Intuitively I think the number of channels (N) in a patch is:
Being n the average of the maximal levels among trials and p the open probability.
Example 1: I apply a voltage pulse of 500 ms to a patch, many times. I get an average of maximal number levels of 3.32. The open probability is 0.2. Then I got aprox 16 channels in the patch.
Example 2: I apply the same protocol to another patch. I get 0.2 levels. p is 0.1. Then N=2 channels
Correct? Any reference you know supporting this?