- This topic is empty.
3rd February 1999 at 13:29 #32349Steven ChiaGuest
Does anyone whether what assumptions are made about the nature of retries in the Erlang C formular? That is, does it take into consideration people making multiple retries as opposed to just a single retry?3rd February 1999 at 13:29 #32350Jim KeaneGuest
Consideration of retries is not appropriate for Erlang C. This is because Erlang C makes the following basic assumptions:
1. Call are never blocked. They always enter some form of queue.
2. Call remain in that queue forever, and never clear down.
So, there is nothing to retry!
Hope that helps,
Jim Keane.4th February 1999 at 13:30 #32351Steven ChiaGuest
Jim thanks a lot for your reply.
What I actually intended to ask was about Erlang B extended model, not Erlang C, a mistake I should apologise.
Do you know about the assumptions made on the nature of retries in the Erlang B extended model? Is it just considering a single retry per failed attempt, or does it take into consideration the possibility of doing multiple retries per failed attempt, which I think is a very natural human behavior.4th February 1999 at 13:31 #32352Jim KeaneGuest
It assumes multiple retries. So, if the retry figure is 70% it assumes that if you make a call and are blocked, there is a 70% chance of your making another call straight away. If that second call is blocked, there is a 70% chance of your making a third call etc…
As you say, it models natural behaviour.
Jim K.5th February 1999 at 13:31 #32353Steven ChiaGuest
Thanks a thousand. This is very helpful for me to size my customer’s call centre. Thanks again.
By the way, do you know of any commercially available tool that helps to size call centres that has complex set up, such as having: 1) PABX 2) IVR 3) Predictive Dialler 4) Internet call centre with feature such as: