joaquinpinto / movieclub

Analysis of Movie Ratings For Movie Club

MovieClub

Analysis of Movie Ratings For Movie Club

This is a project I worked on using data I received from my friends.

What's In Here:

• MovieRatings.csv - Movie rating data from friends, Header = Movie, and friends names.
• AdjustedMovieAverage.csv - Standardized scores and Normalized average scores for each movie.
• MCRatingAdj.inpyb - Jupyter Notebook for creating Adjusted Score sheet from Movie Rating sheet.
• MovieClubRating.py - Python Code for comparing each person's ratings, results below.
• MovieRankWeight.py - Python Code for reading in the Adjusted Scores and creating randomized order.

Who Are My Friends:

We are a group of 5 friends who met in High School (and 1 person who made his way into the group during college) who are all thoroughly interested in cinema. During High School one of our main weekend hang-outs was at the matinee showing of new films at AMC. Together we endulged ourselves in classics such as The Hunger Games and The Purge for the low price of $5.

Movie Club:

After college a few of us found ourselves outside of the city. We thought it would be a great idea to continue our tradition of watching films during the weekend with a bi-weekly, long-distance movie club. We were to watch movies every other Friday and comment on them the following Sunday.

How Do We Select These Movies:

This idea behind this club changed a bit, we were originally going to watch new films, however the cost of watching these would've been too high for our budget, so we decided to go through old movies first. We began by each offering up a set of movies that we were interested in watching, some of us already had lists, others just looked through their recomended movies lists on Netflix.

I then compiled those movies onto a Google Sheet and had each member rate their willingness to watch the movies with these guidelines:

1. The score must be from 1-10.
2. Be reasonable with movies you've watched already, if we've all seen them there's no use in watching them again.
3. You may only use the 1 and the 10 a maximum of one time in your ratings (I'll mention why shortly).

How Do We Make This Fun:

We didn't, it was very tedious.

However, I added a little game to it.

Each person has 2 Power-Ups, the 1-Card and the 10-Card.

• 1-Card: This means you really don't want to watch this film, it functions as an extra -5 to the movie's score.
• 10-Card: This means that the film is your top choice, it functions as an extra +5 to the movie's score.

The Google Sheet would mark that person't column red if they had 2 or more 1 or 10 ratings.

That's Cool and All, But Where's The Analysis:

I'm getting there, give me a second!

After getting the ratings from my friends I sought to standardize each person's ratings, so that we all had the same say in the movie selection.

This was performed by finding the Z-Scores of each rating for each Rater:

• Z-Score = (Score - Rater's AVG Score)/Rater's SD.

I then took the average Z-score for each movie and normalized it back to a 1-10 scale with 5 being the mid-point:

• 5 + (Avg Z-Score * 10)/(Max AVG Z - Min AVG Z)

Afterwards I added/subtracted the power ups and rounded to two-digits:

• Blazing Saddle Adjusted Score = 5.04 + 5/6 = 5.87

This was all completed directly on the Google Sheet so as to be as open as possible, I have recreated this in a Jupyter Notebook.

Movie Order:

Once I found out how each movie was rated I ranked them and removed all the movies with scores below 5.

Since the plan was to watch all the movies, but we weren't sure how long the club would last, I recomended that we make a weighted average ranking. This would allow us to vary the movies, so that we don't watch all the good ones first and then get bored near the end. We gave preference to the top movies, but the ultimate ranking was random.

The code for the weighted ranking is shown here as MovieRankWeight.py.

I ran this 3 times and we chose our favorite from there.

This was our final order after the randomization:

Week 1: District 9
Week 2: Goodfellas
Week 3: Coco
Week 4: City of God
Week 5: Moonrise Kingdom
Week 6: Blazing Saddles
Week 7: Usual Suspects
Week 8: Hoop Dreams
Week 9: Dazed and Confused
Week 10: Shaun of the Dead
Week 11: Gattaca
Week 12: Scott Pilgrim
Week 13: Escape from New York
Week 14: Old School
Week 15: Set it Up
Week 16: 12 Monkeys
Week 17: Annihilation
Week 18: The Fight Club
Week 19: Incredibles 2
Week 20: Halloween
Week 21: Warriors
Week 22: Scream
Week 23: Hot Fuzz
Week 24: The Jerk
Week 25: Forrest Gump
Week 26: Airplane!
Week 27: Basic Instinct
Week 28: Reservoir Dogs
Week 29: Crazy Rich Asians
Week 30: Saw
Week 31: Matrix
Week 32: Blade Runner 2049
Week 33: In Bruges
Week 34: Shawshank Redemption
Week 35: Zodiac
Week 36: Ghost in the Shell (1995)
Week 37: Accident Man
Week 38: Return of the Dragon
Week 39: Gangs Of New York
Week 40: Taxi Driver
Week 41: Bill and Ted
Week 42: Easy A
Week 43: Apocalypse Now
Week 44: American Gangster
Week 45: Donnie Darko
Week 46: Bourne
Week 47: The Naked Gun
Week 48: Dredd
Week 49: Cool Hand Luke

Lastly, I also ran through the original scores to find the closest raters and how far each person strayed from the group.

The code for this is in MovieClubRating.py.

These are the results:

Closest Raters

1. Givnos and Mohinos with an average distance of 1.89 per movie
2. Joaquinos and Marquinos with an average distance of 1.98 per movie
3. Marquinos and Mohinos with an average distance of 2.04 per movie
4. Givnos and Joaquinos with an average distance of 2.05 per movie
5. Givnos and Marquinos with an average distance of 2.05 per movie
6. Givnos and Jerminos with an average distance of 2.16 per movie
7. Jerminos and Marquinos with an average distance of 2.23 per movie
8. Joaquinos and Mohinos with an average distance of 2.26 per movie
9. Jerminos and Joaquinos with an average distance of 2.28 per movie
10. Jerminos and Mohinos with an average distance of 2.35 per movie
11. Givnos and Jaminos with an average distance of 2.36 per movie
12. Jaminos and Joaquinos with an average distance of 2.48 per movie
13. Jaminos and Marquinos with an average distance of 2.51 per movie
14. Jaminos and Mohinos with an average distance of 2.59 per movie
15. Jaminos and Jerminos with an average distance of 2.91 per movie

Average Group Distance

1. Givnos with an average distance of 2.1 from the group
2. Marquinos with an average distance of 2.16 from the group
3. Joaquinos with an average distance of 2.21 from the group
4. Mohinos with an average distance of 2.22 from the group
5. Jerminos with an average distance of 2.39 from the group
6. Jaminos with an average distance of 2.57 from the group