diff --git a/server/card/equipment/artifact/old_version.py b/server/card/equipment/artifact/old_version.py
index c2967811..36fc85e6 100644
--- a/server/card/equipment/artifact/old_version.py
+++ b/server/card/equipment/artifact/old_version.py
@@ -1,7 +1,10 @@
-from typing import Literal
+from typing import Any, List, Literal
+
+from ....action import Actions
 
 from ....struct import Cost
 from .element_artifacts import SmallElementalArtifact as SEA_4_0
+from .others import GamblersEarrings as GE_3_8
 
 
 class SmallElementalArtifact(SEA_4_0):
@@ -9,4 +12,16 @@ class SmallElementalArtifact(SEA_4_0):
     cost: Cost = Cost(same_dice_number = 2)
 
 
-OldVersionArtifacts = SmallElementalArtifact | SmallElementalArtifact
+class GamblersEarrings(GE_3_8):
+    version: Literal['3.3']
+    usage: int = 999
+
+    def equip(self, match: Any) -> List[Actions]:
+        """
+        Equip this artifact. Reset usage.
+        """
+        self.usage = 999
+        return []
+
+
+OldVersionArtifacts = SmallElementalArtifact | GamblersEarrings
diff --git a/tests/server/cards/test_artifacts.py b/tests/server/cards/test_artifacts.py
index 4b273d11..b8c7029d 100644
--- a/tests/server/cards/test_artifacts.py
+++ b/tests/server/cards/test_artifacts.py
@@ -188,7 +188,7 @@ def test_create_small_element_artifacts():
     assert match.state != MatchState.ERROR
 
 
-def test_old_version_artifacts():
+def test_old_version_elemental_artifacts():
     agent_0 = NothingAgent(player_idx = 0)
     agent_1 = InteractionAgent(
         player_idx = 1,
@@ -398,8 +398,117 @@ def test_gambler():
     assert match.state != MatchState.ERROR
 
 
+def test_old_gambler():
+    agent_0 = InteractionAgent(
+        player_idx = 0,
+        verbose_level = 0,
+        commands = [
+            "sw_card",
+            "choose 0",
+            "end",
+            "choose 1",
+            "choose 2",
+            "choose 3",
+            "choose 4",
+            "choose 5",
+            "choose 6",
+        ],
+        only_use_command = True
+    )
+    agent_1 = InteractionAgent(
+        player_idx = 1,
+        verbose_level = 0,
+        commands = [
+            "sw_card",
+            "choose 0",
+            "card 0 1 1",
+            "skill 0 0 1 2",
+            "TEST 1 16 12",
+            "card 0 0 1",
+            "skill 0 0 1 2",
+            "TEST 1 16 10",
+            "skill 0 0 1 2",
+            "skill 0 0 1 2",
+            "TEST 1 16 8",
+            "skill 0 0 1 2",
+            "TEST 1 16 7",
+            "skill 0 0 1 2",
+            "TEST 1 16 6",
+            "end",
+        ],
+        only_use_command = True
+    )
+    deck = Deck.from_str(
+        '''
+        # charactor:Fischl
+        # charactor:Mona
+        # charactor:Nahida
+        charactor:Nahida*10
+        # Gambler's Earrings*2
+        # Wine-Stained Tricorne*2
+        # Timmie*2
+        # Rana*2
+        # Strategize*2
+        # The Bestest Travel Companion!*2
+        # Covenant of Rock
+        # Gambler's Earrings*30
+        '''
+    )
+    old_gambler = {'name': "Gambler's Earrings", 'version': '3.3'}
+    deck_dict = deck.dict()
+    deck_dict['cards'] += [old_gambler] * 30
+    deck = Deck(**deck_dict)
+    for charactor in deck.charactors:
+        charactor.hp = 1
+        charactor.max_hp = 1
+    match = Match(random_state = get_random_state())
+    match.config.max_same_card_number = 30
+    match.config.charactor_number = 10
+    match.config.random_first_player = False
+    match.set_deck([deck, deck])
+    set_16_omni(match)
+    match.start()
+    match.step()
+    while True:
+        if match.need_respond(0):
+            while True:
+                cmd = agent_0.commands[0]
+                test_id = get_test_id_from_command(agent_0)
+                if test_id == 1:
+                    number = [int(x) for x in cmd.strip().split(' ')[-2:]]
+                    assert number[0] == len(match.player_tables[0].dice.colors)
+                    assert number[1] == len(match.player_tables[1].dice.colors)
+                elif test_id == 2:
+                    charactor = match.player_tables[0].charactors[7]
+                    assert charactor.artifact is not None
+                    assert charactor.artifact.name == "Gambler's Earrings"
+                else:
+                    break
+            make_respond(agent_0, match)
+        elif match.need_respond(1):
+            while True:
+                cmd = agent_1.commands[0]
+                test_id = get_test_id_from_command(agent_1)
+                if test_id == 1:
+                    number = [int(x) for x in cmd.strip().split(' ')[-2:]]
+                    assert number[0] == len(match.player_tables[0].dice.colors)
+                    assert number[1] == len(match.player_tables[1].dice.colors)
+                else:
+                    break
+            make_respond(agent_1, match)
+        else:
+            raise AssertionError('No need respond.')
+        if len(agent_1.commands) == 0 and len(agent_0.commands) == 0:
+            break
+
+    assert len(agent_1.commands) == 0 and len(agent_0.commands) == 0
+
+    assert match.state != MatchState.ERROR
+
+
 if __name__ == '__main__':
     # test_small_elemental_artifacts()
     # test_create_small_element_artifacts()
     # test_old_version_artifacts()
-    test_gambler()
+    # test_gambler()
+    test_old_gambler()