A LEARNING GAME ABOUT RECOMMENDATION SYSTEMS

Mr. Pelling’s
Movie Club

Set in a 1990s video store, the game introduces recommendation systems through customer interviews, rules, ratings, and predictive models.

Desktop browser · Logging optional

THE GAME ENVIRONMENTPELLINGS VIDEO

ACTIVITY SEQUENCE

Gameplay and
learning objectives

The eight activities and their intended learning focus.

Activity	What the player does	Intended learning focus
1. Manual recommendation	Ask limited questions and choose a tape for a customer.	Preference elicitation, context, and the limits of manual service.
2. Rule-based recommendation	Construct genre-to-movie rules and inspect batch outcomes.	Automation, brittle rules, and age-appropriate recommendations.
3. Ratings tables	Inspect individual rating rows and compare column averages.	Representing feedback and distinguishing individual taste from aggregate ratings.
4. Collaborative filtering	Compare customers and predict missing ratings.	Similarity-based inference and sparse observations.
5. Matrix factorization	Adjust latent-factor sliders, then run an optimizer.	Hidden factors, prediction error, optimization, privacy, and popularity bias.
6. Content-based recommendation	Compare customer preferences with movie features.	Feature representations, explanations, and the limits of descriptive labels.
7. Market-gap research	Combine setting, character, story, and prop choices into a movie brief.	Using audience evidence to support human creative decisions.
8. Poster generation	Revise a prompt, choose a style, and compare generated drafts when the image service is connected.	Iterative prompting and human evaluation of generated media.

BROWSER VERSION

Build updated September 6, 2026

Use a desktop browser with a mouse or trackpad. The initial download may take a moment.

Select Play without logging to access the activities without providing a name or consenting to gameplay logging.

Live AI poster generation is not connected in this version.