Covid-19 and Sports Update
Thanks for visiting the Oracle page. The site was due to be updated with NBA and NHL information right as we started the quarantine in March. The site will not be regularly updated while the pandemic is going on.
Instead of updating for sports, I will include here some work I have been doing on Covid-19. This work was sponsored by a grant from the Associated Colleges of the South (ACS).
I was inspired by A guide to R — the pandemic’s misunderstood metric and the computational work from rt.live and decided to look how to compute Rt values at a county level.
The effective reproductive rate (Rt) is an important tool to measure the transmission potential of a disease. It is the average number of secondary infections produced by a typical case of an infection.
- If Rt > 1, cases will increase.
- When Rt = 1, the disease is endemic.
- If Rt < 1, cases will decline.
To eliminate a disease from a population, Rt needs to be less than 1.
Here are our estimates for Rt values for Texas and the top four counties (in total number of cases). An explanation follows below the graphs.
New Cases (7-day moving averages) and Rt
When a virus begins to circulate, the basic reproduction number (R0) is the average number of secondary infections produced by a typical case of an infection in a population where everyone is susceptible. Most importantly, for an epidemic to occur in a susceptible population R0 must be >1.
As the epidemic progresses, we measure the effective reproductive rate (Rt), as suggested by the time variable, Rt changes as people change their behaviors. For instance, we start to weak masks, develop better hygiene, improved treatments, increase social distance, etc.
It is an important problem to measure R0 and Rt for any disease, in particular Covid-19. A challenging problem in mathematics is to estimate Rt based on daily cases. The site rt.live provides up-to-date estimates of Rt for every US state. The basis for their work is a paper by Bettencourt & Ribeiro where we assume an SIR model and (roughly) estimate the change in daily cases as a function of Rt. The beautifully written blog by Ramnath Vaidyanathan at DataCamp not only explains some more of the mathematics, but also provides nice R code.
Using these two resources as inspiration, I began collecting data from USAfacts and started to compute Rt at a county level. As the pandemic progresses and there are few sports to occupy the Oracle, we will be working on improving and validating this interesting work.
We also have been playing around with some visualizations, you can check these Tableau maps that I hope to update periodically.
If you read this far, thank you for reading. If you have questions, please contact me. I hope to provide source code on github (once I figure out how that works), but I can share files on request.