Coronavirus Sars-CoV-2/Covid-19 Megathread

[The Minsky Moment]

How about numbers over the past week for Florida:
132, 112, 156, 128, 90, 87, 90, 134.

929 dead Floridians in a week.  Nothing to see here?

[The Minsky Moment]

The last time I had the same discussion with AR was a month ago, on June 23 when the deaths for the prior week were 300.  So the weekly deaths tripled in a month.  The "no worries, death rates are still low" viewpoint was not vindicated. 

[katmai]

I don't know why you people bother with the puerile dickhead.

[garbon]

Florida (and Georgia) are reaching ICU bed capacity. Don't worry. Those death numbers will start going up.

I understand. Disaster for the irresponsible states is always just a little further off.

Would either of you be brave enough to put numbers on this?  What number of deaths will signify death numbers going up, which won't ever happen if you're an optimist or which will happen if you're an optimist?

You brave enough to use numbers, bro?

[garbon]

I don't know why you people bother with the puerile dickhead.

Amen.

[The Brain]

I don't know why you people bother with the puerile dickhead.

I have a name you know.

[Eddie Teach]

Florida (and Georgia) are reaching ICU bed capacity. Don't worry. Those death numbers will start going up.

I understand. Disaster for the irresponsible states is always just a little further off.

Would either of you be brave enough to put numbers on this?  What number of deaths will signify death numbers going up, which won't ever happen if you're an optimist or which will happen if you're an optimist?

You brave enough to use numbers, bro?

He's always looking for action.

[Grey Fox]

I don't know why you people bother with the puerile dickhead.

I have a name you know.

and we don't know what it is.

[DGuller]

How about numbers over the past week for Florida:
132, 112, 156, 128, 90, 87, 90, 134.

929 dead Floridians in a week.  Nothing to see here?

Same question to you.  Cherrypicking the numbers after the fact is easy, making a prediction ahead of time is what separates the bullshitters from the bullshit.

[The Minsky Moment]

How about numbers over the past week for Florida:
132, 112, 156, 128, 90, 87, 90, 134.

929 dead Floridians in a week.  Nothing to see here?

Same question to you.  Cherrypicking the numbers after the fact is easy, making a prediction ahead of time is what separates the bullshitters from the bullshit.

First it is not cherry picking to look at the numbers from last week and compare it the same period when we last had the same discussion.  If you want to use 10 days, 2 weeks, 3 weeks or the full month, fine.  Whatever time period you pick it's going to look similar.

Second, making a prediction ahead of time is what separates the lucky charlatans from the unlucky charlatans.

What happens in the future depends on behavioral response.   NY has been having around 5 deaths per day.  Presumably not because the mayor and governor are great geniuses - its the same guys who were in charge in April.  It's because New Yorkers are behaving differently now then they were in March.

Anyone who says they can predict behavioral response in a matter as complex as this is a charlatan.

I rely on data and we don't have future data.  I have no idea what will happen in the future.

[alfred russel]

Would either of you be brave enough to put numbers on this?  What number of deaths will signify death numbers going up, which won't ever happen if you're an optimist or which will happen if you're an optimist?

Absolutely. I've been willing to the entire time. I made two proposals to Tim earlier to bet him on the ultimate infected fatality rate and the minimum number of deaths he projected by the start of May, he passed on both (and would have lost both--which shows his confidence in ).

In April I highlighted a group of states working to open up early as evidence people were fed up with the situation: South Carolina, Tennessee, Alabama, Georgia, and Florida. Malthus suggested a group of Trumpist governors doesn't prove anything.

My proposal: by the end of the year, the per capita death rate in those states will be below the national per capita death rate. The calculation will be total deaths in all those states divided by the population in all those states.

[alfred russel]

How about numbers over the past week for Florida:
132, 112, 156, 128, 90, 87, 90, 134.

929 dead Floridians in a week.  Nothing to see here?

That is bad. It is a bad week for Florida.

But for perspective. Back in March, Fate was telling us we would be seeing "2,000 dead in a day in New York City for a week straight". Now you are telling me that we didn't even hit half that in all of Florida over a week and we're supposed to freak out? The goal posts have significantly moved.

Florida is still well below the national average in deaths per capita. Maybe it is just moving to the national average and these are the deaths other places had earlier that it is getting now. Maybe it has entered uncontrolled exponential growth of the virus and it is going to be NYC 2.0 on a statewide basis. Time will tell.

[alfred russel]

What happens in the future depends on behavioral response.   NY has been having around 5 deaths per day.  Presumably not because the mayor and governor are great geniuses - its the same guys who were in charge in April.  It's because New Yorkers are behaving differently now then they were in March.

Also a significant number of them have been infected already which will make control of the virus a lot easier and reduce spread.

Keep in mind that if you flatten a curve of cases over time, the area under the flattened curve is the same as the spiked curve.

[DGuller]

First it is not cherry picking to look at the numbers from last week and compare it the same period when we last had the same discussion.  If you want to use 10 days, 2 weeks, 3 weeks or the full month, fine.  Whatever time period you pick it's going to look similar.

The cherrypicking is in what statistics you choose to use, and how you frame it.  You frame it as a three-fold increase in deaths.  AR could frame it as "so, you were saying that these dumbfucks would be dying by the droves, and what you meant by that is that they would be dying at 20% of the rate New Yorkers were dying at?"

Second, making a prediction ahead of time is what separates the lucky charlatans from the unlucky charlatans.

There is obviously an element of luck involved in predicting the future, but at the same time it forces you to put something up, because you're setting up a quantifiable evaluation metric of your ability to analyze things objectively.  For example, you'll have to balance the need to be right in your prediction when AR will come back to quote it in two months, against the need to make a sufficiently dire prediction to prove the point that Floridians are dumbfucks.  If you're going to predict that 200 deaths per day is as bad as it'll get, then you'll already concede the point.  That's why I suspect that you and everyone up against AR will not actually go through with putting up something measurable.

What happens in the future depends on behavioral response.   NY has been having around 5 deaths per day.  Presumably not because the mayor and governor are great geniuses - its the same guys who were in charge in April.  It's because New Yorkers are behaving differently now then they were in March.

Anyone who says they can predict behavioral response in a matter as complex as this is a charlatan.

I rely on data and we don't have future data.  I have no idea what will happen in the future.

You can rely on data precisely in order to gain some insight into what is likely to happen in the future.  I hear some people actually make a living out of it.

[DGuller]

In April I highlighted a group of states working to open up early as evidence people were fed up with the situation: South Carolina, Tennessee, Alabama, Georgia, and Florida. Malthus suggested a group of Trumpist governors doesn't prove anything.

My proposal: by the end of the year, the per capita death rate in those states will be below the national per capita death rate. The calculation will be total deaths in all those states divided by the population in all those states.

How much money are you willing to bet, and at what odds?