Need help coming up with custom QualityScore algorithm

We have a call center and i develop custom metrics applications to help us analyze our call center data. What i would like help figuring out is a math equation to determine the quality of the call on a particular day.

These are the metrics i have to play with:

Volume or number of calls
Total minutes on the phone that day
Average call length
Number of leads generated from said calls
I also have a history of all calls into the call center

This is what we expect to see from looking at all the number previously mentioned:

low call volume and high leads = greater quality
low call volume and low leads = normal quality
high call volume and low leads = low quality
high call volume and high leads = normal quality

We expect that the more volume the call center has to work with the more leads they are able to produce. If the leads are low and the volume is high then this would generally mean the quality of the calls are obviously not very good.

I would also like to take into consideration the average call length. Longer they spend on the phone generally means the calls are more qualified and more likely to convert to a lead.

Additionally, if it is a Monday, i would like to reference and compare all previous Mondays and base the Quality Score on a historical account of that day of the week.

Does anyone have any idea how i can get all these thoughts and ideas in the mathematical equation/algorithm. Could anodyne help me make mathematical sense of this?

Thanks,

Duncan