Hello all SeqFan members. To obtain the various k suitable for first cousins, I started from Balestrieri's consideration for twin primes. However, these are the forms:
k1 = 12*x*y - 2*x - 2*y + 1
k2 = 12*x*y + 2*x + 2*y + 1
k3 = 12*x*y + 2*x - 2*y - 1
k4 = 12*x*y - 2*x + 2*y - 1
The odd values of the complement sequence k1 U k2 U k3 U k4, multiplied by 3, are the average of two first cousins.
What do you think? Can I propose this to the encyclopedia?
Davide