No opinions or assumptions, just data. No agenda other than to inform myself. I put this together because I wasn't seeing the data the way I needed to see it to understand current situation.

In my work I monitor time series data continuously, watching for inflections in reliability, performance, unwanted behaviors, etc. So it is important to me to see the data without mixing in historical data in every data point (cumulative data), to see it on a timeline, and to see it broken down into interesting segments (eg, states).

Some things to keep in mind as you interpret these visualizations:
- When measuring covid-19, deaths are a lagging metric. Changes in direction in a trend line typically correlate to an infection modifier occurring 2-3 weeks earlier to affect infection counts. Examples: new infections introduced to a population; changes in behavior; etc. Perhaps a change in treatment would lag less than infection modifiers.
- Large data sets are never perfect, but this data set appears to be well-scrubbed, well-managed and as consistent as possible for data coming in from so many sources.. The download sight explains methodology, inconsistencies, etc.

Data source: https://github.com/nytimes/covid-19-data

I'm in northern California. The Bay Area locked down quickly, and our numbers are very low. In Santa Clara county for a pupulation of 2 million, we are seeing about 20 new cases and 1 death for a few weeks. We've had a little over 100 deaths total with around 35 from nursing homes.

On the opposite end is LA county who had the LA marathon when this first started and they account for most of the problems in California. Take them out and California is good.