# River School — trained adviser

This game uses Sage, the full trained pytoolz/WSRF equity forest from Distance Draw. It does not claim a separate model fit.

[Read the complete training method, fresh tests, ablation, timing, and source receipts](../distance-draw/TRAINING.md).

The adviser predicts uniform-opponent showdown share. Exact card enumeration separately scores the challenge after you choose. The brain shows actual model inputs and tree paths, not the exact scoring answer.