Country

Philippines, Ecuador, Chile, Gautemala and others

Domain

Cases

Which data is inaccurate or missing?

The cumulative count of deaths and cases doesn’t match for a few countries…And the smoothed columns could be corrected too, I believe...

Why do you think the data is inaccurate or missing?

IF you sum up the new cases of every day, the result won’t be the same as in the total column. It would be less. In your web page it’s hard to notice it, as these columns are used in different graphs, or different filters, better said… Nonetheless, if you take a look at the cumulative new cases of Philippines, you will see that, in august 2023, the graph shows a small descending. Small is a way to speak, since the difference between one day and the next one is of something around 65 thousand… This is the biggest mistake. In other countries the differences are much smaller, most of the time less than one hundred.
I know that there were some issues with the report methods; people that were wrongly diagnosed and so on, but, in the case of Philippines, the number is just too big to be trusted, and, in the other cases, to small to worth the trouble. There is no point in rest 20 cases because of a diagnose mistake, since there probably are a lot of cases that weren’t diagnosed at all… I believe that, most probably, is just a sum mistake… In the days in which the errors occur, always appears a null value in the new cases column, and a descending in the same row of the total column with respect to the previous day (or week or whatever). Anyway, you are better informed than me and maybe there is some other reason for this to happen (though hardly a convincing one), and it’s pretty easy to check it in the csv…
Also, if you sum up the cumulative cases and deaths for the income brackets, the result won’t be the same that the cumulative for the world, and it should be it, since every country falls into one of those categories…
The other thing I found was that, in the smoothed columns, you put the weekly average, since the new cases are reported weekly. But if, for instance, the new cases of today were 35, you would put seven in the smoothed columns, which is all right, but you would put it for the next seven days… It shouldn’t be put in the previous ones?