|
Precincts Max Voters Per Type
Methodology:
X-Axis: straight-party Republican percentage is sp-rep / (total straight party) Trump Percent is individual Trump / (total individual) Y-Axis: difference from center line is Trump Percent minus SP-Rep Percent. ("Expected" difference on line is white, better than expected above line red, worse below line blue.)
Conclusion:
With the above methodology, a completely random sample trends downward as "Republicanism" increases. Therefore, the downward trend is not necessarily algorithmic. Basically, the more Republican the precinct, the less chance it has to beat Republican expectations. There may well be other issues with the Dominion tabulation, e.g., the recurring leading flatline in low-Republican precincts described in their presentation. View page source for the simple scripting involved. MattG |
|