Welcome to the COVID-19 Weekly Challenge.
Your task this week, should you choose to accept it is to estimate the USA total case projections over the next 7 days.
The winner will be deemed the person with the closest estimate based on the Worldometer stats that will be closed off next Sunday (eg: around noon when the numbers tick over to a new day, then the “yesterday” numbers will be deemed the final amount). Closest to the pin, doesn’t matter if you go over or under.
Entries will close at midday on Tuesday, although you can always add your thoughts and estimates if you missed the cut-off.
My entry below.
Today - 123578 (this is the actual)
Monday - 144000
Tuesday - 166000
Wednesday - 189000
Thursday - 213000
Friday - 238000
Saturday - 264000
Sunday - 291000
… don’t look at me like that… I’m bored.
I’ll play. Are we going with the close of day figure for each date based on GMT+0, which that worldometers site is doing?
Today (which is end of 28/03/2020 based on GMT+0) - 123578
Monday - 145000
Tuesday - 168000
Wednesday - 193000
Thursday - 220000
Friday - 250000
Saturday - 284000
Sunday - 320000
Am I meant to model this with some sort of time series ARIMA type model, or am I meant to guess?
I think I’ll guess *shitgibbon-loads"
why did they only go up by 40k on the 6th day? Does the big stupid orange shitgibbon step in and do something? I’m sure he’ll take credit for that day being lower
Shitgibbons are unpredictable.
My answer: they are ■■■■■■.
Here’s a challenge for you all? Is it possible for people with kids to stay the ■■■■ at home? How about the dad stays home with the kids and just have one person do the shopping?? Or the boyfriend and girlfriend just stay inside rather than browse the streets.
While I appreciate the sentiment, the answer really needs to be quantitative rather than qualitative.
Geez mate, next thing you’ll be asking for world peace.
I got the right thread this time.
I will guess 705,000 but could be much lower eg 550k if their report does not get in on by your deadline. And they could have many more than that if they start testing in the numbers that they should be.
I just figured there’s be a day of relative improvement somewhere. And then it would go bad again.
Odds are the shitgibbon will stop testing at some point and then claim no new cases and pretend the thousands of deaths were natural causes or something.
Whole forum needs a name change. Covidblitz?
Gee, you lot are confident -only going point estimates, and no uncertainty intervals!
I think the number is going to be limited by testing capacity rather than patients.
457,000 identified cases.
I actually feel really bad about participating in this.
It’s like driving past a car accident. It’s very hard to look away
999,999, which is a bigly number (possibly the bigliest).