One-Off Coder is helping to fight against the spread of COVID-19 by exploring and experimenting with differential diagnosis models. In this post, we use Bayesian Belief Networks (BBN) to perform diagnostic reasoning from symptoms back to the cause. This approach can help patients be better triaged for screening and testing of COVID-19 by clinics and hospitals, and also patients have a way to self-assess their symptoms for COVID-19. This model can answer a question like the following: given that you have a fever and are fatigued, what is the probability of you having COVID-19? This code is also published online.