If spent more than a few minutes on the COVID Mapping Project website, you can immediately tell we have one over arching goal: create a level playing field where comparing data between states is relatively straightforward. This is because each state has it's own Public Health department with their own data systems and their own way of measuring the severity of COVID-19 in their state. Since we're largely a GA based group, we've paid close attention to how GA is representing their data. To say we've been vocal and critical would be an understatement: GA has a series of bungles that has caused lapses in trust with the state.
To that end, we wanted to see how other states are tracking "new" cases. So we looked at Texas, Louisiana, and Georgia.
Georgia
We've written...a lot...about how we feel about Georgia. They use the date of onset which means the previous 10-14 days will always be lagging. You can see on the right hand of the graph the severe tilting down. They've given this a slightly different markup which suggests that this data is subject to change. You can read more about how we feel about this graph and what we think of it here.
Texas
This graph from Texas is a special kind of bad. The line graph on the same as a bar graph with such wildly different axes doesn't tell you if things are good or bad. You can't tell what is going on. Furthermore, the state does not make it clear what their method of identifying a new case is: date of symptom onset (as GA does it) or when a new case is confirmed.
Louisiana
Louisiana is using the same method that GA is using for tabulating new cases with one very notable exception: they're not showing data from the previous 12 days. I assume (hope?) that this is because they're trying to avoid the pitfall that has bitten GA so badly: displaying data that is incomplete and misleading.
We hope to provide more state by state analysis of what systems each state is using, how they're performing, and an overall grade of how they communicate their data.