covid: why is the conversation over total case numbers? and not death rates?

Politics and religion: two polarizing topics that deserve their own little place
KITNooga
Posts: 6619
Joined: September 19th, 2017, 8:22 am

Re: covid: why is the conversation over total case numbers? and not death rates?

Post by KITNooga » June 30th, 2020, 9:30 pm

xtrawildcat wrote:
June 30th, 2020, 5:02 pm
Do you understand the lag time in the Georgia Department of Health new case data or not.

You can study the numbers and figure it out yourself if you don't beleive the AJC. And the AJC Covid data page lists both sets of numbers. Use the GDH website numbers if you like. They also explain the 14 day delay before numbers are accurate.

But it is a fact Georgia case numbers have more than doubled in the last week and the number of deaths is a lagging indicator.
again, here's the data. from GDH website. as of yesterday.

Image

what makes you think I don't GET the lag issue? it's CLEARLY identified in the graphs. there is the VERIFIED data 'in bold' and there is the 'yet to be verified' data that's 'translucent'. it's all 'there'. it's labeled as the '14 day window' right there on the chart for heaven's sake.

I SEE the verified data and also the potential 'future trend' of things to come. Have been watching this site for a long time now.

fact, here's the FIRST graph I shared with all... from well over a month ago
Image

I understand the lag time numbers. I am wondering if YOU yourself do. it's such a simple concept.

so, again, what do we SEE? first peak. then testing ramps up. we test more. we are testing more. that is a clear driving factor of more cases being reported, observed. along with outcomes of riots/peaceful protest gatherings. but what's KEY is that deaths ARE on the decline.

it's a FACT that total deaths are going up. statistically total deaths just can not EVER go down. RATES however can and ARE. So the numbers clearly indicate.

agree? disagree? what do those charts say to you about death RATES as a reality. and as a function of total tests? perhaps, you look at the charts and are ASTOUNDED that clearly there are so man y more deaths than reported cases? (sarcasm clearly) did you take note not of the shape/amplitude of the relative graphs but also the scale of the y-axis? (again, sarcasm). but when you think I don't 'get the data' tell me. what am I not getting that you 'clearly are'.

I understand the numbers. what are YOU missing? tell me. enlighten me. 'learn' me.

here's a few interesting points to also consider: when was George Floyd killed/murdered? when did protest gatherings start/take place (dates)?
George Floyd died on May 25. protests started right after. then Rayshard Brooks was shot (atlanta). keep tabs on 'that' date and the multi-week lag yet to show up on the chart.
Last edited by KITNooga on June 30th, 2020, 9:44 pm, edited 3 times in total.

Zoltar
Posts: 1639
Joined: September 11th, 2017, 5:01 pm
Location: Omaha, NE

Post by Zoltar » June 30th, 2020, 9:30 pm

powercat5000 wrote:
June 30th, 2020, 8:39 pm
stlcatfan wrote:
June 30th, 2020, 7:03 pm
https://www.youtube.com/watch?v=4H-6HTKNgV0

The only part he forgot to mention is that the virus is not contagious when you are at left-wing protests and riots. I guess he also could have said we will close churches and small businesses, but keep liquor stores, weed shops, and abortion clinics open.
How about 'Hydroxycloroquine is safe to take for malaria prevention in children, adults, and even pregnant women....but if taken to prevent or cure covid it will kill you'
Completely different dosage schedules. Once a week to twice a day. You should do a little research before posting something that is so obviously incorrect.
Win the dang day!

ChemicalKat
Posts: 4972
Joined: September 11th, 2017, 9:18 am

Post by ChemicalKat » June 30th, 2020, 10:04 pm

KITNooga wrote:
June 30th, 2020, 9:30 pm
xtrawildcat wrote:
June 30th, 2020, 5:02 pm
Do you understand the lag time in the Georgia Department of Health new case data or not.

You can study the numbers and figure it out yourself if you don't beleive the AJC. And the AJC Covid data page lists both sets of numbers. Use the GDH website numbers if you like. They also explain the 14 day delay before numbers are accurate.

But it is a fact Georgia case numbers have more than doubled in the last week and the number of deaths is a lagging indicator.
again, here's the data. from GDH website. as of yesterday.

Image

what makes you think I don't GET the lag issue? it's CLEARLY identified in the graphs. there is the VERIFIED data 'in bold' and there is the 'yet to be verified' data that's 'translucent'. it's all 'there'. it's labeled as the '14 day window' right there on the chart for heaven's sake.

I SEE the verified data and also the potential 'future trend' of things to come. Have been watching this site for a long time now.

fact, here's the FIRST graph I shared with all... from well over a month ago
Image

I understand the lag time numbers. I am wondering if YOU yourself do. it's such a simple concept.

so, again, what do we SEE? first peak. then testing ramps up. we test more. we are testing more. that is a clear driving factor of more cases being reported, observed. along with outcomes of riots/peaceful protest gatherings. but what's KEY is that deaths ARE on the decline.

it's a FACT that total deaths are going up. statistically total deaths just can not EVER go down. RATES however can and ARE. So the numbers clearly indicate.

agree? disagree? what do those charts say to you about death RATES as a reality. and as a function of total tests? perhaps, you look at the charts and are ASTOUNDED that clearly there are so man y more deaths than reported cases? (sarcasm clearly) did you take note not of the shape/amplitude of the relative graphs but also the scale of the y-axis? (again, sarcasm). but when you think I don't 'get the data' tell me. what am I not getting that you 'clearly are'.

I understand the numbers. what are YOU missing? tell me. enlighten me. 'learn' me.

here's a few interesting points to also consider: when was George Floyd killed/murdered? when did protest gatherings start/take place (dates)?
George Floyd died on May 25. protests started right after. then Rayshard Brooks was shot (atlanta). keep tabs on 'that' date and the multi-week lag yet to show up on the chart.
Positive tests, as a percent of total tests, is rising. This surge is not the result of more testing but finding sick people easier.

xtrawildcat
Posts: 5789
Joined: February 15th, 2016, 10:41 pm

Post by xtrawildcat » June 30th, 2020, 11:31 pm

I never commented on Georgia death rates except to note we need to see where they go over the next 2 to 3 weeks due to the higher daily case counts in the last week.

The first chart you first posted shows less than 50 cases on May 12. The chart today shows between 500 and 600 cases .

The Georgia department of health chart shows 30 cases today.
The AJC and other Covid trackers shows they had 1874 cases today,

To me, GDH number of 30 is completely meaningless.

IMO, the 1874 cases are a positive sign for Georgi9a as they had over 2200 new Covid cases for the three prior days. But only one day. Now we can see what happens over the next week or so. Maybe all the talk about case numbers and the need to wear masks will have a positive effect.

I don't care which set of numbers you use. That's up to you. I'll continue to use cases reported on the day they are discovered as every other state and the rest of the world is doing.

User avatar
stlcatfan
Posts: 2284
Joined: September 9th, 2017, 7:45 am

Post by stlcatfan » June 30th, 2020, 11:54 pm

powercat5000 wrote:
June 30th, 2020, 8:39 pm
stlcatfan wrote:
June 30th, 2020, 7:03 pm
https://www.youtube.com/watch?v=4H-6HTKNgV0

The only part he forgot to mention is that the virus is not contagious when you are at left-wing protests and riots. I guess he also could have said we will close churches and small businesses, but keep liquor stores, weed shops, and abortion clinics open.
How about 'Hydroxycloroquine is safe to take for malaria prevention in children, adults, and even pregnant women....but if taken to prevent or cure covid it will kill you'
You're right. I forgot about that one.

KITNooga
Posts: 6619
Joined: September 19th, 2017, 8:22 am

Post by KITNooga » July 1st, 2020, 12:00 pm

Image
ChemicalKat wrote:
June 30th, 2020, 10:04 pm
Positive tests, as a percent of total tests, is rising. This surge is not the result of more testing but finding sick people easier.
and thus, exactly why I put both graphs up SIDE BY SIDE.

my point? testing postive is not the end of the world. clearly, the graph right next to it with DEATHS shows that the two (positive tests, deaths) are in no way correlated.

if you look at the data, you see positive results trending/moving UP and on the DEATH chart you see, same time frame, death trending DOWN.

yes? right?

oh sweet lord

ChemicalKat
Posts: 4972
Joined: September 11th, 2017, 9:18 am

Post by ChemicalKat » July 1st, 2020, 1:29 pm

You don't see a correlation because you didn't assign a lag. I plotted Daily New Cases and deaths for the entire country here and then assigned a 7 day lag. The correlation for the first graph is already high at .73. The correlation goes up with the 7 day lag.

Here's another hypothesis: the lag will be greater this time because most of the currently infected are young people. There will need to be a second round of spread to get the virus from the young to the vulnerable, increasing the lag. Hopefully, and this is entirely possible, the young people of this country realize what they just did and decide not to visit grandma/grandpa, making the death rate smaller than the first round of death in New York.
A.jpg
A.jpg (90.88 KiB) Viewed 43 times
B.jpg
B.jpg (88.51 KiB) Viewed 43 times

KITNooga
Posts: 6619
Joined: September 19th, 2017, 8:22 am

Post by KITNooga » July 1st, 2020, 1:32 pm

ChemicalKat wrote:
July 1st, 2020, 1:29 pm
Here's another hypothesis

hypothesis has to be proved. c'mon CK, what's the NULL hypothesis?

and heck, while you're at it: tell me, is everything constant in terms of all the variables contributing to the data? compare/contrast the first section of the graphs with the last section of the graph.

and?

xtrawildcat
Posts: 5789
Joined: February 15th, 2016, 10:41 pm

Post by xtrawildcat » July 1st, 2020, 1:41 pm

KITNooga wrote:
July 1st, 2020, 1:32 pm
ChemicalKat wrote:
July 1st, 2020, 1:29 pm
Here's another hypothesis

hypothesis has to be proved. c'mon CK, what's the NULL hypothesis?

and heck, while you're at it: tell me, is everything constant in terms of all the variables contributing to the data? compare/contrast the first section of the graphs with the last section of the graph.

and?
The concern should be we have just had three days at 45,000 new cases per day. When those cases are resolved, we may feel better about the number of deaths. Lets hope we do. But a little early to do that.

ChemicalKat
Posts: 4972
Joined: September 11th, 2017, 9:18 am

Post by ChemicalKat » July 1st, 2020, 2:22 pm

KITNooga wrote:
July 1st, 2020, 1:32 pm
ChemicalKat wrote:
July 1st, 2020, 1:29 pm
Here's another hypothesis

hypothesis has to be proved. c'mon CK, what's the NULL hypothesis?

and heck, while you're at it: tell me, is everything constant in terms of all the variables contributing to the data? compare/contrast the first section of the graphs with the last section of the graph.

and?
Yes, I know a hypothesis has to be proven. That's why I said hypothesis. We'll see in a week or two. Again, don't bank your position (whatever that position is, still not clear) on decreasing death rates.

Post Reply