Deeper Dive Data (DDD): May 2020

Thursday, May 28, 2020

Estimates of Economic and Other Activity Using Google Mobility Data

stock here.   At ZH I saw an Assin firm making broad speculations about a "second wave" based upon "Mobility Data".

Click on chart for a larger view

OK nice idea, but they should address more important consideration of herd immunity.

AND they are missing the boat completely in terms of a "second wave".   In the Spanish flu, cities that started mitigation quickly, and ended mitigation quickly did the best in both lower deaths, and better economic outcomes EVEN IF there was a second wave.

So many flawed conclusion are drawn by those either with and agenda or lack of a critical eye.   Such as St. Louis is routinely lambasted for their quick end to mitigation (flavors of modern day 'lockdown' meme and narrative)....but their overall death numbers were VERY GOOD in relation to the rest of the USA.

Additionally, A test check against other conventional measure of economic activity should be done to see how well the Mobility Model works.
---------------------------------------------------------------------------

https://www.google.com/covid19/mobility/

So here is Wisconsin. The chart are compared to baseline. They don't say baseline month to month, or month to average over the year, I think the parks data is overblown.

Tuesday, May 26, 2020

Virginia Creates Fear, By Presenting Data In a False Manner, Intentional or Not, It Is a Crime

I sure suspected this when I saw this headline at Zerohedge yesterday.
Click on images for a much better view.
ZH published this early in the afternoon of 5-25-2020, but it was for data for the day 5-25-2020, now that is interesting how the data was compiled for the whole day before the day even ended, and no less, on memorial day?

They republished it today, the 26th, with the 26th as the issue date.

Read the title below, and see how it doesn't match the title of the article itself.

https://www.zerohedge.com/geopolitical/media-exposes-100s-uncounted-covid-19-deaths-japan-lifts-state-emergency-us-death-toll

I thought Virginia was fudging the data to make it appear that a huge spike occurred.

Whomever is participating in this and not reporting it is complicit at best.

Data from this site
https://covidtracking.com/api

I'll send the spreadsheet to whomever wants it.

Friday, May 22, 2020

Charts and Models Based On Data of 25 States Show This "Thing" is Pretty Much Over, USA 60% Down From Peak

heh, if you like this stuff, sign up as a follower, I might even think someone is paying attention or
appreciates the work.

click for a larger clearer picture

----------------------------------------------------------------------------

I charted a bunch of states yesterday, and have requests for 11 more, LOL 11 is my number.

As always blogger sucks, but click for a large clear image, easy to read the dates.

But look at Massachusetts, and especially the weird 7 day moving average. 10 States participating in the Apr 29 Spikefest. It was planned at least a week in advance and coordinated. All by phone call? Must almost for sure be a group phone call, unless one party is creating the plan and then just distributing "orders". Maybe they did it on Zoom.

This looks treasonous to me, what could their excuse possible be? "Oh, we did it in an over abundance of caution, we figured it was better for people to be scared than dead."

New Hampshire, Very few deaths

New Hampshire -- Toughest model I have ever done. Data all over the board. I had to model based on the 7 day MA In Red, Model in Orange. Look at the tabular data, they have to under report, and borrow deaths from the future to show the scary spikes.

Florida, Gov Santos opened up very early.

I have been catching up on my real life, and had to put the charts aside for awhile.

Taking requests....what States do you want to see with the "presented deaths" and the model.

CLICK ON ALL CHARTS FOR A MUCH BETTER VIEW!

For about 10 states I have the models that were done around 4-26-20, almost a month ago. They represent how a real virus would present data based upon a "declining exponent model".

This can change, I'll present my entire Pandemic Model theory someday, but let's sum it up quick. There is a Measure called R-Naught, R sub 0, or looks like this, but a real pain to type

R₀

It says how well the virus can jump from one human to another, how many other humans are likely to get the virus from one infected human.

For each virus there is an inherent R-naught, it the nature of the beast when presented to a totally unaware, never seen this virus before, population. Now my science and modeling is based upon what I call the R-naught effective. How well does the virus jump, how many other humans will one human infect given other variables, and in the models using data, it really doesn't matter what the other factors are. The numbers speak for themselves and create the R-naught-effective

R_0E

_{Any number of factors can be modeled, the most important being, the number of people in the community who are now immune, and the Social Distancing which is a very deceptive term. It doesn't mean the number of feet away from the other people, which has become so popular at the magical 6 feet. A much better measure is the amount of "spray" from one sick person to another, and this can come just from talking, especially talking forcefully. Sneezing is worst, and coughing is second.}

_{So Wisconsin here....Model is from 6 weeks ago, and you can see ---}

Once the "Gang of 10" Governors Crossed the RUBICON of the Apr 29 Spike Fest, they lost all fear of complete data fudging