RCP's shows polling averages by adding up one candidate's lead and subtracting the other's then dividing it by the number of polls. In this example $(2+7+1-2-2)/5 = 1.2$. Is it valid to average out the margin of errors ($(5.7+3+4.3+4+2.9)/5 = 3.98$) and call that the uncertainty range? So Biden could up by 5.18 or down by 2.78? I'm asking because it seems people always talk about errors multiplying when you include other factors, so I want to know if that's the case with polling errors