diff --git a/sc2ts/inference.py b/sc2ts/inference.py
index a6f032d..1670210 100644
--- a/sc2ts/inference.py
+++ b/sc2ts/inference.py
@@ -565,6 +565,8 @@ def extend(
         hmm_cost_threshold = 5
     if min_group_size is None:
         min_group_size = 10
+    if retrospective_window is None:
+        retrospective_window = 30
 
     check_base_ts(base_ts)
     logger.info(
diff --git a/tests/test_inference.py b/tests/test_inference.py
index ba09948..846124a 100644
--- a/tests/test_inference.py
+++ b/tests/test_inference.py
@@ -10,15 +10,14 @@
 
 
 class TestSolveNumMismatches:
-
     @pytest.mark.parametrize(
         ["k", "expected_rho"],
-        [(2, 0.02295918), (3, 0.00327988), (4, 0.00046855), (1000, 0)],
+        [(2, 0.0001904), (3, 2.50582e-06), (4, 3.297146e-08), (1000, 0)],
     )
     def test_examples(self, k, expected_rho):
-        rho, mu = sc2ts.solve_num_mismatches(k, num_sites=2)
-        assert mu[0] == 0.125
-        nt.assert_almost_equal(rho[0], expected_rho)
+        mu, rho = sc2ts.solve_num_mismatches(k)
+        assert mu == 0.0125
+        nt.assert_almost_equal(rho, expected_rho)
 
 
 class TestInitialTs:
@@ -204,10 +203,21 @@ def test_two_samples_one_mutation_one_filtered(self, tmp_path):
 
 
 class TestMatchTsinfer:
-    def match_tsinfer(self, samples, ts, **kwargs):
+    def match_tsinfer(self, samples, ts, mirror_coordinates=False, **kwargs):
         sc2ts.inference.match_tsinfer(
-            samples=samples, ts=ts, num_mismatches=1000, **kwargs
+            samples=samples,
+            ts=ts,
+            mu=0.125,
+            rho=0,
+            mirror_coordinates=mirror_coordinates,
+            **kwargs,
         )
+        if mirror_coordinates:
+            # Quick hack to make the tests here work, as they use
+            # attributes defined by the forward path.
+            for sample in samples:
+                sample.forward_path = sample.reverse_path
+                sample.forward_mutations = sample.reverse_mutations
 
     @pytest.mark.parametrize("mirror", [False, True])
     def test_match_reference(self, mirror):
@@ -406,8 +416,8 @@ def test_n_samples_metadata(self):
                     strain=strain,
                     date=date,
                     metadata={f"x{j}": j, f"y{j}": list(range(j))},
-                    path=[(0, ts.sequence_length, 1)],
-                    mutations=[],
+                    forward_path=[(0, ts.sequence_length, 1)],
+                    forward_mutations=[],
                 )
             )
 
@@ -722,13 +732,13 @@ def test_exact_matches(
         samples = sc2ts.preprocess(
             [fx_metadata_db[strain]], ts, "2020-02-20", fx_alignment_store
         )
-        # FIXME
-        mu = 0.125
+        mu, rho = sc2ts.solve_num_mismatches(num_mismatches)
         sc2ts.match_tsinfer(
             samples=samples,
             ts=ts,
-            num_mismatches=num_mismatches,
-            likelihood_threshold = mu**num_mismatches - 1e-12,
+            mu=mu,
+            rho=rho,
+            likelihood_threshold=mu**num_mismatches - 1e-12,
             num_threads=0,
         )
         s = samples[0]
@@ -757,12 +767,13 @@ def test_one_mismatch(
         samples = sc2ts.preprocess(
             [fx_metadata_db[strain]], ts, "2020-02-20", fx_alignment_store
         )
+        mu, rho = sc2ts.solve_num_mismatches(num_mismatches)
         sc2ts.match_tsinfer(
             samples=samples,
             ts=ts,
-            num_mismatches=num_mismatches,
-            # FIXME
-            likelihood_threshold=0.12499999,
+            mu=mu,
+            rho=rho,
+            likelihood_threshold=mu - 1e-5,
             num_threads=0,
         )
         s = samples[0]
@@ -785,11 +796,12 @@ def test_two_mismatches(
         samples = sc2ts.preprocess(
             [fx_metadata_db[strain]], ts, "2020-02-20", fx_alignment_store
         )
-        mu = 0.125
+        mu, rho = sc2ts.solve_num_mismatches(num_mismatches)
         sc2ts.match_tsinfer(
             samples=samples,
             ts=ts,
-            num_mismatches=num_mismatches,
+            mu=mu,
+            rho=rho,
             likelihood_threshold=mu**2 - 1e-12,
             num_threads=0,
         )