diff --git a/experiments/elo/notebooks/BGG ranking 2.py b/experiments/elo/notebooks/BGG ranking 2.py
new file mode 100644
index 0000000000000000000000000000000000000000..89cf1ecb5f2492ba7195e2a113b40d23f254e8b6
--- /dev/null
+++ b/experiments/elo/notebooks/BGG ranking 2.py	
@@ -0,0 +1,67 @@
+# ---
+# jupyter:
+#   jupytext:
+#     formats: ipynb,py:percent
+#     text_representation:
+#       extension: .py
+#       format_name: percent
+#       format_version: '1.3'
+#       jupytext_version: 1.17.2
+#   kernelspec:
+#     display_name: Python 3 (ipykernel)
+#     language: python
+#     name: python3
+# ---
+
+# %%
+import jupyter_black
+import polars as pl
+from elo.elo_ratings import TwoPlayerElo
+from itertools import combinations
+
+jupyter_black.load()
+
+# %%
+ratings = (
+    pl.scan_ndjson("../../../board-game-data/scraped/bgg_RatingItem.jl")
+    .select("bgg_id", "bgg_user_name", "bgg_user_rating", "updated_at")
+    .drop_nulls()
+    .sort("updated_at")
+    .tail(1_000_000)
+    # .collect()
+)
+self_join = ratings.join(
+    ratings, on="bgg_user_name", how="left", suffix="_prev"
+).filter(pl.col("updated_at_prev") < pl.col("updated_at"))
+df_1 = self_join.filter(
+    pl.col("bgg_user_rating") > pl.col("bgg_user_rating_prev")
+).select(
+    game_1="bgg_id",
+    game_2="bgg_id_prev",
+    updated_at="updated_at",
+    updated_at_prev="updated_at_prev",
+)
+df_2 = self_join.filter(
+    pl.col("bgg_user_rating") < pl.col("bgg_user_rating_prev")
+).select(
+    game_1="bgg_id_prev",
+    game_2="bgg_id",
+    updated_at="updated_at",
+    updated_at_prev="updated_at_prev",
+)
+df = (
+    pl.concat([df_1, df_2])
+    .sort("updated_at", "updated_at_prev")
+    .select("game_1", "game_2")
+    .collect()
+)
+df.shape
+
+# %%
+elo = TwoPlayerElo(elo_k=1.0)
+elo.update_elo_ratings_batch(
+    df.iter_rows(), progress_bar=True, tqdm_kwargs={"total": len(df)}
+)
+
+# %%
+sorted(elo.elo_ratings.items(), key=lambda x: -x[1])[:100]
diff --git a/experiments/elo/notebooks/BGG ranking.py b/experiments/elo/notebooks/BGG ranking.py
new file mode 100644
index 0000000000000000000000000000000000000000..1a10db8d93b180705a37ed8bf7245716f4e773b6
--- /dev/null
+++ b/experiments/elo/notebooks/BGG ranking.py	
@@ -0,0 +1,59 @@
+# ---
+# jupyter:
+#   jupytext:
+#     formats: ipynb,py:percent
+#     text_representation:
+#       extension: .py
+#       format_name: percent
+#       format_version: '1.3'
+#       jupytext_version: 1.17.2
+#   kernelspec:
+#     display_name: Python 3 (ipykernel)
+#     language: python
+#     name: python3
+# ---
+
+# %%
+import jupyter_black
+import polars as pl
+from elo.elo_ratings import TwoPlayerElo
+from itertools import combinations
+
+jupyter_black.load()
+
+# %%
+users = (
+    pl.scan_ndjson("../../../board-game-data/scraped/bgg_RatingItem.jl")
+    .select("bgg_id", "bgg_user_name", "bgg_user_rating")
+    .drop_nulls()
+    .group_by("bgg_user_name")
+    .agg(
+        num_rated_games=pl.len(),
+        ratings=pl.struct("bgg_id", "bgg_user_rating").shuffle(),
+    )
+    .filter(pl.col("num_rated_games") > 1)
+    .select(pl.col("ratings").shuffle())
+    .collect()["ratings"]
+)
+len(users)
+
+# %%
+users.sample(10)
+
+
+# %%
+def matches_from_ratings(users):
+    for user in users:
+        for game_1, game_2 in combinations(user, 2):
+            if game_1["bgg_user_rating"] > game_2["bgg_user_rating"]:
+                yield game_1["bgg_id"], game_2["bgg_id"]
+            elif game_1["bgg_user_rating"] < game_2["bgg_user_rating"]:
+                yield game_2["bgg_id"], game_1["bgg_id"]
+
+
+# %%
+elo = TwoPlayerElo()
+elo.update_elo_ratings_batch(matches_from_ratings(users), progress_bar=True)
+
+# %%
+sorted(elo.elo_ratings.items(), key=lambda x: -x[1])[:100]