The hello's sent over an NBMA network are unicast hello's unlike the broadcast or P2P network type So when unicast hello's are being sent, the router needs to know the IP address of the other router. That's why you configure a neighbor statement in the NBMA environment, as opposed to broadcast or P2P, where the hello's are sent on to the multicast address 18.104.22.168 and the neighbors are automatically discovered. The attempt state exists only on an NBMA. I think it's when a router has sent an hello to the unicast address of the IP address that has been configured manually with the OSPF neighbor statement, and waits for a hello from that router. If it doesn't receive a hello within the dead timer of 120 seconds, it goes into the down state.
You can check this link for the configuration Initial Configurations for OSPF over Non-Broadcast Links - Cisco
It's all detailed well in the link. Check it out.
ip address 22.214.171.124 255.255.255.255
ip address 126.96.36.199 255.255.255.0
ip ospf priority 2
frame-relay map ip 188.8.131.52 16
router ospf 1
network 184.108.40.206 0.0.0.255 area 0