diff --git a/benchmark/problem_stats.py b/benchmark/problem_stats.py
index 2ee8a0890..202942f13 100755
--- a/benchmark/problem_stats.py
+++ b/benchmark/problem_stats.py
@@ -15,7 +15,7 @@ HARD_SET_NUM = 3  # Number of models that defines the hard set threshold
 
 def get_dirs_from_leaderboard():
     # Load the leaderboard data
-    with open("aider/website/_data/edit_leaderboard.yml") as f:
+    with open("aider/website/_data/polyglot_leaderboard.yml") as f:
         leaderboard = yaml.safe_load(f)
     return [(entry["dirname"], entry["model"]) for entry in leaderboard]
 
@@ -92,7 +92,7 @@ def analyze_exercise_solutions(dirs=None, topn=None, copy_hard_set=False):
                     (
                         entry["pass_rate_2"]
                         for entry in yaml.safe_load(
-                            open("aider/website/_data/edit_leaderboard.yml")
+                            open("aider/website/_data/polyglot_leaderboard.yml")
                         )
                         if entry["dirname"] == dirname
                     ),