NBA Game Video Labeler — Watch Games, Log Events

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. We help people start and grow careers teaching AI by connecting contributors to projects, offering clear onboarding, and supporting flexible, remote work.

Why AI training matters

AI training (also called data labeling or annotation) is the human side of building smarter systems. People prepare and review examples—like labeled video events—that modern AI models learn from, making this work a fast-growing way to join the tech ecosystem.

This kind of work is often remote, flexible, and accessible: many projects need no prior experience beyond attention to detail or domain familiarity. For this role, your knowledge of basketball will directly shape how models understand live sports events.

The role

You will watch NBA games (live or within a couple hours after they air) and log specific in-game events along with the quarter and game-clock time. This is entry-level, part-time contractor work designed for basketball fans who enjoy watching games and recording structured event data.

Workload is less than 20 hours per week. The project uses internal proprietary tooling and a simple template to enter each event; we provide the necessary paid streaming subscription so you can access the games required.

• Employment type: Contractor, Part-time
• Data type: Video (action recognition labeling)
• Tooling: Internal proprietary labeling tools
• Schedule: Live or near-live viewing (within a couple hours after air)

What you'll do

Log discrete in-game events while watching NBA games and record associated metadata using the provided template and tooling. Each labeled event captures who was involved, the event type, the quarter, and the clock time.

• Identify the player involved in an event (e.g., LeBron James).
• Classify the type of event (free throw, 3-pointer, rebound, turnover, etc.).
• Record the quarter (Q1–Q4, OT where applicable).
• Enter the game-clock time for the event (e.g., 2:30).
• Use the supplied template and internal tool to submit labels.

Pay and logistics

You are paid per labeled event at a rate of $0.20 per label. We provide a paid streaming subscription (League Pass, YouTube TV, or similar) so you can watch all required NBA games.

• Per-label rate: $0.20 per recorded event
• Streaming access provided by the project (no personal subscription required for assigned games)
• Further submission and payment details will be shared during onboarding

Requirements

Applicants should be familiar with the rules and common events of basketball and comfortable identifying players and plays in live game footage. This is an entry-level role but requires attention to detail and the ability to watch full games live or shortly after they air.

• Familiarity with basketball rules and common in-game events
• Ability to watch full NBA games live or within a couple hours after airing
• Comfort using a web-based labeling tool on a computer or phone with a stable internet connection
• Available for under 20 hours per week

Who should apply and how it works

This project is a good fit for basketball fans, sports statisticians, or anyone who enjoys watching games and wants flexible, part-time remote work. No prior labeling experience is required; we provide clear instructions, examples, and the streaming access you'll need.

To apply, you'll complete onboarding where we review the template, show example labels, and explain how to use the proprietary tool. Once approved, you'll receive assignments, access to the streaming subscription, and begin logging events for pay at $0.20 per label.

• Great for fans who want to earn while watching games
• Entry-level: clear instructions and examples provided
• Worldwide applicants accepted