DataProjectsStudents

Leveraging Big Sports Events to Teach Data Analysis: A Student Project Using JioHotstar Metrics

UUnknown

2026-02-10

9 min read

Turn JioHotstar’s Women’s World Cup streaming surge into a hands-on data analysis project—students clean, visualize, and present viewership insights.

Turn record-breaking streaming into a classroom lab: teach data analysis with JioHotstar Women’s World Cup metrics

Students and teachers: tired of abstract assignments that never stick? You want a project that teaches real-world data skills, stays engaging for learners, and produces portfolio-ready work. Use the Women’s World Cup streaming surge on JioHotstar as your vehicle. In late 2025 JioHotstar reported record engagement—platform-wide numbers like 99 million digital viewers for the final and an average user base exceeding 450 million monthly users—making it an ideal, culturally relevant dataset for a semester-long data analysis project (Variety, Jan 2026).

Why this project matters in 2026

In 2026, educators need coursework that mirrors industry practice: messy data, rapid iteration, and storytelling. Sports streaming platforms now publish richer engagement signals and are central to media-business decisions. Students who can clean, model, and visualize streaming viewership data are prepared for roles in analytics, product, marketing, and research.

“JioHotstar hit its highest engagement for the Women’s World Cup, with record digital viewers—an ideal case for sports analytics and student projects.” — Variety, Jan 2026

Project overview: objectives and outcomes

Design a modular, scaffolded project where students progress from data cleaning to interactive dashboards and predictive models. At the end they should deliver a reproducible notebook, a public-facing dashboard, and a concise slide deck with insights and recommendations.

Duration: 6–10 weeks (modular for workshops or semester)
Skill focus: data cleaning, exploratory data analysis (EDA), visualization, basic statistical testing, time-series forecasting, and storytelling
Tools: Python (pandas, seaborn, plotly), Google Colab, Tableau/Power BI, and GitHub for version control; low-code alternatives for mixed-ability cohorts
Deliverables: cleaned dataset + README, EDA notebook, interactive dashboard, slide deck, graded rubric

Datasets and ethical sourcing

Real platform data is often proprietary. Use a mix of publicly released metrics (press releases, publisher reports), platform-level aggregates cited by industry outlets, and synthetic or scrubbed CSVs that mirror the published aggregates. Always respect platform terms of service and privacy laws—India’s regulatory landscape and international standards have tightened since 2023, so prioritize anonymized and aggregate data.

Where to get data for the classroom

Public reports and press releases: use platform statements (e.g., JioHotstar quarterly report) to set target aggregates. See also best practices for turning press mentions into usable classroom targets.
Social listening APIs: Twitter/X, Instagram public metrics to correlate social buzz with viewership spikes.
Synthetic or scrubbed CSVs: create row-level viewers.csv with columns described below to emulate real telemetry while avoiding PII.
Open sports datasets: match schedules, team metadata, and game events to join with streaming metrics.

Recommended synthetic dataset schema (students build from this)

timestamp — ISO8601 start time of viewing session
match_id — unique match identifier
viewer_id — anonymized session id (not PII)
device — mobile, desktop, smartTV
country — country code for geo-analysis
watch_time_seconds — session length
concurrent_viewers — snapshot of concurrent viewers at minute-level
engagement_event — play, pause, seek, ad_view

Week-by-week syllabus (6–10 week template)

Structure guarantees progressive complexity with checkpoints for instructors to grade and give feedback.

Weeks 1–2: Setup and Data Cleaning

Introduce objectives and platform context (JioHotstar and the Women’s World Cup numbers).
Provide synthetic dataset and a “data contract” describing each column.
Teach data validation: types, missingness, outliers, timestamp parsing, timezone normalization.
Deliverable: cleaned CSV and short report describing cleaning decisions and assumptions.

Weeks 3–4: Exploratory Data Analysis (EDA) and Hypothesis Formation

Perform high-level aggregations: daily unique viewers, average watch time, peak concurrency, device split.
Visualization exercises: line charts for time-series, bar charts for device mix, heatmaps for retention.
Form 3 testable hypotheses (e.g., “Matches with more social buzz have higher peak concurrency within 30 minutes”).
Deliverable: EDA notebook and hypothesis log with visual evidence.

Weeks 5–6: Advanced Analysis

Teach cohort and retention analysis: retention curves, funnel analysis for ‘start to finish’ completion rate.
Introduce time-series forecasting: Prophet or ARIMA models to predict next-match peak concurrency.
Clustering viewers by engagement patterns and device type (K-means or hierarchical clustering).
Deliverable: advanced analysis notebook with reproducible code and commentary.

Weeks 7–8: Visualization & Dashboarding

Build interactive dashboards using Plotly Dash, Tableau Public, or Google Data Studio.
Include drill-downs (match → minute-level), filter by country and device, and a ‘what-if’ scenario panel.
Deliverable: published dashboard URL + README for interaction guide.

Weeks 9–10: Presentation and Business Recommendations

Students prepare a 10-minute presentation and a one-page executive summary with 3 prioritized recommendations.
Business framing: how would JioHotstar use these insights to optimize ad inventory, retention campaigns, or streaming infrastructure?
Deliverable: slide deck, executive brief, and final code repository.

Sample analysis questions and hypotheses

These guide students toward metrics that matter to streaming platforms and advertisers.

What factors predict peak concurrent viewers for a match? (match importance, teams, time-of-day, social buzz)
How does device distribution change across regions, and how does that affect average watch time?
Where are the largest drop-offs in watch sessions (start, mid, or post-match)? Build a retention funnel.
Can we forecast next-match peak concurrency within an acceptable error margin using historical data?
Is higher social media engagement associated with longer viewing sessions or only higher reach?

Tools, templates, and reproducibility

Provide students with starter notebooks and templates to reduce setup friction. Encourage reproducibility and version control.

Recommended starter stack

Google Colab or Jupyter notebooks for Python (pandas, numpy, matplotlib, seaborn, plotly, prophet)
GitHub Classroom for submissions and collaborative work
Tableau Public or Power BI for dashboarding and non-code students
Google Sheets for quick pivoting and initial hypotheses

Starter templates to include

Cleaning checklist: column types, null policy, timezone policy
EDA notebook with example aggregations and visualizations
Dashboard wireframe (PNG) and a visualization style guide
Rubric: code quality, data story, reproducibility, and business insight

Assessment rubric (example)

Use a clear rubric so students know what professional skills they are building.

Data hygiene (20%): documented cleaning, handling of missing values, clear schema
Analysis rigor (25%): correct use of methods, testable hypotheses, statistical soundness
Visualization & storytelling (25%): clear charts, dashboard usability, recommendation clarity
Reproducibility (15%): runnable notebook, dependency list, version control
Business impact (15%): actionable recommendations, prioritized and feasible

Advanced strategies for high-performing students

For students ready to go further, include modules on machine learning, causal inference, and multimodal analysis.

Predictive modeling: build features like rolling averages, social-signal lags, and weather or holiday flags to forecast peak concurrency.
Causal impact: use interrupted time-series or CausalImpact libraries to estimate the effect of a marketing campaign on viewership.
Survival analysis: model session duration as time-to-event (drop-off) and identify covariates that prolong sessions.
Multimodal fusion: combine match telemetry with social sentiment analysis (NLP) to explain sudden viewership spikes.
Edge and streaming analytics: discuss how platforms use real-time telemetry to allocate CDN resources and optimize QoE (quality of experience).

Linking to industry trends in 2026

Late 2025 and early 2026 saw streaming platforms and broadcasters invest heavily in sports rights, analytics, and viewer engagement. JioHotstar’s record numbers highlighted how major events can shift audience behaviour—and how analytics teams translate that into revenue opportunities. As a teacher, connecting classroom work to these trends shows students the career relevance of their work.

Ethics, privacy, and legal considerations

Teach students that powerful analytics come with responsibilities.

Avoid PII: use anonymized session ids and avoid reconstructing identities.
Consent and TOS: scraping or using APIs must follow platform terms; teach students to verify.
Regulatory context: discuss India’s data protection trajectory and global frameworks like GDPR and how they affect telemetry handling.
Bias: identify potential sampling biases (e.g., mobile-only viewers in certain regions) and explain how they affect findings.

Classroom case study: an example outcome

One class used a simulated JioHotstar dataset and public match metadata. Their final dashboard revealed three high-impact findings:

Peak concurrency correlated most strongly with match importance and social mention volume 2–4 hours prior to kick-off—suggesting marketing windows for last-minute push notifications.
Mobile viewers in under-indexed regions showed higher completion rates than expected, indicating an opportunity to expand ad formats there.
A small but consistent mid-match drop-off matched ad breaks; teams recommended testing shorter, targeted ad pods to reduce churn.

These findings were presented to local industry mentors, and several students later referenced the project in internships—an example of experience-based learning translating to career opportunities.

Common instructor FAQs

Q: What if I can’t access real telemetry?

A: Use the synthetic schema above and scale the volume to emulate production telemetry. Seed the dataset with event patterns pulled from press-release aggregates so classroom results are realistic.

Q: How do I accommodate mixed-ability students?

A: Offer a no-code track (Tableau/Sheets) and a code track (Python). Provide extension tasks for advanced students and plug-and-play notebooks for beginners.

Q: How to grade subjectively creative outputs like visuals?

A: Use rubrics emphasizing clarity, accuracy, and insight. Provide exemplars and run a peer-review session to teach critique skills.

Final checklist for instructors

Prepare starter dataset and cleaning checklist
Create exemplar notebooks and dashboards
Align grading rubric with real-world deliverables
Schedule interim checkpoints for formative feedback
Ensure ethical data use and privacy training

Why this project prepares students for tomorrow

Sports streaming analytics sits at the intersection of data science, product, and business strategy. By teaching students to extract insights from viewership and engagement metrics—especially using a culturally resonant event like the Women’s World Cup and JioHotstar’s 2025–26 surge—you give them a portfolio that employers recognize. They learn technical skills, critical thinking, and the soft skills needed to communicate findings that drive decisions.

Next steps and call-to-action

If you’re ready to run this project, start with a lightweight pilot: a one-week mini-hackathon using the synthetic schema above, then iterate to the full semester version. Download our free project pack (starter CSVs, Colab notebooks, and grading rubric) or sign up for thepower.info’s workshop for educators. Equip your students with a real-world sports analytics case study that turns curiosity into career-ready capability.

Ready to teach a project that matters? Get the starter pack, sample syllabus, and student rubrics at thepower.info/projects and run your first session in under a week.

Unknown

Contributor

Senior editor and content strategist. Writing about technology, design, and the future of digital media. Follow along for deep dives into the industry's moving parts.