I changed the centered moving average from 14 days to 15 days. Why? With 15 days there is a natural center date so that is one reason. A 15 day moving average is not going to look significantly different from a 14 day moving average, so that is reason two. The idea behind the moving average at this length is that the state of North Dakota reports the 14 day average, and it coincides with the typical incubation period.
We see the consistent increase in the moving average since about July. The moving average allows us to smooth out the variation that occurs in the daily positivity rate (17.2% for 11/05/2020).
I also made an identical calculation for the larger counties in the state. The situation is more variable at the county level with some higher and some lower than the state average.
I am kind of at a loss for words with Ward county. A smoothed out positivity rate above 25%? This is an astounding number. The other counties are all clustered more in the 12.5% to 15% range.
These counties are the economic and health care centers in the state and with high levels there, as people from other counties come there to shop or receive care they come in contact with COVID and bring it back to rural areas with them. (We can discuss this more later). The high positivity levels in these counties goes a long way to confirm the reason for the increase in deaths (see recent post here). Higher positivity indicates increased exposures overall and translates into higher incidences of illness and possibly death.
I will keep updating the state and county data and look for other ways to present the information that make the situation clear.