- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Coaching Changes
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
re: UW model updates again overnight
Posted on 4/8/20 at 8:43 am to The Maj
Posted on 4/8/20 at 8:43 am to The Maj
quote:
Sounds like this projection evolves about as much as the cone does tracking a hurricane.
shite worse than that
Yes, it was like a prediction for a CAT 5 hitting NYC, instead it landed south of the Yucatan Peninsula.
1
Posted on 4/8/20 at 8:45 am to The Maj
quote:
So, make up worst case data and then do little to no comparison against other models in order to vet the information?
It also had a "best case scenario" built-in on March 26th which had the low end of U.S. deaths somewhere around 25,000-30,000 total.
But because people have no idea how to read models, everyone ignored it and went straight with the worst case scenario.
Posted on 4/8/20 at 8:47 am to Norbert
quote:
I don't remember hearing that.
Selective memory tends to happen when your side is proven wrong. This whole thing was a charade.
This post was edited on 4/8/20 at 8:48 am
Posted on 4/8/20 at 8:47 am to Antonio Moss
quote:
Antonio Moss
quote:
because people have no idea how to read models
Posted on 4/8/20 at 8:47 am to 91TIGER
quote:
Yes, it was like a prediction for a CAT 5 hitting NYC, instead it landed south of the Yucatan Peninsula.
I'd say it's more like we issued mandatory evacuations for the entire Gulf Coast and Atlantic Coasts because of a tropical wave off Africa. When, in all reality we needed to simply properly remind those in high risk areas to have their preparations in place.
Posted on 4/8/20 at 8:47 am to Dawgfanman
quote:These are flat out lies.
specifically being told that it would NOT lead to a lower overall death toll or infection rate, but that it would just spread it out over time.
Posted on 4/8/20 at 8:48 am to Antonio Moss
quote:
It also had a "best case scenario" built-in on March 26th which had the low end of U.S. deaths somewhere around 25,000-30,000 total.
But because people have no idea how to read models, everyone ignored it and went straight with the worst case scenario.
This is your take now?
Posted on 4/8/20 at 8:48 am to Shaun176
quote:
If they have enough time to reflect and interview, it must be getting better.
I saw a TV story this week about a cured COVID-19 woman being wheeled out of the hospital to her husband, who hadn't seen her in about 15 days.
Emotional, of course..........But the hospital exit was lined by 20-or-more white coats. I thought, they have time for this when they claim they don't have time to eat or use the bathroom? I thought it was supposed to be Armageddon at General Hospital.
Nevertheless, here's to a cure, your good health and a return to normalcy.
Posted on 4/8/20 at 8:49 am to Golfer
quote:
I'd say it's more like we issued mandatory evacuations for the entire Gulf Coast and Atlantic Coasts because of a tropical wave off Africa. When, in all reality we needed to simply properly remind those in high risk areas to have their preparations in place.
Touche'
Posted on 4/8/20 at 8:49 am to Antonio Moss
quote:
But because people have no idea how to read models, everyone ignored it and went straight with the worst case scenario.
I am not arguing that they didn't and that most have no clue how to read the models BUT this model is built on faulty assumptions (I will give them credit for attempting to correct with real data) and best I can tell was not vetted against other models and assumptions...
Also, I haven't really seen them trying to get out in front of it and "say, hey why are you continuing to use the worst case scenario"... They have as much responsibility as in how their data is utilized as they do in the development...
Posted on 4/8/20 at 8:50 am to Golfer
Antonio posted this the other day, and it seems extremely apparent every time he's arguing in a modeling thread:
quote:
First, I'm not an expert is statistics. I don't have much of a background in it
Posted on 4/8/20 at 8:51 am to The Maj
quote:
How do they base anything on this? The data out of China is suspect at best, so how do you base a model on this to start with?
Here is the link to the supporting paper for the model -
LINK
quote:
The value of
the covariate multipliers in each type of model was assumed to closely follow the fit obtained
from data from Wuhan, which is the time series to reach a stable state in the training dataset. To
be specific, the generalizable information from Wuhan was the impact that social distancing had
on maximum death rate and time to reach the inflection point. For each type of model, we both
considered ‘short-range’ and ‘long-range’ variants, to explain existing data and forecast longterm trends, respectively. In the former case, covariate multipliers could deviate from those fit to
Wuhan, while in the latter, the data from Wuhan had a larger impact on the final covariate
multiplier.
quote:
The consequent main limitation of our study is that observed epidemic curves for COVID-19
deaths define the likely trajectory for US states. In this study, we do include a covariate meant to
capture the timing of social distancing measures to take into account that Wuhan implemented 4
out of 4 social distancing measures within 6 days of reaching a threshold death rate of 0.31 per
million.
Posted on 4/8/20 at 8:51 am to Golfer
quote:
It also had a "best case scenario" built-in on March 26th which had the low end of U.S. deaths somewhere around 25,000-30,000 total.
But because people have no idea how to read models, everyone ignored it and went straight with the worst case scenario.
This is your take now?
No, just pointing how people don't know what they are looking at. There have been 100+ posts of "golly, them science folks, they said we was gonna have 2 million deads and now we's only gonna have 60,000"
There have been zero stating that those same science folks "predicted" 30,000 deaths and now are saying we are going to have 60,000.
Posted on 4/8/20 at 8:52 am to Norbert
quote:i think you have that exactly the opposite. hospitalization has not been an issue but fatality has absolutely been a huge issue for the target demographic - seniors
The biggest problem with this disease isn't its fatality rate, it's the contagiousness combined with the hospitalization rate
quote:this does not seem true. moreover, even nyc, by far the worst place in the country, never got overwhelmed
deaths will certainly fall if you keep hospitals from becoming overwhelmed
Posted on 4/8/20 at 8:52 am to msudawg1200
quote:
I don't know where you saw that, but this is right off the model
COVID-19 projections assuming full social distancing through May 2020
It's in the FAQ section.
quote:
SOCIAL DISTANCING
Does your model show the effect of social distancing and other measures?
The model includes the effects of social distancing measures implemented at the “first administrative level” (in the US this generally means the state level) and assumes continued social distancing through the end of the modeled period (August 4, 2020).
Posted on 4/8/20 at 8:53 am to The Maj
quote:
BUT this model is built on faulty assumptions
It was built on the available data at the time and the current trends (in LA and in the US) still fall within the original data set. The model narrowed as data became more prevalent.
Posted on 4/8/20 at 8:53 am to rds dc
quote:
from data from Wuhan,
Which was suspect from the start and still is...
quote:
the data from Wuhan had a larger impact on the final covariate multiplier.
Which means they relied very heavily on suspect data from the start, which is one of the reasons they are having to adjust their models with frequency...
quote:
that Wuhan implemented 4 out of 4 social distancing measures within 6 days of reaching a threshold death rate of 0.31 per million.
Again, building their curve on suspect data at best...
Posted on 4/8/20 at 8:54 am to The Maj
so for this model to be accurate, we would have to extend sheltering/distancing until May 30?
Posted on 4/8/20 at 8:56 am to Antonio Moss
quote:
The model narrowed as data became more prevalent.
The model was absolute shite in the beginning, so when you feed shite in you get shite out...
Remember this about numbers and models "I can make anything look as good or as bad as I would like to make them look"...
Page 4 of 7
Popular
Back to top
Follow TigerDroppings for LSU Football News