Comparing Italy & UK in the 6 pack colour charts at the end & focusing on deaths Italy, less so than China, has peaked but UK not so. This sort of contradicts the daily charts we get (herein Scotland) issued at the UK 5pm Downing Street Press briefings. The “Daily cv-19 deaths in all settings” shows a clear broad peak emphasised by a 7 day rolling average curve.
The total in the past included only Hospital deaths, it now has added Care Home ones (currently larger on a daily basis). The Guardian has doubted that all deaths are included. But on the basis of his data we are contemplating leaving lockdown (we will find out Sunday ….)
Vox(dot) com (an online magazine) has a COVID section—one refers to an article by someone at JHU who compares the Imperial college UK study with one by a group at U Wash, Seattle (apparently funded by Gates foundation). Both of these studies were later shown to be likely incorrect.
Peter Turchin of U Conn and Y Bar-Yam of NECSI also have articles/papers/blogs on this. (Turchin especially compares the cases of Sweden and Denmark which had different policies on ‘quarantining’, though they also have different demographics, so its hard to compare).
I just wonder how all these models compare, and how anyone can process all this information, especially if they are half scientifically illiterate.
There are so many models that there is a website that is tracking the number of trackers https://covidtrackertracker.com/. This website is tracking the predictions of a select few https://github.com/reichlab/covid19-forecast-hub. The CDC also compares models to each other https://www.cdc.gov/coronavirus/2019-ncov/covid-data/forecasting-us.html. My take is that almost all models are overfit in the sense that the data are not sufficient to constrain them so they all have weak predictive power. The UWash/IHME paper is a curve fit and has no predictive power. Curve fits are fine if you have the correct curve but they are choosing their curve based on how well it fits the front end of the pandemic to predict the back end and dynamically there is no reason the front end determines the back end from shape alone. The world is plateauing at 100000 new cases per day and I don’t think anyone has an explanation for that.
Scientific Clearing House 
Thank you very much!
LikeLike
Comparing Italy & UK in the 6 pack colour charts at the end & focusing on deaths Italy, less so than China, has peaked but UK not so. This sort of contradicts the daily charts we get (herein Scotland) issued at the UK 5pm Downing Street Press briefings. The “Daily cv-19 deaths in all settings” shows a clear broad peak emphasised by a 7 day rolling average curve.
The total in the past included only Hospital deaths, it now has added Care Home ones (currently larger on a daily basis). The Guardian has doubted that all deaths are included. But on the basis of his data we are contemplating leaving lockdown (we will find out Sunday ….)
PS good work
LikeLike
Vox(dot) com (an online magazine) has a COVID section—one refers to an article by someone at JHU who compares the Imperial college UK study with one by a group at U Wash, Seattle (apparently funded by Gates foundation). Both of these studies were later shown to be likely incorrect.
Peter Turchin of U Conn and Y Bar-Yam of NECSI also have articles/papers/blogs on this. (Turchin especially compares the cases of Sweden and Denmark which had different policies on ‘quarantining’, though they also have different demographics, so its hard to compare).
I just wonder how all these models compare, and how anyone can process all this information, especially if they are half scientifically illiterate.
LikeLike
There are so many models that there is a website that is tracking the number of trackers https://covidtrackertracker.com/. This website is tracking the predictions of a select few https://github.com/reichlab/covid19-forecast-hub. The CDC also compares models to each other https://www.cdc.gov/coronavirus/2019-ncov/covid-data/forecasting-us.html. My take is that almost all models are overfit in the sense that the data are not sufficient to constrain them so they all have weak predictive power. The UWash/IHME paper is a curve fit and has no predictive power. Curve fits are fine if you have the correct curve but they are choosing their curve based on how well it fits the front end of the pandemic to predict the back end and dynamically there is no reason the front end determines the back end from shape alone. The world is plateauing at 100000 new cases per day and I don’t think anyone has an explanation for that.
LikeLike