Shanghai Disneyland will close in effort to contain coronavirus

Brian Noble said:

I will not be surprised to see Michigan pick up some more come mid-to-late October when people start spending a lot more time inside.

But, I also don't expect the rate of increase to be as bad as it has been in some other places, because it is likely that other mitigation factors will be brought to bear. For example, I believe our county is going to switch to a mask requirement in all schools as of this upcoming Monday.

Our school district, the second largest in Michigan, is refusing to enact a mask mandate, using the "logic" of pointing to high case counts in districts where they had mask mandates, and low case counts in districts without them, because, you know, it's apparently impossible that those mask mandates were enacted in districts where case counts were high, and super likely that the global consensus of subject matter experts is totally wrong. Morons.

And of course without the mask mandate, our "Meet & Greet" at the school yesterday was absolutely full of unmasked people everywhere, including teachers and staff, in a county with only ~50% vaccinated. So that means it's inevitable that my kid will be shoulder-to-shoulder with unmasked children of unvaccinated parents.

It’s the old “hey let’s look at cases and not look at deaths” so we can make this virus even more political retort…

When you start looking at deaths from Covid…that logic looks flawed.

This virus doesn’t care who you voted for and where you live. It won’t discriminate…

I'm not even sure what that chart is talking about. Can some explain what "age adjusted" means and what "raw" means? Is this a cumulative chart or a 7 day chart? I'm certainly willing to accept that there's data which contradicts my conclusion, and I like to think I'm willing to change my conclusion if such data surfaces, but I have no idea what I'm even looking at here.

Top 10 states in per capita deaths over last 7 days (by, you know, counting dead people):
MS, LA, FL, AR, NV, TX, SC AL, WY, KY

Bottom 10:
ME, DC, CT, MN, MA, VT, CO, NY, NE, OH

If whatever this metric is, shows no correlation to those fairly simple metrics, then you need to justify why it's more reliable and more important and why we should throw out the mountain of (more straightforward) data which points to a different conclusion.

Also, why are case counts, the primary metric we've been using forever and ever to talk about severity of the pandemic, suddenly "a political tool?" Are per capita deaths also a political tool? Is this mysterious "cumulative age adjusted death rate over the last 18 months" or whatever your chart is about the only non-politicized data available?

Edit: Also, in just trying to understand what was going on I did the raw calculation for MS (pop: ~ 3M) and NY (~pop: 19.5M) and I got a higher number for MS than NY, so I don't even know what raw means here.

Calculations, Sources, and Methods Here:

COVID deaths from CDC "Provisional COVID-19 Death Counts by Sex, Age, and State" https://data.cdc.gov/NCHS/Provisional-COVID-19-Death-Counts-by-Sex-Age-and-S/9bhg-hcku

Populuaton from 2019 (latest estimate) U.S. Census Bureau "State Population by Characteristics: 2010-2019" https://www.census.gov/data/tables/time-series/demo/popest/2010s-state-detail.html

Age-adjustment methodology https://www.nlm.nih.gov/nichsr/stats_tutorial/section2/mod5_age.html

Updated from Len Cabrera's sheet posted here https://alachuachronicle.com/yes-the-covid-mortality-rate-is-higher-in-california-than-florida/

None of your links work.

For no reason at all, I'm reminded of the Gish gallop. From Wikipedia:

The Gish gallop is a term for a rhetorical technique in which a debater attempts to overwhelm an opponent by excessive number of arguments, without regard for the accuracy or strength of those arguments. The term was coined by Eugenie Scott; it is named after the creationist Duane Gish, who used the technique frequently against scientists and other defenders of the scientific fact of evolution. It is similar to a method used in formal debate called spreading.

ApolloAndy said:

None of your links work.

Sorry…I was fixing them when you posted. Looks like they are all working now.

Aamilj said:

The Desantis dunking seems like political zealotry to me. It is almost as if some are hoping that Florida is the WORST to justify their political slant.

Aamilj said:

Updated from Len Cabrera's sheet posted here https://alachuachronicle.com/yes-the-covid-mortality-rate-is-higher-in-california-than-florida/

Linking to an editorial from a small town publication that refers to a mask as a "muzzle" in the headline is proving a point for us, but not the point you think you are proving.

For the record…I’m not implying Florida and Desantis don’t/won’t suck. They might when all is said and done. I am implying that current “dunking” on Desantis seems premature at best…and possibly downright silly when you look at the age-adjusted data. Those into advanced statistics as typically applied to health data…can see that Florida is “middle of the road”…at least in terms of suspected deaths…

BrettV said:

Aamilj said:

Updated from Len Cabrera's sheet posted here https://alachuachronicle.com/yes-the-covid-mortality-rate-is-higher-in-california-than-florida/

Linking to an editorial from a small town publication that refers to a mask as a "muzzle" in the headline is proving a point for us, but not the point you think you are proving.

That 4th link was posted for those who read the first three links and had questions as to why 0-54 year olds are grouped together. Specifically…

“The CDC link above gives the COVID-19 deaths by age groups. Since most (over 90%) of COVID-19 deaths are 55+ and the CDC leaves blanks for single-digit deaths in a category, it’s convenient to use a 0-54 age group, then traditional 10-year groups: 55-64, 65-74, 75-84, and 85+. The population data comes from the U.S. Census Bureau “State Population by Characteristics: 2010-2019,” using 2019 as the most recent data available.”

The content, and frankly the source, of the 4th link are of no particular interest other than to explain why the lowest age group is so wide…

As for “proving a point for us”…? You can only lead a horse to water…

So, having read the articles I have some vague notion of what is meant by age adjustment in some sense. I'd like to see the methodology and math and I'd be interested in double checking Mr. Cabrera's figures (I assume that's where you got the chart you posted above), given that he clearly has a horse in the race, but let's just take it at face value.

The chart you posted uses cumulative deaths over ~18 months. Many of those deaths (particularly in NY and NJ) occurred before we had any idea what was going on and certainly before current treatments were developed. Heck, many of those occurred before anyone had put on a mask. That has nothing to do with politics and everything to do with better science. I bet if you used the same metric over the last six months, you'd see the red states, even the cold ones, rise to the top like they do in per capita deaths. In spite of all that, the chart STILL has correlation to political leaning. Maybe Florida specifically looks okay by this one particular (very roundabout) metric but look at where your usual suspects (AL, MS, LA, GA, AR, OK, TX) land.

And even beyond all that, it seems like the age-adjusted metric, in light of the raw per capita numbers is basically the defense "Sure, Florida is killing a lot of people...but they're mostly old people so they only count half." Which is probably why nobody actually uses that metric. I've been closely following the science of this pandemic and this is the first I've ever heard of this metric. Not a single public health official, politician, or scientist has used it and I'm (not really) curious why. This aside from the fact that old people still spread it just as easily.

It genuinely seems like you're trying to find a single, tiny datum to make Florida in particular look "not awful" and ignoring the mountain of obvious and intuitive evidence (current per capita cases, current per capita deaths) showing that common sense policies (masks and vaccines) save lives and obstructing such policies is dangerous.

Those into advanced statistics

Newsflash: waving vaguely at "advanced statistics" probably isn't going to work given that I'm a professor in a top-10 college of engineering who has taught statistical methods to PhD students. I can slice and dice any pile of data to get the result I want, and I know other people can too. (As an aside, this is partly why the social sciences are currently facing a replication crisis, even aside from the high-profile fabrications.)

In other words, Andy's observation is spot on:

it seems like the age-adjusted metric, in light of the raw per capita numbers is basically the defense "Sure, Florida is killing a lot of people...but they're mostly old people so they only count half."

I bet if you used the same metric over the last six months, you'd see the red states, even the cold ones, rise to the top like they do in per capita deaths. In spite of all that, the chart STILL has correlation to political leaning. Maybe Florida specifically looks okay by this one particular (very roundabout) metric but look at where your usual suspects (AL, MS, LA, GA, AR, OK, TX) land.

That’s a fair critique. I bet the usual suspects would look worse. I would argue that standard practice of counting total deaths, making no adjustment, etc…as if each state’s demographics leads to similar chances for certain outcomes…is not the best way to do it either.

It genuinely seems like you're trying to find a single, tiny datum to make Florida in particular look "not awful" and ignoring the mountain of obvious and intuitive evidence (current per capita cases, current per capita deaths) showing that common sense policies (masks and vaccines) save lives and obstructing such policies is dangerous.

I don’t care about Florida. As stated above the entire state and Desantis may actually suck. I just know that the current per capita cases/deaths that so many are using to dunk on Desantis…are certainly not a fair assessment of the hand he’s been dealt. This virus kills old people at a much higher rate. Florida is full of old people. Certainly some sort of adjustment/curve accounting for the differences in age makes sense.

I would add that there should be an adjustment made for the general health of the state too…but as far as I know…no one has tried. I would guess the general health of the population accounts for much of the south’s poor performance…and Oregon’s heretofore good performance…but that’s speculative.

Regardless of Florida…or any other state you’d like to assert… My opinion is that the chart I provided, while certainly not perfect, is a much more useful look at State’s performance than the typical New York Times data (per capita cases/deaths/etc).

"Sure, Florida is killing a lot of people...but they're mostly old people so they only count half."

Of course no death counts as half. But protecting Duke University and their population subset versus protecting the country’s #1 retirement state do not call for standard (non-adjusted) statistics to explain differing outcomes. Adjustments need to be made based on the population subset in need of protecting.

Thanks for taking the time to read and try to understand what I was saying.

I actually don't disagree with you fundamentally. I've been beating the drum (giggity) for better statistics and better math across this entire pandemic and there is certainly good reason to normalize based on expectation*. But at the same time, unlike something like cancer where there isn't a direct and obvious counter measure (other than "quit smoking," I guess) this pandemic (a) has been known from the beginning to particularly target the elderly and (b) does have an obvious counter measure. That Florida is a particularly susceptible population should mean that its leadership should be even more careful and proactive against the virus and not hide behind "Well, what are you gonna do? They're old."

We don't see nursing homes let residents dry hump and then say, "Well, a bunch of them were expected to die anyway." Instead, they totally lock everybody down and slide the food trays under the door to keep them all safe.

And ultimately, to look at this age adjusted data and conclude that masks don't work or that the virus will get everyone eventually is just bad statistics. Nothing in here justifies that conclusion.

*I'm also not even sure how you come up with an expectation of death by corona virus as a function of age.

Florida sucks, it’s americas wang for a reason. Covid is just one more thing to add to the list of suck for that state.

Yeah, endless summer and not shoveling snow is such a drag!

Jeff said:

Yeah, endless summer and not shoveling snow is such a drag!

I actually agree with this, non-sarcastically.

Kstr 737: You know that I can see the ridiculous post you made, right?