diff --git a/src/dcegm/final_periods.py b/src/dcegm/final_periods.py
index e79d2523..1cefe7ba 100644
--- a/src/dcegm/final_periods.py
+++ b/src/dcegm/final_periods.py
@@ -229,7 +229,7 @@ def solve_final_period_discrete(
     )
     # Choose which draw we take for policy and value function as those are not
     # saved with respect to the draws
-    middle_of_draws = int(value.shape[2] + 1 / 2)
+    middle_of_draws = int((value.shape[2] - 1) / 2)
     # Select solutions to store
     value_final = value[:, :, middle_of_draws]
 
diff --git a/tests/test_discrete_versus_continuous_experience.py b/tests/test_discrete_versus_continuous_experience.py
index 7113c25e..9e6537fd 100644
--- a/tests/test_discrete_versus_continuous_experience.py
+++ b/tests/test_discrete_versus_continuous_experience.py
@@ -17,7 +17,7 @@
 N_DISCRETE_CHOICES = 2
 MAX_WEALTH = 50
 WEALTH_GRID_POINTS = 100
-EXPERIENCE_GRID_POINTS = 6
+EXPERIENCE_GRID_POINTS = 5
 
 
 PARAMS = {
@@ -129,6 +129,7 @@ def test_setup():
 @pytest.mark.parametrize(
     "period, experience, lagged_choice, choice",
     [
+        (1, 0, 1, 0),
         (1, 0, 1, 0),
         (1, 1, 0, 0),
         (2, 1, 0, 1),
@@ -220,5 +221,5 @@ def test_replication_discrete_versus_continuous_experience(
             params=PARAMS,
         )
 
-        aaae(value_cont_interp, value_disc_interp, decimal=1e-6)
-        aaae(policy_cont_interp, policy_disc_interp, decimal=1e-6)
+        aaae(value_cont_interp, value_disc_interp, decimal=4)
+        aaae(policy_cont_interp, policy_disc_interp, decimal=4)