Tell me some good news Bookmark and Share

The bulk of the musical talk

Re: Tell me some good news

Postby goodgigs » Mon Mar 23, 2020 11:05 pm

War, like sports, does NOT build character - it reveals it !
This goes to show that there good people everywhere: you just have to find them.
We won WWII the same way. 8)
Brian "Goodgigs" Kane
User avatar
goodgigs
6 valves
6 valves
 
Posts: 2177
Joined: Mon Mar 22, 2004 11:30 pm
Location: Santa Cruz CA

Re: Tell me some good news

Postby Wu299 » Thu Mar 26, 2020 2:23 pm

Have we run out of good news in just 4 pages of replies?

The virus actually made my life much easier here in the Czech republic – my lifestyle is now officially everyone's lifestyle, I am comfortably working from home and cleaning up several parts of my house in the spare time. Most of my colleagues appear to be happy about spending more time with their families rather than commuting and being whole day with random strangers, as well.

Aaaand... I play my new (to me) Willson rotary F I bought two months ago! Now that I'm thinking about it, perhaps I might start a separate thread with pictures, etc., as I haven't seen many of these on tubenet. :tuba:
Wu299
bugler
bugler
 
Posts: 63
Joined: Thu Apr 22, 2010 2:51 pm

Re: Tell me some good news

Postby Three Valves » Thu Mar 26, 2020 2:35 pm

Gas is under $2/gallon for regular everywhere.

All the WAWAs are open.

This is like being retired, only without the planning or potential income one may have otherwise expected!!
Who needs four valves??

Mack Brass Artiste
MackBrunner 210L
TU25
User avatar
Three Valves
6 valves
6 valves
 
Posts: 3952
Joined: Wed Jan 07, 2015 10:44 am
Location: The Land of Pleasant Living

Re: Tell me some good news

Postby bort » Thu Mar 26, 2020 3:01 pm

I talked to someone on the phone today.

Aah, telephones... The original social distancer!
Alexander 163 CC
User avatar
bort
6 valves
6 valves
 
Posts: 10624
Joined: Wed Sep 22, 2004 11:08 pm
Location: Minneapolis, Minnesota

Re: Tell me some good news

Postby bort » Thu Mar 26, 2020 3:06 pm

Wu299 wrote:Have we run out of good news in just 4 pages of replies?

The virus actually made my life much easier here in the Czech republic – my lifestyle is now officially everyone's lifestyle, I am comfortably working from home and cleaning up several parts of my house in the spare time. Most of my colleagues appear to be happy about spending more time with their families rather than commuting and being whole day with random strangers, as well.

Aaaand... I play my new (to me) Willson rotary F I bought two months ago! Now that I'm thinking about it, perhaps I might start a separate thread with pictures, etc., as I haven't seen many of these on tubenet. :tuba:


Ooh, yes please!

And tell me what you think of the rotax valves. I liked the valves on my old Willson rotary CC just okay.... But they we're kind of slow and heavy. I popped then out and sent them to Martin Wilk, who expertly cut them down to make them into regular rotary valves. That helped the action and response of the tuba quite a bit. The ergonomics were never great for me, but otherwise, it was a great tuba. If only Willson could have figured out the rotary valves a little better.

If all is well on the rotary F though... I bet it's a sweet tuba!
Alexander 163 CC
User avatar
bort
6 valves
6 valves
 
Posts: 10624
Joined: Wed Sep 22, 2004 11:08 pm
Location: Minneapolis, Minnesota

Re: Tell me some good news

Postby bloke » Thu Mar 26, 2020 3:36 pm

It's not convenient (20 minutes to the south of me), but there is some .69⁹¢ gasoline for sale, around here.

Hey yankees:
I'm just kidding. Everything around here is crazy-expensive...You would just hate it. You wouldn't even enjoy visiting, so...
User avatar
bloke
musician/technician/innovator
musician/technician/innovator
 
Posts: 45711
Joined: Sat May 08, 2004 6:04 pm
Location: western Tennessee

Re: Tell me some good news

Postby Three Valves » Thu Mar 26, 2020 3:58 pm

If they would finally take the booze out of our gas and make sanitizer out of it, this whole megillah could wind up with a positive spin on it!!
Who needs four valves??

Mack Brass Artiste
MackBrunner 210L
TU25
User avatar
Three Valves
6 valves
6 valves
 
Posts: 3952
Joined: Wed Jan 07, 2015 10:44 am
Location: The Land of Pleasant Living

Re: Tell me some good news

Postby Three Valves » Thu Mar 26, 2020 6:21 pm

http://www.wees.org

Commercial and “news” free stream community radio out of OCMD.

Piano music for the dinner hour now. A return to 50s/Pops/Hit Parade even a little Classic Country now and then coming at 8pm.
Who needs four valves??

Mack Brass Artiste
MackBrunner 210L
TU25
User avatar
Three Valves
6 valves
6 valves
 
Posts: 3952
Joined: Wed Jan 07, 2015 10:44 am
Location: The Land of Pleasant Living

Re: Tell me some good news

Postby bloke » Thu Mar 26, 2020 8:13 pm

The more people they test, the higher the percentage of people with no symptoms, and the lower the death rate. Imagine how low the death rate would be if they showed the number for those with NO "underlying heath conditions".
User avatar
bloke
musician/technician/innovator
musician/technician/innovator
 
Posts: 45711
Joined: Sat May 08, 2004 6:04 pm
Location: western Tennessee

Re: Tell me some good news

Postby SWE » Thu Mar 26, 2020 8:32 pm

bloke wrote:The more people they test, the higher the percentage of people with no symptoms, and the lower the death rate. Imagine how low the death rate would be if they showed the number for those with NO "underlying heath conditions".
Bro you ok?
SWE
bugler
bugler
 
Posts: 64
Joined: Tue Oct 15, 2019 12:35 pm

Re: Tell me some good news

Postby bloke » Thu Mar 26, 2020 11:38 pm

remarkably well...as is my memory...including recalling days in a couple of statistics courses - which demonstrated how statistics can be used, as well as how they can be abused.
User avatar
bloke
musician/technician/innovator
musician/technician/innovator
 
Posts: 45711
Joined: Sat May 08, 2004 6:04 pm
Location: western Tennessee

Re: Tell me some good news

Postby bort » Fri Mar 27, 2020 12:23 am

bloke wrote:remarkably well...as is my memory...including recalling days in a couple of statistics courses - which demonstrated how statistics can be used, as well as how they can be abused.


"If you torture the data long enough, it will confess to anything"

My problem is this -- there are many cases out there confirmed, and people keep saying how many more people must be infected and not know it, infected and untested (wait... Don't they say to just stay home?!), and people who may have already had it and recovered and just never knew.

All of this is just "more people", making a larger denominator... With the numerator being the number of deaths (which is a more absolute and accurate count... Deaths are always recorded). So the same number of deaths, with a much larger denominator... Lower fatality rate. And yes Joe, if you remove the people who were already very sick with other things, then it goes down more.

I'm not saying that the deaths and sick people aren't real. It's real and it's sad. But the numbers don't add up they way they try to report them.

To say it another way:
100% of people who died from COVID-19 had COVID-19. That's about all you know for sure, and it doesn't say much. Real useful statistics about contagion and death rates needs a valid sampling method to draw inferences for the whole population. Testing already-sick people... Not really helpful for statistics. Helpful for containment and quarantine... But not these kids of statistics.

Also, the total number of infected people is modeled by a logistic curve, not exponential... right? There are a finite number of people on Earth; logistic curves have an upper bound... Hopefully far less than the total earth population!
Alexander 163 CC
User avatar
bort
6 valves
6 valves
 
Posts: 10624
Joined: Wed Sep 22, 2004 11:08 pm
Location: Minneapolis, Minnesota

Re: Tell me some good news

Postby bloke » Fri Mar 27, 2020 8:30 am

↑ this
=========

I have a longtime friend - somewhat high-profile in this industry - who has lived seven years (continuing to travel and perform concertos) with a cancer that was supposed to kill him in 1 to 2 years. He has been on life-support at least twice - for a while each time - due to past viral attacks, as he has no immune system. To my delight, my friend has continued to live - and has cheated death several times - due to modern medical science, and probably also likely due to his own incredible determination....
...but if this virus attacks him - and he succumbs to it, how do you suppose his death would be categorized?

I strongly suspect NOT “deadly incurable cancer”, but “CV - with ‘underlying health conditions’ ”.

:arrow: THIS is why I would like to see a statistic for ONLY the CV deaths with NO underlying health conditions.
User avatar
bloke
musician/technician/innovator
musician/technician/innovator
 
Posts: 45711
Joined: Sat May 08, 2004 6:04 pm
Location: western Tennessee

Re: Tell me some good news

Postby hup_d_dup » Fri Mar 27, 2020 10:08 am

bloke wrote: THIS is why I would like to see a statistic for ONLY the CV deaths with NO underlying health conditions.


You need to run it through the Bloke filter to see if it is a statistic or a "statistic."

bloke wrote:I tend to be suspicious of any "statistics" put forth by rulers' lackeys (as rulers' lackeys' jobs are to control the serf class' thoughts and actions), whether someone's else rulers' lackeys, or my own rulers' lackeys.


Hup
Facebook is your friend, your friend your friend your friend your friend your friend your friend your friend your friend your friend your friend your friend your friend your friend your friend
hup_d_dup
4 valves
4 valves
 
Posts: 784
Joined: Wed Dec 14, 2011 10:10 am
Location: Tewksbury, NJ

Re: Tell me some good news

Postby hrender » Fri Mar 27, 2020 10:24 am

hal.
1927 King 1240 - 1935 Martin Mammoth
User avatar
hrender
6 valves
6 valves
 
Posts: 2658
Joined: Fri Mar 19, 2004 5:18 pm
Location: Fruited Plains

Re: Tell me some good news

Postby Donn » Fri Mar 27, 2020 11:37 am

bort wrote:Real useful statistics about contagion and death rates needs a valid sampling method to draw inferences for the whole population.


It's a shame that this thread is located in "Tell me some good news".

Here's a statistical analysis from a group at our university, that's simply based on deaths, more or less for the reason you mention.
https://covid19.healthdata.org/projections

For Minnesota, you'll see that they expect at peak in late April, assuming current practices, you'll have about half the ICU beds needed. In my state, we'll have about 3/4 the ICU beds we need.

What I can't explain is the colored areas around the curves. I assume they represent some range of uncertainty, could hardly be anything else but I don't see that explained in any more detail. I'd be particularly interested to know why the interval is so much larger for some states than others.
User avatar
Donn
TubeNet Sponsor
TubeNet Sponsor
 
Posts: 5818
Joined: Fri Aug 19, 2005 3:58 pm
Location: Seattle, ☯

Re: Tell me some good news

Postby Rick Denney » Fri Mar 27, 2020 11:58 am

bloke wrote: :arrow: THIS is why I would like to see a statistic for ONLY the CV deaths with NO underlying health conditions.


There is a lag between the infection totals (however much those are undercounted) and outcome statistics, because the disease takes several weeks to reach an outcome and that is an eternity in an exponential growth.

But deaths have been growing exponentially just like infections (outside of China), increasing by an order of magnitude every 15 days or so. So, if we divide those who died by the total infection rate, we are missing about three or four weeks of infection growth in the calculation. That makes the percentage low (just as an undercount of the infections makes it high).

One website looks only at the cases that reached a conclusion--deaths versus recoveries. But while death has a finite determination, recoveries are being very conservatively assessed as being no trace of virus in consecutive tests. So the lag goes in the other direction, making the percentage high.

The truth is somewhere in the middle. Generally, we don't know what the rates are until after the pandemic has resolved and we can consider all of the statistics. That was true for SARS and MERS, which are also coronaviruses, and also for influenza pandemics, such as H1N1 (2009). We don't know between 50 and 100 million deaths the true outcome of the Spanish Flu pandemic of 1918-1920.

The problem with statistics is that like lots of quantitative data, they present a pastiche of precision, while the underlying accuracy is subject to large errors.

But deaths among those with no underlying conditions seems to be a little under 1%. Still about ten times higher than influenza.

Rick "who has more than a smattering of statistical training" Denney
User avatar
Rick Denney
Resident Genius
 
Posts: 6459
Joined: Mon Mar 22, 2004 2:18 am

Re: Tell me some good news

Postby bort » Fri Mar 27, 2020 12:23 pm

Donn wrote:
bort wrote:Real useful statistics about contagion and death rates needs a valid sampling method to draw inferences for the whole population.


It's a shame that this thread is located in "Tell me some good news".

Here's a statistical analysis from a group at our university, that's simply based on deaths, more or less for the reason you mention.
https://covid19.healthdata.org/projections

For Minnesota, you'll see that they expect at peak in late April, assuming current practices, you'll have about half the ICU beds needed. In my state, we'll have about 3/4 the ICU beds we need.

What I can't explain is the colored areas around the curves. I assume they represent some range of uncertainty, could hardly be anything else but I don't see that explained in any more detail. I'd be particularly interested to know why the interval is so much larger for some states than others.


Never miss a good opportunity for data visualization!

The colored areas around the curves are for standard error. Deaths per day seems pretty stable for the standard error (it's just one guess that's used for the duration). For the total deaths, the farther out you go (time-wise), the more uncertainty there is, and the bigger the shaded parts -- the errors compound over time -- you're making guesses from guesses, and things just get a lot less reliable. For example, total deaths by the end of July could be from 38,000 to 162,000. That's a huge range, and I'm not really sure it's helpful except that it's "more than right now." These totals don't just stop, they gradually level off.

Why are the intervals wider for some states than others? My guess is that there is just so much variation between states about case loads, testing (<-- again, reporting is really wacky here, and inconsistent), and then just differences in states to begin with. Alaska and California, for example, are tremendously different states (CA has over 40x the population, lots of cities, lots of population density, etc...), and the kinds of data collected in California are just a lot different than what you'd ever get in Alaska.

The cool thing about plots like this is that as you add more data, you can re-run the models and get more accurate (realistic) projections. It's basically like watching the hurricane landfall projection maps. As time goes on and as the storm gets closer, you know better and better where it's going to make landfall. At some point, it becomes clear enough (and early enough) that it's time to evacuate that area. There's still uncertainty about where it will hit, exactly... but if you waited for an exact hit, it would be too late. Here, it's the same thing. About when do we expect things to be the "most bad", and let's plan around the badness and see what we can do to make it less bad.
Alexander 163 CC
User avatar
bort
6 valves
6 valves
 
Posts: 10624
Joined: Wed Sep 22, 2004 11:08 pm
Location: Minneapolis, Minnesota

Re: Tell me some good news

Postby Donn » Fri Mar 27, 2020 1:22 pm

OK, I see that the range specifically portrays a 95% Uncertainty Interval. I think you'll see that this is not just a factor applied to the base curve. For example the curve at the bottom of the range peaks on a different day. They're likely modeling different spread rates.

As for the data differences - they're simply using deaths, which are reported fairly consistently. It looks to me like the conspicuous difference in uncertainty intervals is about whether they think your state may be able to get it under control early, based on measures that are being taken. If there's a possible but not guaranteed early peak, then the uncertainty over that is compounded over time.
User avatar
Donn
TubeNet Sponsor
TubeNet Sponsor
 
Posts: 5818
Joined: Fri Aug 19, 2005 3:58 pm
Location: Seattle, ☯

Re: Tell me some good news

Postby SWE » Fri Mar 27, 2020 1:48 pm

Donn wrote:As for the data differences - they're simply using deaths, which are reported fairly consistently.

Many believe coronavirus deaths are underreported in the U.S.
https://www.buzzfeednews.com/article/ni ... -hospitals
(Prima facie more than likely given the dearth of testing.)

There is emerging evidence that is also the case in Italy:
https://www.corriere.it/politica/20_mar ... 1130.shtml

This happy positive thread was great while it lasted, eh.
SWE
bugler
bugler
 
Posts: 64
Joined: Tue Oct 15, 2019 12:35 pm

PreviousNext

Return to TubeNet

Who is online

Users browsing this forum: ggdadams, Google Adsense [Bot], MSN [Bot], paulver and 17 guests