Remove obsolete documentation about GameOverEvaluator

crates/engine/src/evaluator.rs

@@ -3,8 +3,7 @@ use crate::position::Position;
 use crate::probabilities::Probabilities;
 
 /// [Evaluator] is one of the central parts of the engine. Implementing structs only have to
-/// implement the function [Evaluator::eval], Examples are [crate::composite::GameOverEvaluator]
-/// and `RolloutEvaluator`.
+/// implement the function [Evaluator::eval], Examples are `OnnxEvaluator` and `RolloutEvaluator`.
 ///
 /// The function [Evaluator::eval_batch] is implemented by default, building on `eval`.
 /// If there is way to optimize evaluations by looking at all legal moves at once, then don't

docs/dev/engine.md

@@ -41,9 +41,9 @@ For a given position in cubeless money game, it returns the probabilities to win
 https://github.com/carsten-wenderdel/wildbg/blob/d5c7280a60a52cb61c92af78018fb811cf3dd223/crates/engine/src/evaluator.rs#L31
 
 Examples of implementations are
-- `GameOverEvaluator` for positions where the game is over and the probabilities are known.
+
 - `OnnxEvaluator` which uses a neural net in `ONNX` format to calculate the probabilities.
-- `CompositeEvaluator` which consists of a `GameOverEvaluator` and other evaluators for different game phases. It decides which evaluator to use based on the position.
+- `CompositeEvaluator` which consists of various `OnnxEvaluator`s for different game phases.
 - `RolloutEvaluator` which uses another evaluator to do rollouts.
 
 ## Inputs generation