Analytics Coding Education STEM

Tracking Actual vs. My Predictions 4/29/20

Back in March 2020, I did some predictive modeling using simple math and Excel where I presented 3 scenarios: Likely, Best-case, Worst-case. You can read the original blog here: What-If Models (COVID-19): Results

Let’s take a look at today’s latest data from WHO and compare my Model 1: Likely Scenario numbers. We’ll compare the USA numbers as my modeling were based on my observation of current social behaviors in the USA.

WHO Data as of 4/29/2020 for USA:

# of Cases: 983,457

# of Deaths (to-date): 50,492

Now, flashback to my Model 1 chart below:

My Model 1 Predictions:

# of Cases: between 622,800 to 942,805 

# of Deaths (to-date): 50,213

I know, I had to double-take too! This is incredibly close to what’s occurring/occurred despite innumerable variables. However, you might think it’s just a coincidence and random guess. It’s not, this is the second time I’ve compared so far at different timeframes and both times, they were essentially spot-on. You can see the first time I compared in the blog here.

To recap:

# Cases to-date:

  • Actual Cases to-date: 983,457
  • My Prediction: between 622,800 to 942,805

# Deaths to-date:

  • Actual Total Deaths to-date: 50,492
  • My Prediction: 50,213

My Model 1 also sees a drop in cases from its day 100 (which is about 2 weeks from today).

Also note that in my modeling, 100% population or 1M (of hypothetical country of population 1 Million) will be infected by around 90 days. Which is exactly the trend we’re seeing in USA…approaching 1M cases very soon.


Stay in. Stay safe. Stay informed!

Leave a Reply

Your email address will not be published. Required fields are marked *

Back To Top