Different trackers (like COVID Tracking Project, Johns Hopkins University and others) have different cut-off times, so cumulative totals reported may be slightly different. Our cut-off time is 12 a.m. to 11:59 p.m. We capture the cumulative totals at throughout the day and end each day with the total as publicly reported at 11:59 p.m. Other trackers end earlier, for example, COVID Tracking Project cuts off at 5:30 CT. This has become less of an issue as the Alabama Department of Public Health moved to once daily updates.

It is two different measurements of the same phenomenon. We also display ADPH’s calculations of daily case counts in addition to ours. To explain how we calculate that number: We capture the cumulative total as displayed on ADPH’s public-facing dashboard from 12 a.m. to 11:59 p.m. each day. That is the main, top-line number you see at the top of their dashboard under “CONFIRMED.”

We take the cumulative total of confirmed cases at 11:59 p.m. on one day and subtract the cumulative total at 11:59 p.m. from the previous day. That is how we arrive at the “daily increase in cases.” We update this number for the current day until 11:59 p.m. For example, the cumulative count at 11:59 p.m. on May 23 was 14,149, according to ADPH’s dashboard. The cumulative count at 11:59 p.m. on May 22 was 13,670, according to ADPH’s dashboard. That is an increase in the cumulative total of 479 in a 24 hour period.

Generally speaking, our daily case counts are not much different from ADPH. But on some days, they are quite different because the top-line number on the public ADPH dashboard may not have updated correctly. ADPH’s daily case total is kept internally and is separate from the cumulative total displayed on the public-facing dashboard. So, for example, they show May 22 ending with a cumulative total of 13,777 on their daily charts. The topline number on the dashboard showed 13,670 at 11:59 p.m.

Though daily increases can be interesting and telling, looking at an average over a longer period is a better reflection of the situation and better reflects trends. This is why we suggest looking at 7-day and 14-day averages instead of individual day increases.