I should elaborate a little.

The average gain is the ratio of the total power in all directions at a
great distance (beyond the point where the surface wave has decayed to a
negligible value) to the power into the antenna from all the sources.
(There's a factor of two also involved when using a ground plane with
NEC but not with EZNEC.) So the average gain is the efficiency if you
consider ground reflection and the decay of the surface wave to be part
of the loss.

Roy Lewallen, W7EL

Roy Lewallen wrote:
Frank wrote:

Correct Owen. NEC shows 97.3% for free space, and 100 %,
as expected, with perfect conductors. Certainly the loss does
include absorption of the reflected rays. As mentioned before, in
previous threads, it is very tedious to determine what percentage
of the "Loss" is due to ground wave radiation. One of these
days I will write the code necessary to compute the actual
TRP including ground wave.

That capability is already built into NEC, as the average gain calculation.

Roy Lewallen, W7EL