diff --git a/trumania/core/relationship.py b/trumania/core/relationship.py
index 6e6b943..78f8107 100644
--- a/trumania/core/relationship.py
+++ b/trumania/core/relationship.py
@@ -11,15 +11,15 @@
 
 # There are a lot of somewhat ugly optimizations here like in-place mutations,
 # caching, or usage of numpy instead of a more readable pandas alternative. The
-# reason is the methods of this filetend to be called a large amount of time
+# reason is the methods of this file tend to be called a large amount of time
 # in inner loop of the simulation, optimizing them make the whole simulation
 # faster.
 
 
-class Relations(object):
+class OutgoingRelations(object):
     """
-     This entity contains all the "to" sides of the relationships of a given
-     "from", together with the related weights.
+     For a given "from", this entity contains all the "to" sides of the
+     relationship with the related weights.
 
      This data structure seems to be the most optimal since it corresponds to a cached
      group-by result, and those group-by are expensive in the select_one
@@ -45,8 +45,10 @@ def from_tuples(from_ids, to_ids, weights):
          a relationship is built here for each "line" read across those 3
          arrays.
 
-         This methods builds one instance of Relations for each unique from_id
+         This methods builds one instance of OutgoingRelations for each unique from_id
          value, containing all the to_id's it is related to.
+
+         :returns Dictionary { id1 -> OutgoingRelations1, id2 -> OutgoingRelations2, ... }
         """
 
         from_ids = np.array(from_ids)
@@ -60,11 +62,12 @@ def from_tuples(from_ids, to_ids, weights):
         order = from_ids.argsort()
         ordered = zip(from_ids[order], to_ids[order], weights[order])
 
+        # Find for every unique id in from_ids their matching "to" relations
         def _relations():
             # itertools.groupby is much faster than pandas
             for from_id, tuples in itertools.groupby(ordered, lambda t: t[0]):
-                to_ids, weights = list(zip(*tuples))[1: 3]
-                yield from_id, Relations(list(to_ids), list(weights))
+                to_ids, weights = list(zip(*tuples))[1:3]
+                yield from_id, OutgoingRelations(list(to_ids), list(weights))
 
         return {from_id: relz for from_id, relz in _relations()}
 
@@ -72,7 +75,7 @@ def plus(self, other):
         """
         Merge function for 2 sets of relations all starting from the same "from"
         """
-        return Relations(
+        return OutgoingRelations(
             np.hstack([self.to_ids, other.to_ids]),
             np.hstack([self.weights, other.weights]))
 
@@ -83,7 +86,7 @@ def minus(self, other):
         """
         removed_indices = np.argwhere(
             [idx in other.to_ids for idx in self.to_ids])
-        return Relations(
+        return OutgoingRelations(
             np.delete(self.to_ids, removed_indices),
             np.delete(self.weights, removed_indices))
 
@@ -157,7 +160,7 @@ def add_relations(self, from_ids, to_ids, weights=1):
 
         self.grouped = utils.merge_2_dicts(
             self.grouped,
-            Relations.from_tuples(from_ids, to_ids, weights),
+            OutgoingRelations.from_tuples(from_ids, to_ids, weights),
             lambda r1, r2: r1.plus(r2))
 
     def add_grouped_relations(self, from_ids, grouped_ids):
@@ -185,7 +188,7 @@ def remove_relations(self, from_ids, to_ids):
 
         self.grouped = utils.merge_2_dicts(
             self.grouped,
-            Relations.from_tuples(from_ids, to_ids, weights=0),
+            OutgoingRelations.from_tuples(from_ids, to_ids, weights=0),
             lambda r1, r2: r1.minus(r2))
 
     def get_relations(self, from_ids=None):
diff --git a/trumania/core/util_functions.py b/trumania/core/util_functions.py
index f5495d5..727c8ef 100644
--- a/trumania/core/util_functions.py
+++ b/trumania/core/util_functions.py
@@ -89,23 +89,33 @@ def merge_2_dicts(dict1, dict2, value_merge_func=None):
     if dict1 is None:
         return dict2
 
-    def merged_value(key):
-        if key not in dict1:
-            return dict2[key]
-        elif key not in dict2:
-            return dict1[key]
-        else:
-            if value_merge_func is None:
-                raise ValueError(
-                    "Conflict in merged dictionaries: merge function not "
-                    "provided but key {} exists in both dictionaries".format(
-                        key))
-
-            return value_merge_func(dict1[key], dict2[key])
-
-    keys = set(dict1.keys()) | set(dict2.keys())
-
-    return {key: merged_value(key) for key in keys}
+    if dict1 == dict2:
+        for k, v in dict1.items():
+            dict1[k] = value_merge_func(v, v)
+        return dict1
+
+    dict1_set = set(dict1)
+    dict2_set = set(dict2)
+
+    keys_to_merge = dict1_set.intersection(dict2_set)
+
+    if len(keys_to_merge) != 0 and value_merge_func is None:
+        raise ValueError(
+            "Conflict in merged dictionaries: merge function not "
+            "provided but keys {} exists in both dictionaries".format(
+                keys_to_merge))
+
+    values_merged = dict()
+
+    for key_to_merge in keys_to_merge:
+        old_value1 = dict1[key_to_merge]
+        old_value2 = dict2[key_to_merge]
+
+        new_value = value_merge_func(old_value1, old_value2)
+
+        values_merged[key_to_merge] = new_value
+
+    return {**dict1, **dict2, **values_merged}
 
 
 def df_concat(d1, d2):