Create a model and tag it in DB. The DB has a tag, which in the context of he server can ce used o identify and construct a predictor.
1.1 Identify features. The model code for a given predictor knows how to extract features from a 'Schedule' and XStatStore.
1.2 Train model. The parameters
1.3 Save parameters
1.4 Show Errors
1.5 Show Predixitons
Prediction workflow:
Types of Models
Favorite/Dog Spread Odds Over/Under
Types of Errors