Well, as my mum taught me, I must make sure I get all my facts right before I frame a theory.
I have found two Fionas who are decidedly dodgy, which while I can't yet think of any others I know still makes an 80% hit rate, with also a 100% miss rate from the ones I tried (or didn't even waste my time knowing the result..).
Then I decided to do a control name to see if it worked for any name. I chose Catherine/Katherine as that was the name with the most girls attached to it I knew, and here are the results:
Total names found: 17
From which top quality: 13
comparitive percentage: 76% (pretty close), while the other 3 were OK but not division one.
Then I tried Sarah, who was also a fairly popular target for me over the years but not so many generally.
Total names found: 9
Top quality: 6
Top percentage 66.6%
Alison: Total names found: 7
Top quality: 5
Top percentage: 71%
Caroline: Total (so far, probably more): 7
Top quality: 5 I think
Karen: Total (so far) 8
Top quality: 3
There is my favourite name but only a small sample, which is Sophie but I can only think of four, all who were top quality. But at least I had a little success with one though it was at a party where we all swapped around officially so she was left with little choice!
So the system isn't perfect but still pretty spooky. I haven't fully checked all the last few names but it still seems some are connected and most of the others I don't think I know enough but will try a few more as this is a fascinating test, and one Tommy Boyd already did a rough version of a few months ago. It also keeps me out of trouble...