I am having a lot of trouble trying to understand how load match correction works. In particular I am measuring directicity of directional couplers integrated into duplexers accurately. I don't see how the VNA knows to apply a load correction. Lets say I have a 2 port network and want to see the s11 response with port 2 terminated in as close to an ideal 50 Ohm load as I can. I am comfortable with connecting UUT port 2 to a good load such an a cal kit load. I have also read that with a 2 port analyzer and full 2 port calibration that the VNA can correct the port 2 load match to look like like a good load to the UUT port 2? I can only assume this is happening via the magic of matrix math. (The analyzer is not actively sourcing any signal for port 2 during a s11 measurement). How can the VNA "know" if the UUT port 2 is terminated with the cal kit termination with no port 2 load match correction needed or with a cabled connection to the VNA port 2 that will require the load match correction to accurately measure the s11 for a low loss UUT? How does this extend to a multi-port analyzer?