As far as I understand - next call will get next available agent. I understand, completely pointless answer. But look, the system does not have plans. If you think about it - it cannot have plans, because the situation can change any millisecond. Next agent will be determined when the call need to be assigned.

If you think about queuing and agent assignment algorithmically you will notice that actually there are two possible situations - "little load" or "heavy load".

In "little load" scenario, there is no call queue, there are lots of free agents (at least 2). In this case the call is agent picking initiator. The call will "usually" go to "longest available". You can have field showing "longest time in ready state". The agent at the top will be the one.

In "heavy load", the are calls  in the queue and no available agents. In this case it is inconvenient to go through all the calls in the queue and ask "is there agent for me?" - the agent becomes "picking initiator". In this case it is like agent is picking suitable call for him. This way time spent looking for a call is not that dependant on queue depth. But, because of this - you cannot have prediction. Agent picking is just reaction to "agent ready" event.

Sorry for long post. Hopefully this explains things. I didn't write CCX or CCE, but that's how I think they work. Hopefully somebody smarter than me can explain it better... or prove me wrong.

Ervins