Posted 5 June 2020
The ability to convey often complex statistical ideas in a way that non-statisticians can follow is I believe one of the true skills of a good statistician. It does not matter how clever you are with analyses, if you cannot convey the meaning in a way that other people can follow, then the piece of work is of very limited value.
A couple of days ago I posted a piece about the chances of you dying from COVID19 in terms of ‘extra days lived’. A few friends have since sent me private messages saying they did not follow my point. By my own criteria therefore, with that piece of analysis, I failed!
Let me try another explanation, this time with an analogy.
I want you to imagine a deck of playing cards. In this deck there is one Joker card, all the other cards are normal playing cards. In true style of a magic trick, I then ask you to “pick a card, any card”.
However, this magician’s trick is rather sinister as the rules are if you pick the Joker, then I am afraid your time is up, it is time to meet thy maker. To make the rules worse, you have to go through this card section routine once every morning. [To be strictly accurate with my analogy, I have to point out that you put the card you have selected back in the pack before tomorrow’s draw – assuming of course it was not the Joker!]
The key question: How big is the deck of cards? If there are only the standard 52 cards plus one Joker, then things are not looking too good for you as it won’t take many morning selections before the Joker is likely to be picked.
Fortunately for us all, the pack is rather large, but how large depends on your age.
If you are under 20 years old there are 1,190,000 cards in your pack along with that one Joker. Every day, as an under 20 year old, you therefore have a one in 1,190,000 chance of picking the Joker and therefore dying.
For those over 80, the size of the deck is around 5,000 cards. Every time you select from the pack you have a one in 5,000 chance of picking the Joker and so that is your chance of dying each day in a pre-COVID world. [All based on 2018 data].
What difference has COVID19 made? 2020 has obviously been a COVID year – by how much has it increased our chances of dying?
Well, for under 20 year olds, think of it as being forced to take one extra selection from the pack of cards during the whole of 2020. So rather than sampling the pack 366 times you will be asked to sample it 367 times between 1st January and 31 December 2020. That one extra ‘pick a card’ moment is the entirety of your risk of dying with COVID19 mentioned on your death registration.
For the over 80 year olds your risk is of course much greater. Rather than selecting from the pack 366 times, I need you to select from the pack 399 times during 2020. Not only is your pack much smaller but you need to sample from it nearly 3 extra times every month of 2020. Those 33 extra ‘pick a card’ moments represent your extra risk of dying of COVID19 in 2020.
Hence my previous analysis – for an under 20 year old, the risk of dying from COVID19 has been like living one extra day in 2020; for an over 80 year old, it has been like living an extra 33 days.
For sake of completeness, the deck sizes and number of ‘extra picks’ for all age groups are shown in the second figure attached to this post.
How does that sound?
