA inheritance is a random process, while relationship predictions are based on averages. So our actual relationship with a DNA match will often vary from the one predicted.

 

Quite so. In plenty of cases, I share more DNA with a 4th cousin than a 3rd cousin.

 

Its not the chart that I don't understand, just didn't realise you could share less matches with a close cousin than a distant one. DNA has not worked out for me, as I expected it to. Following the masterclass, the surnames which I have struggled to find more out about, there are no matches, as of yet.
Also I didn't understand is how the chart shows one relationship and the software FTM gives another, which is probably question for another page.

 

I recall Peter either in a Newsletter or in this Forum using a visual display of playing cards to represent the distribution of DNA which I found both entertaining and informative. It might help to resurrect it again.

On a different (but essential the same) tack, on my recent visit back to Birmingham, I made contact with four different female cousins. One a first cousin with whom we stayed whilst we visited a first cousin once removed (now in her 90th year). Then via 'Skype' contact with a second cousin even more elderly and frail (approaching 97) who lives in Italy. This was pre-arranged with her daughter who lives in the UK (with whom I have regular contact) via her sister who lives with her mother in Italy - both second cousins once removed. Not sure if there is a collective name for a group of cousins, but if there is this qualified.

It was only afterwards when my first cousin asked about the likely distribution of DNA amongst them all was I able to quote average statistics (as per Peter's Chart) which in round terms from 1st cousin to 2nd cousin once removed was 12%-6%-3%-1.5%. But then trying to explain that this was merely an average and in practice shared DNA depended on distribution. I remembered the pack of cards explanation and did my best to explain how it came down to the luck of the draw. I think this helped and was certainly much easier than trying to explain cM's and segments.

 

I agree, a brilliant demonstration of DNA inheritance, in an easy-to-digest format. It was in the Newsletter of 10th July this year. I remember recommending it to a friend who's recently had her DNA tested and is just starting out in family history research.

 

Oh dear think there are crossed wires, that's not what I meant, nothing to do with the DNA cM. Obviously too difficult to explain. I'm talking about the relationship someone is to me, through a direct line, no half ancestors. That's why I said it was probably not a question for the DNA page.

 

Thanks I'll have a look for it.