From 6aa9002e92567770c538a82aab580158a170707b Mon Sep 17 00:00:00 2001
From: David Kruger <david@krugerlabs.us>
Date: Mon, 30 Jun 2025 21:41:49 -0700
Subject: [PATCH] GameState::reward_for_player is no longer a trait function

We need GameState::rewards_for_players implemented, the library
currently doesn't need to specify for a singular player.
---
 benches/e2e.rs               | 26 +++++++++++++-------------
 examples/auto_tic_tac_toe.rs | 26 +++++++++++++-------------
 examples/tic_tac_toe.rs      | 26 +++++++++++++-------------
 src/state.rs                 | 11 ++++-------
 4 files changed, 43 insertions(+), 46 deletions(-)

diff --git a/benches/e2e.rs b/benches/e2e.rs
index b8f27a8..11397c0 100644
--- a/benches/e2e.rs
+++ b/benches/e2e.rs
@@ -125,6 +125,19 @@ impl TicTacToe {
 
         None
     }
+
+    fn reward_for_player(&self, player: &Player) -> RewardVal {
+        if let Some(winner) = self.get_winner() {
+            if winner == *player {
+                return 1.0; // Win
+            } else {
+                return 0.0; // Loss
+            }
+        }
+
+        // Draw
+        0.5
+    }
 }
 
 impl GameState for TicTacToe {
@@ -161,19 +174,6 @@ impl GameState for TicTacToe {
         self.get_winner().is_some() || self.moves_played == 9
     }
 
-    fn reward_for_player(&self, player: &Self::Player) -> RewardVal {
-        if let Some(winner) = self.get_winner() {
-            if winner == *player {
-                return 1.0; // Win
-            } else {
-                return 0.0; // Loss
-            }
-        }
-
-        // Draw
-        0.5
-    }
-
     fn rewards_for_players(&self) -> HashMap<Self::Player, RewardVal> {
         HashMap::from_iter(vec![
             (Player::X, self.reward_for_player(&Player::X)),
diff --git a/examples/auto_tic_tac_toe.rs b/examples/auto_tic_tac_toe.rs
index 1183ea5..819a5f6 100644
--- a/examples/auto_tic_tac_toe.rs
+++ b/examples/auto_tic_tac_toe.rs
@@ -148,6 +148,19 @@ impl TicTacToe {
 
         None
     }
+
+    fn reward_for_player(&self, player: &Player) -> RewardVal {
+        if let Some(winner) = self.get_winner() {
+            if winner == *player {
+                return 1.0; // Win
+            } else {
+                return 0.0; // Loss
+            }
+        }
+
+        // Draw
+        0.5
+    }
 }
 
 impl GameState for TicTacToe {
@@ -184,19 +197,6 @@ impl GameState for TicTacToe {
         self.get_winner().is_some() || self.moves_played == 9
     }
 
-    fn reward_for_player(&self, player: &Self::Player) -> RewardVal {
-        if let Some(winner) = self.get_winner() {
-            if winner == *player {
-                return 1.0; // Win
-            } else {
-                return 0.0; // Loss
-            }
-        }
-
-        // Draw
-        0.5
-    }
-
     fn rewards_for_players(&self) -> HashMap<Self::Player, RewardVal> {
         HashMap::from_iter(vec![
             (Player::X, self.reward_for_player(&Player::X)),
diff --git a/examples/tic_tac_toe.rs b/examples/tic_tac_toe.rs
index c0bc3ec..09053b9 100644
--- a/examples/tic_tac_toe.rs
+++ b/examples/tic_tac_toe.rs
@@ -191,6 +191,19 @@ impl TicTacToe {
 
         None
     }
+
+    fn reward_for_player(&self, player: &Player) -> RewardVal {
+        if let Some(winner) = self.get_winner() {
+            if winner == *player {
+                return 1.0; // Win
+            } else {
+                return 0.0; // Loss
+            }
+        }
+
+        // Draw
+        0.5
+    }
 }
 
 impl GameState for TicTacToe {
@@ -227,19 +240,6 @@ impl GameState for TicTacToe {
         self.get_winner().is_some() || self.moves_played == 9
     }
 
-    fn reward_for_player(&self, player: &Self::Player) -> RewardVal {
-        if let Some(winner) = self.get_winner() {
-            if winner == *player {
-                return 1.0; // Win
-            } else {
-                return 0.0; // Loss
-            }
-        }
-
-        // Draw
-        0.5
-    }
-
     fn rewards_for_players(&self) -> HashMap<Self::Player, RewardVal> {
         HashMap::from_iter(vec![
             (Player::X, self.reward_for_player(&Player::X)),
diff --git a/src/state.rs b/src/state.rs
index 2048962..250b4f4 100644
--- a/src/state.rs
+++ b/src/state.rs
@@ -40,11 +40,11 @@ pub trait GameState: Clone + Debug {
     /// instead should modify a copy of the state and return that.
     fn state_after_action(&self, action: &Self::Action) -> Self;
 
-    /// Returns the reward from the perspective of the given player for the game state
+    /// Returns the rewards for all players from their perspective for the game state
     ///
-    /// This evaluates the current state from the perspective of the given player, and
-    /// returns the reward indicating how good of a result the given state is for the
-    /// player.
+    /// This evaluates the current state from the perspective of each player, and
+    /// returns a HashMap mapping each player to the result of this evaluation, which
+    /// we call the reward.
     ///
     /// This is used in the MCTS backpropagation and simulation phases to evaluate
     /// the value of a given node in the search tree.
@@ -55,9 +55,6 @@ pub trait GameState: Clone + Debug {
     /// - 0.0 => a loss for the player
     ///
     /// Other values can be used for relative wins or losses
-    fn reward_for_player(&self, player: &Self::Player) -> RewardVal;
-
-    /// Returns the rewards for all players at the current state
     fn rewards_for_players(&self) -> HashMap<Self::Player, RewardVal>;
 
     /// Returns the player whose turn it is for the game state