Z Charts and COVID-19

Introduction

Z Charts used to be quite a popular form of communication years ago but now it can be difficult to find reference to them at all. On my old web site, I had a few examples of Z Charts and I just thought the other day that they might be a useful addition to the covid-19 discussion.

First of all, let’s see what Z Charts are, one for world covid-19 cases and one for covid-19 deaths and then see how to construct and interpret them.

Z Chart Components

There are three separate but related curves on a Z Chart

• The daily, in this case, numbers (cases and deaths)
• The cumulative numbers
• The Moving Total numbers

A Z Chart always looks like the letter Z when we draw it properly. So how do we construct them and what do they tell us?

Why am I showing an 18 Day Moving Total? Because I am plotting March 2020 data and I am adding the data as it comes in.

Z Chart Data

Here is the part of the table I used to create the covid-19 cases Z Chart:

The Daily column contains the daily new cases being reported

• The cumulative simply adds the new cases from day one to day two … then to day three …
• The 18 Day Moving Total starts on 27th February by adding together all of the daily results for the range E25:E42 and as we fill that down, it becomes E26:E43 and so on.
• There are three checks built into this chart: the two sets of cells that I have coloured on 27/2/2020 and on 15/3/2020: they show that they must be equal to each other on their respective days. The third check in the Difference column that confirms the check on 15/3/2020 that the cumulative and the moving totals are the same as each other. If these checks fail, the Z Chart will be wrong!

How do we Interpret a Z Chart?

We see that the number of daily new cases is the bottom line and it slopes upwards from left to right: it is indicating that the trend of new cases is rising: more and more cases per day. If we plot just the daily cases, this is what that curve looks like:

The changes are quite dramatic, aren’t they?

The cumulative results also show an increasing trend from left to right: from the past to the present. Also looking at that curve in isolation:

This curve confirms the sharply increasing rate of change of cases.

Finally, what about the 18 day moving total, what does that tell us? Let’s see:

The story here is that we see a levelling off and then a fall in the moving total cases but then, in tune with the other two curves, we see a steep rise. That tells us that something sudden seems to have happened around 4th March 2020 when the curve inverted.

You can download the latest data from various sources, including this really user friendly site: https://www.worldometers.info/coronavirus/

Duncan Williamson

16th March 2020