Alan Marty wanted his daughter to get more height on her basketball shot. He put a rake on a ladder so she could practice sending the ball over it. The arrangement gave her a target for the shape of the shot. It also gave Marty a problem to keep working on: how could a player tell whether a shot had the right arc?
“After every shooting session, you get scored as you would in a video game”
Two fathers compare notes
Marty and Ridge McGhee were friends who played pickup basketball together. McGhee had also built a homemade aid for his daughter. Once the two compared ideas, they could bring different skills to the same question. McGhee knew machine vision, and they brought in Tom Edwards, then a researcher at NASA Ames, to help turn the driveway experiment into something they could test.
A number after every shot
The early prototype used a camera to follow the ball and called out the angle after a shot. The collaborators tested it with local high schools and Stanford’s men’s team. Marty had begun with a physical obstacle his daughter could shoot over. The camera let the team give a player feedback on the flight of the ball during an otherwise ordinary practice session.
AI in the product · November 13, 2026
Marty’s driveway experiment asked whether a player could get useful feedback on every shot. Noah now sells installed systems that track ball flight and, in some configurations, automatically identify the shooter without a wearable. Its HOOPS app brings the company’s shot-tracking engine to an iPhone, giving individual players another way to track their shooting.
A Vanderbilt collaboration took the research in a different direction. Students tested a model that classified different kinds of shots from video. Vanderbilt described it as a research project with potential applications. The deployed products and the research prototype show two different ways to build on Marty’s original question.