The formula for calculating correlation is: The correlation between ballots cast and registrations for all counties in Ohio is .9993109593 which is of course EXACTLY what one would expect. You will notice he doesn't apply his curve fitting formula to counties like Geauga or Athens which had high and low turnouts respectively. So no, it's not evidence of fraud. It means turnout is reemarkably consistent across the State and across age groups. That's not weird or odd. Lindell is brain dead stupid, and Dr. Frank has to be a FRAUD in order to MISUSE correlation in such a manner.
Lets not forget that Ohio uses around seven different voting machine vendors[*] with at least three different methods of voting all using different software. This would require hundreds of lizard people embedded into all these companies and all 88 counties to enter Dr. Frank's magic "Sixth degree polynomial equation".
[*] ftp://sosftp.sos.state.oh.us/free/publications/website/maps/votingSystems.pdf
“The formula for calculating correlation is: The correlation between ballots cast and registrations for all counties in Ohio is .9993109593 which is of course EXACTLY what one would expect. “ -rwc1
No - not ever in the real world with real independent data would one expect a correlation that high.
0.8 is really strong; 0.95 amazing; 0.999 fantasy-with real data even when there is a expected behavioral relationship. This is not testing a physical property, like sunrise to change in brightness. This is behavior.