This post explores the application of numerical methods to analyze the initial growth trends of the COVID-19 outbreak using real-world data from Johns Hopkins CSSE. It utilizes a Quadratic Least Squares polynomial to model the time-series data of new confirmed cases, effectively smoothing inherent data noise. Subsequently, Richardson’s Extrapolation is applied to a 3-point midpoint formula to calculate the instantaneous rate of change (first derivative) of the new cases at various time intervals. The analysis reveals a significant deceleration in the growth rate over the studied period, suggesting the positive impact of early social distancing measures. The computational methods were implemented in MATLAB and Java, demonstrating robust modeling and high-precision results from low-order formulas.