Class B address: AAA.AAA.XXX.XXX, where A stands for network bit and X stands for host bit, and an effective subnet is to use host bit X as a subnet.

XXX is a decimal digit of 0-254, which is yyyyyy (8-bit y, y represents 0 or 1) when converted into binary representation.

Then you can use AAA. AAA. Y1Y2Y3Y4Y5Y6Y7Y8. Z1Z2Z3Z4Z5Z6Z7Z8 to represent the effective subnet (in order to distinguish it, it is represented by Y and Z respectively, but in fact, the actual meaning of Y and Z is the same).

Then Y 1-Y8 and Z 1-Z6 can be used as subnet bits (you can't go to Z7, otherwise this subnet only contains 2 IP addresses, which is not a valid subnet), so the total number of bits used for subnet by * * * is 14 bits (8 y and 6 z, 8+6 =/kl).