Scale Matters

It's Monday, and we're waiting to see how the petty tyrant's proclamation will affect us today.  But, the Baton Rouge Advocate put out some government chars today.  They purport to reflect the  the impact of the covid virus o Louisiana, but the charts themselves are seductive.  They are horrific, but if you know how to read a chart, you learn that they're not reflective.  Oh, they are accurate.  But they are not reflective of what is going on.

Scale matters.  If you are running a series of charts, the scale has to be consistent so an observer can see what is happening.  For example, this chart of the Covid impact on the New Orleans

Look at this chart.  It's probably a standard bell chart, or perhaps a Poisson distribution.  It shows that at one point, the New Orleans metro area went over 120 per 100k residents.  And, they are in decline.  That's good nes for the New Orleans metro area.

Now, let's look at this next chart, which pupots to tell the same story for central Louisiana.

Omigod, we're all going to die.  Central Louisiana has had a double spike.  Infection is rampant, much worse than New Orleans.  Until you look at the scale on the right side of the chart.  Our high point is 8 cases per 100K (where the NOLA  high point is 120 cases per 100K.  If those two charts were done to the same scale, the scale, the Region 6 chart would basically be a flat line.

Based on the numbers, central Louisiana hasn't been affted nearly to the point that NOLA has been affected.  Our cases are statistically  irrelevant.

Beware of charts.  Scale matters.