Thread created to discuss this whitepaper.
Over the course of the 2017 season, I weekly published an event simulator which provided live predictions for events. In addition to predicting the winner of each match, I provided predictions for multiple distinct scoring categories. Most of these predictions followed the same basic formula, but the parameters for the predictions in each category were unique, and changed throughout the season as I provided updates. These parameters were those which provided the most predictive power for the previous week’s qualification matches. I feel that sharing the parameters from each category can provide some level of insight into the dynamics of that scoring category, and by extension, the dynamics of the entire game. Doing this will also give me an opportunity to shed some light on how I created my predictions.
If you have any questions or suggestions for improvements to this paper, feel free to let me know. It’s been a while since I have made something like this, so I would appreciate useful feedback.