Read on Twitter
it's interesting to compare covid deaths in the US based on date of actual death vs the date the death was reported
the CDC uses the former, covid tracking project and worldometer use the latter
the divergence since aug 1 is becoming very significant
CTP is overstating deaths
the CDC uses the former, covid tracking project and worldometer use the latter
the divergence since aug 1 is becoming very significant
CTP is overstating deaths
for this still not familiar with this issue, it works like this:
the top chart is when deaths are reported. but they get reported on days later than they occurred.
if you then assign them to the actual day of death, they look like the bottom curve
this can go back a LONG way.
the top chart is when deaths are reported. but they get reported on days later than they occurred.
if you then assign them to the actual day of death, they look like the bottom curve
this can go back a LONG way.
this means that deaths reported on late october can actually be deaths from april or july.
as you can see, the real deaths by date are over-weighted in those months.
but the current divergence is getting quite wide.
as you can see, the real deaths by date are over-weighted in those months.
but the current divergence is getting quite wide.
we can zoom in to see it more clearly.
the divergence is getting quite large. the CDC data lags, so the last 2 weeks are meaningfully incomplete.
it takes time to fill it out. but this does lead one to ask: so where are all these CTP deaths coming from?
the divergence is getting quite large. the CDC data lags, so the last 2 weeks are meaningfully incomplete.
it takes time to fill it out. but this does lead one to ask: so where are all these CTP deaths coming from?
the cumulative difference since 8/1 is getting quite large.
it's 22,000 additional deaths, 24% of the CTP total.
all these deaths happened before the week ending 8/1.
the lag is not so long that it's materially affecting august and september or probably h1 of oct.
it's 22,000 additional deaths, 24% of the CTP total.
all these deaths happened before the week ending 8/1.
the lag is not so long that it's materially affecting august and september or probably h1 of oct.
it's death mining. states and hospitals get big additional payouts for finding covid deaths. so they are back searching all the available deaths looking for them.
and it's skewing the trendline to slope up.
reality looks like it's still down.
and it's skewing the trendline to slope up.
reality looks like it's still down.
so be very careful with the date of report data.
it is not actually representative of current epidemiology and seems to be diverging meaningfully.
it's pretty dazzling that at this stage of the game, these sites are still not providing real data.
it is not actually representative of current epidemiology and seems to be diverging meaningfully.
it's pretty dazzling that at this stage of the game, these sites are still not providing real data.
