This page shows win probabilities, odds, and a bit about the random forest model built to make predictions for the next DGPT event or major.. These predictions are experimental and will probably change a lot moving forward. It is definitely giving some interesting, even spicy, expectations – which is not necessarily a bad thing.

The probabilities (and odds) are generated using a random forest model because, so far, it has been the most predictive. In leave-one-out cross-validation using data from 2020 to the present it has predicted the winner of DGPT or major events 62% of the time. More about the model can be found in the stats glossary.