Variance is also called MS, Mean Square; approximated by the sum of squares (SS) divided by the degrees of freedom (i.e., n-1); the MSE from the empty model can be thought of as roughly the average squared deviation.
Confounding variable is a variable that, though not measured, causes variation in both the explanatory variable and the outcome variable; also know as a "lurking variable" or a "third variable".
When we can make a better prediction of the outcome variable if we knew something about the explanatory variable; accounting for variation in the outcome variable by using variation in another variable.