COVID-19, The Corona Virus Pandemic through the eyes of a Six Sigma Blackbelt – UPDATE#2 (Predictions)

Related Topics: Covid-19 Update#1

Data Analysis, Predictions, Visuals and Key Takeaways for us:

Covid-19 Predictive Analysis

Assuming social distancing measures, COVID-19 deaths in the United States of America by the first week of August 2020 is projected to be about 60,000 (ranging between 31,000 and 126,000 deaths), Data Source: IHME. State by State Projection for the United States is available here. It would peak by April 12th, 2020, with over 2200 deaths per day.

Covid-19 Deaths Per Day Projections for USA. Data Source: IHME

WHO has developed a Health Alert system (in collaboration with Praekelt.Org) using Turn machine learning technology on WhatsApp Messenger. It’s free to use. Add a contact on your phone with the number: +41798931892. Open WhatsApp messenger and initiate a conversation with the first message as “hi”. You’ll get a response with a number of options as an interactive menu with topics ranging from Travel Advice, latest Covid-19 numbers, how to stay protected and more.

Transmission Rate of the Covid-19 virus SARS-CoV-2:

The Transmission Rate or the Reproductive Number (Ro – read as ‘R nought’) – indicates the number of newly infected people from a single case. Preliminary studies had estimated the Ro for Covid-19 to be between 1.5 and 3.5. For comparison, the Ro for The common flu has a Ro of 1.3 and SARS had a Ro of 2.0. Generally, an outbreak with a Ro less than 1 should disappear gradually. So you can imagine why social distancing is so crucial – one person can affect up to 3, and those 3, in turn, start affecting up-to a 3 each and thus it grows exponentially.

The number of Covid-19 confirmed cases reported around the globe is only what gets tested which obviously does not comply with a robust Measurement System Analysis – thus leaving us with the reported official figures to rely on, for our analysis and inferences. For example, the reported vs actual could have a huge discrepancy according to this report from businessinsider as a case in point.

To know how far you are from the nearest Covid-19 confirmed case (For India), go-to

How Misleading Visuals Can Be…

Many popular Covid-19 Dashboards have visuals that are grossly misrepresented, although the numbers are right. As a case in point, take a look at the wildly popular dashboard from the prestigious Johns Hopkins University, as on Apr 9, 2020. (btw, I’m a huge fan of their work and faculty – taken a few data science courses on

It gives you a sense that over half of the land area in the World is affected in very large numbers by Covid-19. This is totally not the case. The issue is with how we choose to represent volumes on a map. While the number is true, when showing it up on a map – it would make more sense if they would compare it with the total population or the size of the geographical area affected. I have written to the contact @JohnsHopkins Website to make it more meaningful – let’s see where that goes! Take a look at the table below to see what I’m saying.

United States33100265121897666615.550.66%
South Korea512691854668049104.960.91%
United Kingdom678860112327083427.920.34%
Czech Republic10708981986819214.790.92%
New Zealand4822233468759720.600.97%
Costa Rica509411859631170.570.12%
Data as on first week of April 2020, Sorted by Tests Performed. Raw Data Source: Wordometer. Computation and Illustration: @vimaloctavius

2.18Mn tests performed so far is a huge number in the US, but it’s merely 0.7% of the population – and that’s just total tested and not positive. Out of these are the 432,596 Covid-19 positive cases and that’s about 0.13% (Population of US taken as 331,002,651).

While Japan has the capacity to test in thousands per day, as per this report, their health officials are choosing not to test unless the case is severe and are reserving the Health worker’s time and effort to care for the neediest. One thing is for sure – for many years, wearing a mask, using hand sanitizers often and keeping public places and surroundings super clean has been a part of their culture – which could be playing a big role in containing the spread. Even these measures are put to test with Japan declaring a State of Emergency until May 6 2020 and advising people to reduce their contact with others by 70-80%.

The Movie “The Gods Must Be Crazy” comes to my mind. While the entire world seems to have come to a standstill and we feel that Covid-19 is everywhere and in everything we read – there is an unimaginably big portion of the population who are maybe unaware of what’s exactly going on. The movie starts by comparing the Urban life with that of the Bushmen, of the Kalahari desert – who live in harmony in the wild, completely unaware of the civilized world. It’s humbling to think that such a tiny amount of the affected population in the world (0.02% Confirmed Covid-19, 1503900/7776637000) could stop the world – by an ever so tiny virus that no one has ever seen with bare eyes. Is it the Economic Divide, or the fear of the unknown? I would want to visualize the correlation between the GDP density and Covid Cases – hmm… mañana. I’m in no way implying to downplay the seriousness – a lockdown is definitely essential or more lives will be lost.

When the disease has run its course affecting whoever it can takedown, nature would eventually balance itself out and things would be normal again.

Community Mobility Reports:

Google has started publishing the impact of social distancing – in visuals, through their Community Mobility Reports.

Google’s Community Mobility Report: India

Google’s Community Mobility Report: USA

In my previous update – I had covered the aspect that the virus moves from household to household too. Social Distancing away from work and recreational areas is great to flatten the curve. Not quarantined at home and moving around and socializing within our neighborhood – how effective is that going to be in stopping the virus? What options do we have?

AarogyaSetu is a mobile application developed by the Government of India. It gives you realtime alerts if you are at close proximity to a Covid-19 Confirmed Case. Download from links below (Available in Play Store)


On a side note, I had shared this idea on , taking best practices from South Korea which had implemented a similar app to alert citizens and stay safe. Whether the app is an outcome of the suggestion or not I do not know but I’m glad we have already started rolling it out.

Snapshot of Idea Submitted on a fortnight ago. Go Ahead! Submit your ideas

Some good precautions help as listed out by WHO in their YouTube Channel.

Below are some forecasts:

For USA – Blue line is the point forecast, going up-to 1,252,874 Covid-19 Confirmed by the 6th of May, and could range between 983,997- 1,521,750 (Dark Grey Shade of 80% and light Grey Shade of 95% Prediction Intervals). Raw Data Source: JHU CSSE, Computation and Illustration in R: @vimaloctavius
See Terms Of Use and Disclaimer*
For India – Blue line is the point forecast, going up-to 34,413 Covid-19 Confirmed by the 6th of May, and could range between 18,198 – 50,627 (Dark Grey Shade of 80% and light Grey Shade of 95% Prediction Intervals). Raw Data Source: JHU CSSE, Computation and Illustration in R: @vimaloctavius
See Terms Of Use and Disclaimer*

Published: Apr 10, 2020 by @vimaloctavius. Connect with me on LinkedIn

Related Topics: Covid-19 Update#1

*terms of use and disclaimer:

Prediction Models are subjected to change as data evolves daily. Although the models are built on actual data, they have not been verified or endorsed by any governing or health care bodies – it is purely from a study perspective and not to be used for decision making. All the information provided is free and shareable – please include the source.