Survey numbers are given as a percentage and are derived by mathematical models that are predicated upon assumptions. The flaw of survey (polling) numbers is that election fraud results become the baseline. The polling maintains the fraud and the fraud maintains the polling. The other flaw is independent and undecided voters. It’s not the 1960s and Trump has completely shattered both the DNC and RNC. Plus in this election, a sizable percentage of potential voters have never voted. Crossover voting will also be unprecedented.

The robocalls ask a random sample of R, D and independents who they intend to vote for and their enthusiasm (very likely, or leaning). The polling data are useless because they only reflect change in responses relative to the last fraudulent election. Since the last election had 100,00 votes switched, destroyed, etc, it’s impossible to accurately gauge what is going on. Because only Military intelligence knows who is actually voting and what the actual vote is before it’s adulterated, cured, cooked, and fixed by the traitorous scum.

And the percentage is a derived number that is algorithmic and assumes what will happen in the future according to the makeup of registered voters and prior (historical) vote fraud.

The polling numbers, I’m guessing, are relative to the agreed upon baseline. Since 50+ years of election fraud, all polling agencies pretend and collude. Everyone knows that during voting, turnout is almost nonexistent, the polls are empty, and yet the machines report a million people voted. The polling is all bullshit narrative programming.