> You might be confusing the population variance with the sample mean variance.

That's exactly what I'm doing - thanks for pointing that out.