Rebel A. Cole Profile picture
Follow the facts. Report the truth. Former Fed Staff Economist. 35 years as Data Scientist. Personal account. Retweets do not equal endorsements.
24 Nov
1/ Florida Covid Update: No. 23, 2020

Number of cases has been rising largely in lockstep with number of tests.

Correlation coefficient between two series: > 0.80
2/ Florida Covid Update: No. 23, 2020

Positive percentage (7-day MA) had been rising since Oct. 7 from low of 4.4% to 8.3% on Nov. 18.

Falling since then, today at 7.5%.

July 8 high was 16.1%.

The snowbirds have returned with more than just their buying power.
3/ Florida Covid Update: No. 23, 2020

Hospitalization data continues to show stability.

Occupancy of all beds at 74%, in line with past several months.

Same for occupancy of ICU beds at 76%.
Read 12 tweets
1 Nov
1/ Today's Palm Beach Post newspaper's latest #PanicPorn from its resident "fear-mongerer" Jane Musgrave aka @pbpcourts.…
2/ 800,000 have tested positive.

Current positive percentage is more about 5% in FL so that means more than 1 million of 21.5 million population are CURRENTLY INFECTED.

Most are asymptomatic.

Jane misses this "small" fact.
3/ Serology tests indicate that about 20% of Floridians have been infected with Covid-19.

That is not 1 million.

That is 4 million.
Read 12 tweets
30 Oct
1/ Palm Beach County, FL Update:

Today, the @PBPost published yet another #PanicPorn article penned by @pbpcourts Jane Musgrave, trying to paint our situation as grim.
"We're in trouble in Florida," opines one "expert."

"This comes when you open too soon," scolds another.
2/ So what are they worrying about?

A rise in the number of "cases."

Ok, so let's look:

Cases have risen from recent 7-day average low of 115 up to 309.

But much of this is the result of increased testing.
From 3,460 to 6,181.
3/ Much more relevant is the positive percentage.

This metric, on a 7-day average to smooth daily spikes, had averaged less than 5% from Aug. 23 until Oct. 26 when it rose above 5%.

Has been less than 6% each day since.

Worth keeping an eye on it as the snowbirds return.
Read 9 tweets
21 Oct
1/ Florida Covid-19 Situation:


Appear to be rising by "date of lab datadump."

But continue to fall by "date of event," which for most cases is the date of the PCR test.

Labs are dumping old tests to create the illusion of "rising cases"?
2/ Florida Covid-19 Situation:


Covid-occupied hospital beds hit new low on Sunday under 2,000 after mid-July peak of 9,500.

Slight uptick on Mon-Tue as happens each week.

Relative to "event date," hospitalizations continue to decline to new low of under 100.
3/ Florida Covid-19 Situation:


By date of death, continue to decline. Could miss a recent uptick.

Relative to event date, continue to decline in a big way.

Strong evidence that death are really continuing to decline.
Read 8 tweets
18 Oct
1/ Today's Florida Covid-19 Update:

Positive percentage (7-day rolling average) below 5% for fourth straight week.

Slight blip attributable to Nov. 9 lab reporting debacle that inflated daily positive percentage to 7.88%.
2/ Hospitalizations:

Covid-occupied hospital beds fall to new low of 1,996, down from mid-July peaks of 9,500+.

First time below 2,000 since public reporting began.

Less than 5% of ~42,000 total occupied hospital beds.
3/ Deaths:

Deaths by "date of death" continue to decline from early August peak, but state bureaucrats continue to create new "Covid deaths" by matching old death certificates with old Covid cases.
Read 10 tweets
17 Oct
1/ Today's Florida Covid-19 testing data update.

After a disastrous lab "data dump," we are "officially back below 5% positive percentage.
2/ Testing by month and then by day for past 14 days.

Do you see a "spike" from re-openings schools and colleges?


Me neither.

Because there wasn't one.

So much for that #PanicPorn headline.
Hospitalization data.

Covid-occupied beds down from peak of 9,500+ in mid-July to less than 2,100.

No discernable rise in past few weeks.

Maybe this is as low as it goes with bogus PCR testing "false positives."
Read 7 tweets
24 Sep
1/ #PaniPorn fear-mongerer Jane Musgrave @pbpcourts of the trashy @pbpost spreads misinformation about Covid deaths in FL & PBC.

"After a 4-day lull, COVID-19 deaths soared in Florida on Wednesday with 203 fatalities reported across the state, including a near-record 22 in PBC" ImageImage
2/ @pbpcourts reports:

"The number of deaths statewide is more than were reported in the past four days combined, and the most logged since Sept. 10, when 213 people were reported dead."


@HealthyFla saved them up for your #PanicPorn headline.

9, 21 reported Sun-Mon. Image
3/ @pbpcourts reports:

"However, the reporting of deaths is often delayed by two weeks or more. It is unlikely that many of the deaths reported on Wednesday occurred within the past 24 hours."


You KNOW SO, yet you LIE to your readers.

Stop LYING, Jane Musgrave. ImageImage
Read 4 tweets
18 Sep
1/ Why @COVID19Tracking has Florida testing data WRONG.

I think I've figured out what the simpletons at CTP are doing wrong.

This chart appears on the first page of each day's State Report from @HealthyFla

It shows CUMULATIVE number of "persons tested" for Covid in FL.
2/ "Data scientists" at CTP are simply subtracting cumulative numbers of residents plus non-residents that tested positive and negative.

Then calculating daily percent positivity based on these two numbers.

So, for today:

3,244 / (23,695+3,244) =
3,244 / 26,950 = 12.08%.
3/ Because these are "cumulative number" of people tested, it DELETED persons tested more than once.

So if I got tested last month and again yesterday, my test yesterday is not counted.

I only get counted once for negative and positive.

Think about how many get tested > 1 time
Read 10 tweets
12 Sep
If you have time to read ONLY ONE article on Covid testing, I recommend this one.

It is FANTASTIC information!

And its been out there since early April.

Was the COVID-19 Test Meant to Detect a Virus?… via @tracybeanz
Read 7 tweets
12 Sep
1/This is STUNNING.

Rather than use "actual cases" to determine "false positive" rates, labs are using comparisons from different laboratories.

No lab knows what a true case is.

But ALL of the FDA EUAs are based upon comparisons with other lab results.… Image
So we are assessing "false positives" based upon other lab results that also cannot identify "false positives."


Because "experimenters" supposedly "cherry-pick" the samples.


"There are 2 ways to find out what false positive rate of a test is.
One is to run test on cases where there is certainty about the diagnosis and figure out how many results are wrong."
Read 7 tweets