PlayVision Is Doing for Basketball Scouting What Moneyball Did for Baseball, but With Computer Vision

The Macro: Basketball Analytics Still Runs on Eyeballs and Spreadsheets

Professional basketball has gotten very good at analytics. NBA teams employ entire departments of data scientists, have proprietary camera systems tracking every movement on the court, and build custom models that inform everything from lineup decisions to contract negotiations. The tracking infrastructure alone (Second Spectrum, now owned by the NBA itself, and the league-wide player tracking system) generates granular data that teams use constantly.

College basketball is a different story. Below the top 20 or 30 programs, most schools do not have the budget for a full analytics operation. They have a graduate assistant breaking down film manually. They have coaches rewatching games on Hudl or Synergy at 2 AM, pausing and rewinding to chart plays by hand. The transfer portal has made scouting more important than ever (over 2,000 players entered it this past season), and most programs evaluate portal candidates by watching highlight reels and calling other coaches.

This is a market gap that has been obvious for years. The technology to automatically track player movements on a basketball court using computer vision has existed in some form since at least 2015. But the products built on that technology have either been prohibitively expensive (targeting NBA and elite college programs) or too limited in scope (handling one piece of the puzzle, like shot tracking, but not the full analytical picture).

The demand side is clear. Over 350 Division I programs, hundreds more in D2 and D3, plus international leagues, prep schools, and AAU organizations all need better scouting and film analysis tools. ShotTracker sells hardware-based tracking for college programs but requires sensors on every player. Catapult focuses on wearable performance data. Synergy Sports provides curated play-by-play data but charges enough that smaller programs balk. The gap is in software-only, affordable, comprehensive film analysis that any program can use.

The Micro: Upload Film, Get a Million Data Points

Anish Gupta (CEO) and Marc Zoghby cofounded PlayVision and brought it through Y Combinator’s Fall 2025 batch. The product is straightforward in concept and ambitious in execution. You upload game film. PlayVision’s computer vision system processes it. You get back frame-by-frame player tracking, automatic play tagging, formation analysis, and over a million data points per game.

A million data points per game is a lot of data. To put that in context, that is roughly the same order of magnitude as what the NBA’s proprietary tracking system generates. The difference is that PlayVision derives this from standard broadcast or huddle camera video rather than dedicated multi-camera tracking rigs. If the quality of tracking holds up (and that is the key technical question), it means a mid-major program can get NBA-level positional data from the same game film they already have.

The transfer portal database is particularly interesting. PlayVision has built searchable profiles for over 2,400 college players. A coach looking for a specific type of player (say, a wing who shoots above 38 percent from three, defends in the top quartile of closeout speed, and creates at least two open looks per game for teammates) could theoretically query that database and get results with supporting data. That is something only the richest programs can do today, and they do it with expensive proprietary tools and dedicated staff.

Pricing is aggressive. The College Portal is $29.99 per month. The NBA Portal is $29.99 per month. The Pro Suite (both plus extras) is $49.99 per month. Team plans are custom. At thirty bucks a month, this is accessible to literally any college program, any AAU team, any international club. It is priced as a consumer product, not an enterprise product, and that is a deliberate choice to prioritize adoption over revenue per customer.

The competitors worth watching are Synergy Sports (the incumbent for play-by-play data), ShotTracker (hardware-based tracking), InStat (video analysis for soccer that has expanded into basketball), and Hudl (the dominant platform for film distribution that has been adding analytics features). None of them currently offer fully automated computer vision analysis at PlayVision’s price point. If the tracking quality is competitive, PlayVision undercuts every alternative by an order of magnitude on price.

The Verdict

I think this product has a real shot at changing how basketball scouting works at the college level and below. The price point removes the budget objection. The computer vision approach removes the hardware installation objection. The transfer portal database addresses the most urgent need in college basketball right now.

At 30 days, the question is tracking accuracy. Computer vision on basketball film is a hard problem because of occlusion (players overlap constantly), camera angle variation (broadcast vs huddle camera vs gym camera), and speed of play. If PlayVision’s tracking is accurate enough that coaches trust it for personnel decisions, the product spreads through word of mouth in the coaching community fast. At 60 days, I want to see how many programs are using the transfer portal database and whether it is influencing actual roster decisions. At 90 days, the big question is whether any conference or league does a bulk deal to give all member schools access, which would be the fastest path to broad adoption.

The “Moneyball for basketball” label has been applied to a dozen products and none have stuck. PlayVision has a better shot than most because it is not asking coaches to change their workflow. Coaches already watch film. PlayVision just makes the film tell them more. That is the kind of value proposition that actually survives contact with real users.