The Evolution of Yuing: Revolutionizing Basketball Predictions

The initial collaboration between Mag and Vulgaris was challenging, with Mag initially believing he could create a working model alone. The program, named Yuing, was designed to simulate basketball matches and predict outcomes by breaking down the game into possessions.

Each simulation consisted of a series of mini-simulations, predicting outcomes of possessions and using historical data. The program was named Yuing after the 'Yuing theory' described by Bill Simmons, which posits that the departure of an overrated superstar strengthens the team. An additional model, Van Gandi, was created to predict team compositions and playing time for each player.

A database named Mori was developed to track player movements, draft picks, and acquisition trends. The initial version of Yuing had a bug where the free throw percentage was set at 88% instead of the correct 75%. Yuing evaluates player usefulness by considering their offensive and defensive value, which changes based on who they are guarding.