diff --git a/reasoning_gym/games/countdown.py b/reasoning_gym/games/countdown.py
index 75552494..c8da71a5 100644
--- a/reasoning_gym/games/countdown.py
+++ b/reasoning_gym/games/countdown.py
@@ -3,6 +3,7 @@ from dataclasses import dataclass
 from random import Random
 from typing import Any, Optional
 
+import numpy as np
 import sympy
 from sympy import Symbol, symbols
 from sympy.parsing.sympy_parser import parse_expr
@@ -23,6 +24,16 @@ Final answer format instructions:
 
 DATASET_NAME = "countdown"
 
+_num_re = re.compile(r"\b\d+\b")  # pre-compile once, reuse
+
+
+def _extract_ints(expr_str: str) -> list[int]:
+    """
+    Fast path: grab the literal integers that appear in the source text.
+    Handles duplicates correctly (e.g. “1 + 1 + 81” ⇒ [1, 1, 81]).
+    """
+    return [int(m) for m in _num_re.findall(expr_str)]
+
 
 @dataclass
 class CountdownConfig:
@@ -192,12 +203,14 @@ class CountdownDataset(ProceduralDataset):
             return reward
 
         try:
-            answer = answer.strip()
-            user_answer = int(parse_expr(answer))
-            used_numbers = [int(num) for num in re.findall(r"\b\d+\b", answer)]
-            target_numbers = set(entry["metadata"]["numbers"])
+            user_answer = float(parse_expr(answer))
+            used_numbers = _extract_ints(answer)
+            target_numbers = entry["metadata"]["numbers"]
 
-            if (user_answer == entry["metadata"]["target"]) and (set(used_numbers) == target_numbers):
+            if sorted(used_numbers) != sorted(target_numbers):
+                return 0.05
+
+            if np.isclose(user_answer, entry["metadata"]["target"], atol=1e-6):
                 return 1.0
 
             return 0.05 if answer else 0.01
diff --git a/tests/test_countdown.py b/tests/test_countdown.py
index 2070a6dc..6e3a1e0a 100644
--- a/tests/test_countdown.py
+++ b/tests/test_countdown.py
@@ -100,6 +100,20 @@ def test_answer_without_all_numbers():
     assert dataset.score_answer(answer=answer, entry=item) == 0.05
 
 
+def test_edge_cases_1():
+    dataset = CountdownDataset(CountdownConfig(size=10, seed=42))
+    answer = "1*81"
+    item = {"metadata": {"numbers": [1, 1, 1, 81], "target": 81}}
+    assert dataset.score_answer(answer=answer, entry=item) != 1.0
+
+
+def test_edge_cases_2():
+    dataset = CountdownDataset(CountdownConfig(size=10, seed=42))
+    answer = "6*34/11-1"
+    item = {"metadata": {"numbers": [6, 34, 1, 11], "target": 17}}
+    assert dataset.score_answer(answer=answer, entry=item) != 1.0
+
+
 def test_countdown_game_randomization():
     """Test number randomization configuration"""
     config = CountdownConfig(min_numbers=4, max_numbers=4, shuffle=False, size=10, seed=42)  # Fixed size for testing