- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- 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
The current data modeling for COVID-19 is akin to weathermen predicting rain
Posted on 4/2/20 at 8:24 am
Posted on 4/2/20 at 8:24 am
The advocate should also be ashamed of themselves
Here is the link on the landing page
From the story
LINK
These so called data scientist aren't going to be any more accurate that Al Roker trying to predict if the farmers almanac summer forecast.
Here is the link on the landing page
quote:
Models paint grim picture of coronavirus' impacts on Louisiana
From the story
quote:
One such model, designed to give early warning to public leaders and health officials about how quickly they will need to ramp up the capacity in their hospitals, predicts between 360 and 3,100 residents of Louisiana could die from the coronavirus by early May. And the influx of patients in need of lifesaving care is not likely to peak in the state until next week.
quote:
In Louisiana, the model predicts the need for medical resources will peak in about nine days, when up to 1,730 ventilators and 8,000 hospital beds will be needed for coronavirus patients.
quote:
Gov. John Bel Edwards has been pleading with the federal government for 5,000 more ventilators to meet the demands predicted by the state’s own, in-house modeling. Those estimates now predict the parishes around New Orleans, which has been the epicenter of the crisis in the state, will run out of ventilators on Monday, Edwards said Wednesday afternoon.
LINK
These so called data scientist aren't going to be any more accurate that Al Roker trying to predict if the farmers almanac summer forecast.
Posted on 4/2/20 at 8:27 am to lsufan1971
They have to assume worst case scenario to be prepared
Not sure what the huge fuss is about
Not sure what the huge fuss is about
Posted on 4/2/20 at 8:30 am to lsufan1971
#1 Louisiana needs to have more vents in general.
#2 I haven’t seen a model that requires 5000 total vents in Louisiana.
#2 I haven’t seen a model that requires 5000 total vents in Louisiana.
Posted on 4/2/20 at 8:30 am to lsufan1971
I have been saying this for weeks.
I always use that prick Jim Cantore as an example. Cheerleaders of destruction.
I always use that prick Jim Cantore as an example. Cheerleaders of destruction.
Posted on 4/2/20 at 8:32 am to Powerman
quote:
They have to assume worst case scenario to be prepared
I have no issues with being prepared. But don't call it data modeling if your skew for deaths is 88%.
This post was edited on 4/2/20 at 8:33 am
Posted on 4/2/20 at 8:35 am to lsufan1971
The advocate reads the poli board. They just changed the wording on the link on the home page from
To
quote:
Models paint grim picture of coronavirus' impacts on Louisiana
To
quote:
How many people could die from coronavirus in Louisiana by May? Models say hundreds to thousands
Posted on 4/2/20 at 8:37 am to Powerman
quote:
They have to assume worst case scenario to be prepared
Not sure what the huge fuss is about
Because the government shut downs are being prolonged due to these "worst case scenarios" that likely will never happen, and people are losing their jobs because of it.
Posted on 4/2/20 at 8:38 am to lsufan1971
quote:
between 360 and 3,100 residents of Louisiana could die from the coronavirus by early May. And the influ
Well that’s a huge gap. And on the high end it only .068% of the La population.
Posted on 4/2/20 at 8:41 am to BamaCoaster
quote:
I always use that prick Jim Cantore
My brother livers within spitting distance of the Gulf of Mexico. The day before Katrina came to town Jim Cantore was reporting directly in front of my brother's home. My brother walks down from his porch and asks Cantore what are the chances he can ride out the storm in his house.
Cantore replies that my brother would have less than 24 hours before he meets death if he stays in place
My brother considered Cantore's reputation for fearmongering while looking at the vastness of the gulf and decided to err on the side of caution and left.
He never saw his house again. Even people paid to predict the worst can get it right at times.
Posted on 4/2/20 at 8:46 am to lsufan1971
quote:
I have no issues with being prepared. But don't call it data modeling if your skew for deaths is 88%
And strangely...3100 deaths isn't terribly "high" considering Louisiana is a hot spot for the virus. Louisiana loses 800 a year to the flu, and has higher average death rate than the US per capita at that as well.
With the spread being marshalled, it will probably be half 800 when it's all said and done, but if they keep blaming falling off ladders to CV, we might see 800.
And I agree, deaths are grim inherently, so to say that is either redundant or hyperbole.
Posted on 4/2/20 at 9:12 am to lsufan1971
quote:
These so called data scientist aren't going to be any more accurate that Al Roker trying to predict if the farmers almanac summer forecast.
Which assumptions in the model do you believe are incorrect?
Posted on 4/2/20 at 9:18 am to lsufan1971
What exactly are you complaining about and why did you put the part about JBE requesting 5,000 vents in bold? Was that supposed to be some sort of gotcha?
Posted on 4/2/20 at 9:32 am to longwayfromLA
quote:
Which assumptions in the model do you believe are incorrect?
Predicting between 360-3100 deaths. That is like predicting LSU football can go either 3-9 or 15-0 in the same breath. My point is they aren't predicting they are guessing. Don't call it data modeling if you are going to guess.
Posted on 4/2/20 at 9:51 am to lsufan1971
quote:
Predicting between 360-3100 deaths. That is like predicting LSU football can go either 3-9 or 15-0 in the same breath. My point is they aren't predicting they are guessing. Don't call it data modeling if you are going to guess.
Right. 360 - 31000 are outcomes of the model. I asked about model inputs - specifically the assumptions you think are wrong. It is not in anyway abnormal for models to have large ranges of outcomes depending on how impactful and variable the key assumptions are. The "guesswork" as you put it, is in the estimate for:
- doubling rate of the disease
- impact of social distancing on doubling rate
- mortality rate
- current penetration
etc.
Surely you can see that if one person has an assumption of 14 days for doubling rate and another has 3 days, the output of their versions of the model will be a large variance. So I ask again, which assumptions do you believe are overstated.
Posted on 4/2/20 at 9:54 am to longwayfromLA
quote:
I asked about model inputs - specifically the assumptions you think are wrong.
I don't think that's the point he's trying to make. The point is we are making policy decisions that have impacted tens of millions of people based on models that are producing quite a bit of uncertainty.
Posted on 4/2/20 at 9:59 am to lsufan1971
This datahas been pretty accurate with influenza and it's also being used to now gather CV-19 data and it does look like Louisiana is gonna get hit pretty hard.Florida looks like the next hotspot as well.
No idea about deaths.
No idea about deaths.
Posted on 4/2/20 at 10:02 am to lsufan1971
Well, that’s an interesting opinion on your part. We’ll keep It in mind for a few seconds.
Posted on 4/2/20 at 10:03 am to RollTide1987
quote:
I don't think that's the point he's trying to make. The point is we are making policy decisions that have impacted tens of millions of people based on models that are producing quite a bit of uncertainty.
That's a somewhat fair point. But I can't help but note that we have clearly, consistently underreacted to this thing. And to be frank, it's been some time since I've seen outcomes in any of the models that I would describe as optimistic, even taking in the most favorable assumptions.
The math is PUNISHING. This thing doubles, absent mitigation, at best every 9 days, probably 2-3 times that fast in reality. That's observable with real data now. There were 6 confirmed case in Louisiana 3 weeks ago today. There are over 6,000 now. THAT is what people are making policy decisions off of. THAT is why the models look so scary.
Posted on 4/2/20 at 10:04 am to lsufan1971
It’s like listening to Greta on climate
Posted on 4/2/20 at 10:04 am to lsufan1971
quote:
The current data modeling for COVID-19 is akin to weathermen predicting rain
So it's going to be pretty accurate?
Back to top
Follow TigerDroppings for LSU Football News