Read on Twitter
This is a deep dive into the Covid situation in Florida, Texas, Arizona, and California.
Afterall, we're seeing a huge jump in cases!
But let's look behind the case numbers being reported and let's see what's really going on.
Afterall, we're seeing a huge jump in cases!
But let's look behind the case numbers being reported and let's see what's really going on.
First, a big picture look: Daily deaths by state
-clearly the deaths do not match the cases overall
-California has more deaths despite having far fewer cases than Texas and about the same number as Florida
Next up, the deep dive
-clearly the deaths do not match the cases overall
-California has more deaths despite having far fewer cases than Texas and about the same number as Florida
Next up, the deep dive
What I did is I put together three graphs for each state:
1) Tests vs Cases vs Deaths (this is essentially Tests vs Deaths since Cases are so tiny by comparison)
2) Tests vs Cases
3) Cases vs Deaths
I added re-opening data to see if change in behavior affected trends
Data: CTP
1) Tests vs Cases vs Deaths (this is essentially Tests vs Deaths since Cases are so tiny by comparison)
2) Tests vs Cases
3) Cases vs Deaths
I added re-opening data to see if change in behavior affected trends
Data: CTP
Florida - Tests vs Cases vs Deaths
-Reopenings do not affect deaths
-No association between testing and deaths
-Reopenings do not affect deaths
-No association between testing and deaths
Florida Tests vs Cases
-Surge in cases after Phase 2 Reopening - Rest, Bars, Theaters 50% capacity
-Testing also surges after Phase 2
-Testing dump after Phase 1 - not sure how to interpret that but it didn't affect cases...I wonder how they record cases identified by this dump?
-Surge in cases after Phase 2 Reopening - Rest, Bars, Theaters 50% capacity
-Testing also surges after Phase 2
-Testing dump after Phase 1 - not sure how to interpret that but it didn't affect cases...I wonder how they record cases identified by this dump?
Florida Cases vs Deaths
-No observable correlation between cases and deaths
-Reports that age of cases is falling
-Since reopening, deaths are stable and low (30-40 deaths/week is nothing)
-No observable correlation between cases and deaths
-Reports that age of cases is falling
-Since reopening, deaths are stable and low (30-40 deaths/week is nothing)
Florida Conclusions:
-I see no reason for panic in Florida
-Since the deaths are very low, what we see is more like static than an observable trend
-Large growth in cases started 2.5 weeks ago - not reflected in increased deaths - these cases coincide with BLM protests!
-I see no reason for panic in Florida
-Since the deaths are very low, what we see is more like static than an observable trend
-Large growth in cases started 2.5 weeks ago - not reflected in increased deaths - these cases coincide with BLM protests!
Texas - Tests vs Cases vs Deaths
-No correlation between testing and death
-Deaths very low for a state the size of Texas
-Testing has increased consistently
-No correlation between testing and death
-Deaths very low for a state the size of Texas
-Testing has increased consistently
Texas - Tests vs Cases
-Cases did not increase after Phase 1
-Pretty direct correlation between testing and cases
-Big increase in case growth started 14 days ago...larger relative growth than testing
-This coincides with BLM protests
-Cases did not increase after Phase 1
-Pretty direct correlation between testing and cases
-Big increase in case growth started 14 days ago...larger relative growth than testing
-This coincides with BLM protests
Texas - Cases vs Deaths
-No correlation between cases and deaths
-Rapid growth in cases started 2 weeks ago - BLM?
-Increased cases in last 2 weeks aren't reflected in deaths
-No correlation between cases and deaths
-Rapid growth in cases started 2 weeks ago - BLM?
-Increased cases in last 2 weeks aren't reflected in deaths
Texas Conclusions:
-Texas deaths on the graph also look like static because of low daily deaths
-Second rise (not wave) in deaths that I predicted last week did not materialize
-Its very possible that young people were infected at BLM protests which is why deaths don't also rise
-Texas deaths on the graph also look like static because of low daily deaths
-Second rise (not wave) in deaths that I predicted last week did not materialize
-Its very possible that young people were infected at BLM protests which is why deaths don't also rise
Arizona - Tests vs Cases vs Deaths
-On first look it seems that Arizona is experiencing a second rise (not wave) in deaths
-Phase 1 could be linked to the second rise of deaths...second rise started 3 weeks after reopening
-On first look it seems that Arizona is experiencing a second rise (not wave) in deaths
-Phase 1 could be linked to the second rise of deaths...second rise started 3 weeks after reopening
Arizona Tests vs Cases
-Testing dump in mid May
-Pretty direct correlation between testing and cases
-Cases started rising at the end of May consistent with increased testing,
-but a sharper growth started two weeks ago which grows faster than testing - aligns with BLM protests
-Testing dump in mid May
-Pretty direct correlation between testing and cases
-Cases started rising at the end of May consistent with increased testing,
-but a sharper growth started two weeks ago which grows faster than testing - aligns with BLM protests
Arizona Cases vs Deaths
-Deaths are still very low so we might be looking at static, but it seems that there is a spike in deaths which started a little over a month ago...
-There is also a huge surge in the last week which is not consistent with case growth two weeks earlier
-Deaths are still very low so we might be looking at static, but it seems that there is a spike in deaths which started a little over a month ago...
-There is also a huge surge in the last week which is not consistent with case growth two weeks earlier
Arizona Conclusions
-Despite increased testing, cases didn't rise in the period when the new increase in deaths would have been infected (5/13-5/29)
-There is definitely an increase, but Arizona is not behaving like the other three states
-Alternative explanations?
-Despite increased testing, cases didn't rise in the period when the new increase in deaths would have been infected (5/13-5/29)
-There is definitely an increase, but Arizona is not behaving like the other three states
-Alternative explanations?
California Tests vs Cases vs Deaths
-Deaths consistently decline after March 24th
-Reopenings did not affect death
-Testing has increased consistently (note, Tests are reported with a 14 day rolling average to smooth out those two big bumps)
-Deaths consistently decline after March 24th
-Reopenings did not affect death
-Testing has increased consistently (note, Tests are reported with a 14 day rolling average to smooth out those two big bumps)
California Tests vs Cases
-There is a direct correlation between testing and cases
-Exception is last two week where cases have grown much faster than testing. This is again consistent with BLM protests.
-There is a direct correlation between testing and cases
-Exception is last two week where cases have grown much faster than testing. This is again consistent with BLM protests.
California Cases vs Deaths
-Deaths do not correlate with cases
-Surge in case over the last two weeks is not yet reflected in deaths. If they are BLM related, will probably not affect deaths because of young age of population
-Deaths do not correlate with cases
-Surge in case over the last two weeks is not yet reflected in deaths. If they are BLM related, will probably not affect deaths because of young age of population
California Conclusions
-Only state where a pattern emerges for deaths - Sharp rise followed by a steady decline
-Testing and cases strongly correlated
-Again, a large surge in cases in the past two week - probably BLM related
-Only state where a pattern emerges for deaths - Sharp rise followed by a steady decline
-Testing and cases strongly correlated
-Again, a large surge in cases in the past two week - probably BLM related
Overall Conclusions
-No reason for panic
-The surge in cases for all states in the last two weeks is probably BLM related
-Overall, cases increase with testing
-Overall, deaths do not increase with more new cases
-Arizona deserves a deeper dive to make sense of the second rise
-No reason for panic
-The surge in cases for all states in the last two weeks is probably BLM related
-Overall, cases increase with testing
-Overall, deaths do not increase with more new cases
-Arizona deserves a deeper dive to make sense of the second rise
Somebody asked me why Minnesota wasn't spiking after BLM started there: https://twitter.com/gummibear737/status/1277309398365417472?s=20
