'Historically, it's possible that both the US and Russia are heading for historic disaster. Couldn't happen to two nicer bullies. '

Oh Joe and I thought you had not read the end of the book. You certainly want to hope for Australia's sake that the US does not go under. Then again with so many fools embracing Marxism maybe we deserve what we get.

Dear mhaze,

This is what you claimed; “Load that CSV into a spreadsheet and do some sorting of the raw numbers and you'll find that so far this year they've predicted 7770 deaths in excess of their so-called threshold ie expected number based on previous periods. So 7770 people over the year in excess of what would happen in a normal year.”

You are a scream. To get such a low figure you would have done something like count deaths which didn't make the threshold as surplus to be then taken off those above the threshold. Am I right?

Utterly dishonest but not unexpected.

The threshold used is the “Upper Bound Threshold for Excess Deaths”. This is the threshold which warns of an outbreak. It is NOT the 'expected number based on previous periods' at all.

Further they quite clearly state; “Negative values, where the observed count fell below the threshold, were set to zero.”

But you ignored that didn't you. You decided to use figures which were the result of a less than potent flu season to subtract from the Covid toll and state the case for Covid19 not being that extraordinary.

Perhaps sticking to the pretty pictures might be advised for your good self. It seems you get beguiled and bedazzled when hard numbers are placed in front of you.

Silly, silly boy.

Runner,

No, you misunderstand me: yes, Russia and the US are heading for a disaster by opening up far, far too early, but the US will come out of better than Russia will, and I think it will be able to take its place again in the forefront of economies in barely ten years.

It will be a different world by 2030.

Joe

You see, this is why its so hard to have these type of conversations with you. You utterly misunderstand the maths, convince yourself otherwise and then make assertions based on the misunderstanding.

Yes, in this scenario the CDC does ignore weeks where the numbers of deaths don't exceed the prediction. But that's because they are answering different questions. What they are looking for is what they call "the burden of mortality" ie weeks where the medical system may be under stress due to higher than expected deaths.

But that's not the question I was seeking to answer. I was looking for the overall numbers of excess deaths in the period by comparing the number of deaths and the predicted number. Oh and despite yet another misunderstanding from you, the predicted number is based on past periods..."Counts of deaths in more recent weeks are compared with historical trends to determine whether the number of deaths is significantly higher than expected." (from the technical notes). Want me to explain 'historical trends' to you?

So I totalled the number of deaths for the year to date and compared that to the number of deaths predicted YTD.

To confirm the calculation I did it for previous years as well.

It probably won't make any difference to your understanding but just for the record the number of deaths to the end of April was:

2020 - 1036723, predicted 1028953
2019 - 975537, predicted 1037458
2018 - 1002269, predicted 1024196
2017 - 976086, predicted 1002452

When you adjust for population changes these figures become remarkably similar.

So as someone brilliantly opined two months ago, basically a normal flu season.

LM,

Perhaps this will help you to understand why your post "So 80,000 Americans haven't died from this virus ?" is utter bunkum.

___________________________________________________________

I saw a post from a statistician yesterday that said that the population adjusted US deaths from the Hong Kong Flu was 160000. But no one talked about closing the entire economy.

Dear mhaze,

Well that gave me a very hearty chuckle so thank you for being such a goose.

The "Threshold" figure is set at the 95% Confidence Interval for crying out loud!!

This is not the average figure at all mate. It is the figure which will be exceeded only 5% of the time. It is set purposefully high so if an observed figure exceeds it it will be a red flag.

What you have done is taken the actual figures and subtracted them from this purposefully high figure CI figures and un-bloody-surprisingly came up with far a lesser figure of 7770. Right up until the middle of March you kept totaling illusory negative figures didn't you.

You have even included the week-ending the 25/4/2020 which had not been fully updated so it took another 15,259 off the total.

Mate, give it up. That is why graphical representations are there, for people like you to check your assumptions. The fact that you were so dismissive of them now means you are in the dog house and looking like an idiot, especially after you had been so cockahoop. I knew as soon as you posted that you had to be talking out your backside.

COVID-19 is becoming an excuse !