AI-GEOSTATS:tunning a semivariogram

[Javier]

Hi all; I guess this is a very simple question: Is it supposed the sill is always the variance of the set of data? If so, is it always better to perform a manual tunning, (after the inicial automatic adjusment of a semivariogram establish a sill -say in gstat-),  to force it to reach the variance of the data?   Thanks and regards ---------------- Javier Garc�a