Coronavirus in the United States - news and thoughts

Monthly totals aren't great at trends if the variation is happening within a month, which it is. There is weekly variation that can be handled with 7-day moving averages. There is no reason to aggregate the data out to a monthly level except to obscure trends that happen within the month.

 

I picked the days before the surge and the peak of the surge (which is appropriate when the topic of discussion is the surge, which is how you started the discussion). You utilized a monthly average that contained both the trough and the beginning of the peak. You were looking at a weekly level when you wanted to (talking about Mid-July). When the weekly data didn't support you, you switched to monthly data to hide the trend that happened mostly in the month of July.

 

well i guess you are right, deaths do double or triple from sat/sun when very little actual reporting is done to mon/tues when the numbers are caught up, how could i be so blind not to see your imperial logic and wisdom.

 

There are lies, damn lies, and statistics. Or as Homer Simpson once put it, "Oh, people can come up with statistics to prove anything, Kent. Forty percent of all people know that."

Rolling averages when it comes to COVID-19 is the best metric because not all days get reported equally. A rolling weekly average ensures that you are not dropping a Sunday-Monday and adding a Tuesday-Wednesday, which would skew the numbers, since weekend death totals tend to be lower than weekday. Weekly numbers also provide a better snapshot of what is happened or what happened versus monthly, which encompasses a longer period of time. Monthly numbers are better for overall trends.

The bottom line is deaths went up as the number of cases went up. We're arguing which data sets to use to prove the point, but they are all showing the same trend. It's just the variance where we disagree. And how any can argue that deaths increase as infections increase is a good thing is beyond me. Just goes to continue to show the best way to decrease COVID-19 related deaths is do what we can to avoid the spread of the virus.

 

He's going to die on that hill.

 

Stop it.

Early July to early August 7 day average:

 

I was utilizing 7-day moving averages. 7-day moving averages account for weekly variance. That is why people use them.

 

so again, what days did you use? from worldometer it appears the surge started on july 10 to july 31. you used your base 3 day stats from july 4-6, a weekend plus a holiday, so how did you expect numbers on deaths to be high, no one worked, so of course any other 3 day period would look good

 

Oh he knows it’s being disingenuous and hoped no one would notice. He got called out on it and keeps making up BS.

 

I have told you multiple times which days I used. A 7-day moving average for July 4 would be the average daily deaths that occurred from June 28-July 4. The 7-day moving average for July 5 is June 29-July 5. The 7-day moving average fr July 6 is Jun 30-July 6. The 7-day moving average smooths weekly variation into comparable metrics regardless of the day of the week.

 

It's like talking to TheJosh

 

go back and look at your post 17334, you stated that the average was 521-524 from july 4-6, 3, count them 3 days, then you compare to july31-august 5, 6, not even 7 days.

 

The 7-day moving average on those days is the metric that I was using. I thought that was obvious in that post when I said "If you look at the 7 day moving average, it was 521-524 from July 4-6," but I guess that was a little too vague. So let me repeat, I was utilizing the 7-day moving average for those three days as the trough.

 

me thinks you are trying too hard to impress people, if you say 7 days, reference 7 days, not 3 days as a whatever the hell a trough is.

 

Maybe they're just thejoshing you? I mean. They have to be, right?

 

He did. You can't read.

 

no, you must be blind. go to post 17334 and tell me it does not say july4-6, which, even i guess in your universe, is 3 days, not 7.

 

plus, in my universe, july 31-august 5 is still 6 days, not 7.

 

Well, to make his numbers fit his narrative he had to subtract a day I guess...

 

You are still failing to read, thejosh. I'm loling. Thanks.