diff --git a/addons/project/tests/test_project_flow.py b/addons/project/tests/test_project_flow.py
index f4f539d3ec4..0b88260ac1f 100644
--- a/addons/project/tests/test_project_flow.py
+++ b/addons/project/tests/test_project_flow.py
@@ -248,7 +248,9 @@ class TestProjectFlow(TestProjectCommon, MockEmail):
         self.assertEqual(rating_good.parent_res_id, self.project_pigs.id)
 
         self.assertEqual(self.project_goats.rating_percentage_satisfaction, -1)
+        self.assertEqual(self.project_goats.rating_avg, 0, 'Since there is no rating in this project, the Average Rating should be equal to 0.')
         self.assertEqual(self.project_pigs.rating_percentage_satisfaction, 0)  # There is a rating but not a "great" on, just an "okay".
+        self.assertEqual(self.project_pigs.rating_avg, rating_bad.rating, 'Since there is only one rating the Average Rating should be equal to the rating value of this one.')
 
         # Consuming rating_good
         first_task.rating_apply(5, rating_good.access_token)
@@ -257,10 +259,14 @@ class TestProjectFlow(TestProjectCommon, MockEmail):
         # Our One2Many is linked to a res_id (int) for which the orm doesn't create an inverse
         first_task.invalidate_cache()
 
+        rating_avg = (rating_good.rating + rating_bad.rating) / 2
         self.assertEqual(first_task.rating_count, 2, "Task should have two ratings associated with it")
+        self.assertEqual(first_task.rating_avg_text, 'top')
         self.assertEqual(rating_good.parent_res_id, self.project_pigs.id)
         self.assertEqual(self.project_goats.rating_percentage_satisfaction, -1)
         self.assertEqual(self.project_pigs.rating_percentage_satisfaction, 50)
+        self.assertEqual(self.project_pigs.rating_avg, rating_avg)
+        self.assertEqual(self.project_pigs.rating_avg_percentage, rating_avg / 5)
 
         # We change the task from project_pigs to project_goats, ratings should be associated with the new project
         first_task.project_id = self.project_goats.id
@@ -271,7 +277,9 @@ class TestProjectFlow(TestProjectCommon, MockEmail):
 
         self.assertEqual(rating_good.parent_res_id, self.project_goats.id)
         self.assertEqual(self.project_goats.rating_percentage_satisfaction, 50)
+        self.assertEqual(self.project_goats.rating_avg, rating_avg)
         self.assertEqual(self.project_pigs.rating_percentage_satisfaction, -1)
+        self.assertEqual(self.project_pigs.rating_avg, 0)
 
     def test_task_with_no_project(self):
         """
diff --git a/addons/project/views/project_views.xml b/addons/project/views/project_views.xml
index 9e2261a573c..0bb19d55a8a 100644
--- a/addons/project/views/project_views.xml
+++ b/addons/project/views/project_views.xml
@@ -33,7 +33,10 @@
                     <separator/>
                     <filter string="Open Tasks" name="open_tasks" domain="[('is_closed', '=', False)]"/>
                     <separator/>
-                    <filter string="Rated Tasks" name="rating_task" domain="[('rating_last_value', '!=', 0.0)]" groups="project.group_project_rating"/>
+                    <filter name="rating_satisfied" string="Satisfied" domain="[('rating_avg', '&gt;=', 3.66)]" groups="project.group_project_rating"/>
+                    <filter name="rating_okay" string="Okay" domain="[('rating_avg', '&lt;', 3.66), ('rating_avg', '&gt;=', 2.33)]" groups="project.group_project_rating"/>
+                    <filter name="dissatisfied" string="Dissatisfied" domain="[('rating_avg', '&lt;', 2.33), ('rating_last_value', '!=', 0)]" groups="project.group_project_rating"/>
+                    <filter name="no_rating" string="No Rating" domain="[('rating_last_value', '=', 0)]" groups="project.group_project_rating"/>
                     <separator/>
                     <filter string="Unread Messages" name="message_needaction" domain="[('message_needaction', '=', True)]"/>
                     <separator/>
@@ -1121,8 +1124,8 @@
                     <field name="legend_done"/>
                     <field name="activity_ids"/>
                     <field name="activity_state"/>
-                    <field name="rating_last_value"/>
-                    <field name="rating_ids"/>
+                    <field name="rating_count"/>
+                    <field name="rating_avg"/>
                     <field name="allow_subtasks"/>
                     <field name="child_text"/>
                     <field name="is_private"/>
@@ -1181,10 +1184,10 @@
                                     <div class="oe_kanban_bottom_left">
                                         <field name="priority" widget="priority"/>
                                         <field name="activity_ids" widget="kanban_activity"/>
-                                        <b t-if="record.rating_ids.raw_value.length">
-                                            <span style="font-weight:bold;" class="fa fa-fw mt4 fa-smile-o text-success" t-if="record.rating_last_value.value == 5" title="Latest Rating: Satisfied" role="img" aria-label="Happy face"/>
-                                            <span style="font-weight:bold;" class="fa fa-fw mt4 fa-meh-o text-warning" t-if="record.rating_last_value.value == 3" title="Latest Rating: Okay" role="img" aria-label="Neutral face"/>
-                                            <span style="font-weight:bold;" class="fa fa-fw mt4 fa-frown-o text-danger" t-if="record.rating_last_value.value == 1" title="Latest Rating: Dissatisfied" role="img" aria-label="Sad face"/>
+                                        <b t-if="record.rating_count.raw_value &gt; 0" groups="project.group_project_rating">
+                                            <span style="font-weight:bold;" class="fa fa-fw mt4 fa-smile-o text-success" t-if="record.rating_avg.raw_value &gt;= 3.66" title="Average Rating: Satisfied" role="img" aria-label="Happy face"/>
+                                            <span style="font-weight:bold;" class="fa fa-fw mt4 fa-meh-o text-warning" t-elif="record.rating_avg.raw_value &gt;= 2.33" title="Average Rating: Okay" role="img" aria-label="Neutral face"/>
+                                            <span style="font-weight:bold;" class="fa fa-fw mt4 fa-frown-o text-danger" t-else="" title="Average Rating: Dissatisfied" role="img" aria-label="Sad face"/>
                                         </b>
                                     </div>
                                     <div class="oe_kanban_bottom_right" t-if="!selection_mode">
diff --git a/addons/rating/models/rating_mixin.py b/addons/rating/models/rating_mixin.py
index 46b9f43e5f5..ff5f55d1583 100644
--- a/addons/rating/models/rating_mixin.py
+++ b/addons/rating/models/rating_mixin.py
@@ -1,10 +1,21 @@
 # -*- coding: utf-8 -*-
 # Part of Odoo. See LICENSE file for full copyright and licensing details.
+import operator
 from datetime import timedelta
 
 from odoo import api, fields, models, tools
 from odoo.addons.rating.models.rating import RATING_LIMIT_SATISFIED, RATING_LIMIT_OK, RATING_LIMIT_MIN
 from odoo.osv import expression
+from odoo.tools.float_utils import float_compare
+
+OPERATOR_MAPPING = {
+    '=': operator.eq,
+    '!=': operator.ne,
+    '<': operator.lt,
+    '<=': operator.le,
+    '>': operator.gt,
+    '>=': operator.ge,
+}
 
 
 class RatingParentMixin(models.AbstractModel):
@@ -59,7 +70,8 @@ class RatingMixin(models.AbstractModel):
     rating_last_feedback = fields.Text('Rating Last Feedback', groups='base.group_user', related='rating_ids.feedback')
     rating_last_image = fields.Binary('Rating Last Image', groups='base.group_user', related='rating_ids.rating_image')
     rating_count = fields.Integer('Rating count', compute="_compute_rating_stats", compute_sudo=True)
-    rating_avg = fields.Float("Rating Average", compute='_compute_rating_stats', compute_sudo=True)
+    rating_avg = fields.Float("Average Rating",
+        compute='_compute_rating_stats', compute_sudo=True, search='_search_rating_avg')
     rating_percentage_satisfaction = fields.Float("Rating Satisfaction", compute='_compute_rating_satisfaction', compute_sudo=True)
     rating_last_text = fields.Selection(string="Rating Text", groups='base.group_user', related="rating_ids.rating_text")
 
@@ -79,6 +91,19 @@ class RatingMixin(models.AbstractModel):
             record.rating_count = mapping.get(record.id, {}).get('rating_count', 0)
             record.rating_avg = mapping.get(record.id, {}).get('rating_avg', 0)
 
+    def _search_rating_avg(self, operator, value):
+        if operator not in OPERATOR_MAPPING:
+            raise NotImplementedError('This operator %s is not supported in this search method.' % operator)
+        rating_read_group = self.env['rating.rating'].sudo().read_group(
+            [('res_model', '=', self._name), ('consumed', '=', True), ('rating', '>=', RATING_LIMIT_MIN)],
+            ['res_id', 'rating_avg:avg(rating)'], ['res_id'])
+        res_ids = [
+            res['res_id']
+            for res in rating_read_group
+            if OPERATOR_MAPPING[operator](float_compare(res['rating_avg'], value, 2), 0)
+        ]
+        return [('id', 'in', res_ids)]
+
     @api.depends('rating_ids.res_id', 'rating_ids.rating')
     def _compute_rating_satisfaction(self):
         """ Compute the rating satisfaction percentage, this is done separately from rating_count and rating_avg