Closing out on a driver who is already moving

basketball.casa is a basketball league where every game is simulated. Nobody scripts the plays. Each one comes out of who the players are at that moment: how tall, fast and tired they are, what they can see, and what they decide. You can watch any game back exactly as it happened. Last time the update was about momentum. This one follows it onto the defender guarding the ball.

Speed that counts is speed toward the driver

A defender sprinting at a driver who is running away from them is not closing much. One running straight at a driver coming toward them closes very fast. Until now the game judged a drive by how fast the defender could run, as if the driver were standing still. Now it uses how fast the gap between them is shrinking: the defender's own speed and direction against the driver's, carried over from the dribble that came before. The same defender can be quick enough to cut off one drive and too slow for the next, depending on where both of them were already heading.

The defender starts from where they were

The defender guarding the ball now sets off at the speed they already had, the way every other defender has since the last update. And when a driver gets past them, the chase in the replay starts from the run the drive was judged on. Before, the game judged a drive against a defender who was moving, then replayed them chasing from a standstill. Those two pictures disagreed in about half of all drives. Now they are the same run, so the drive you watch is the drive that was decided.

Why it matters

Whether a drive gets by is a question about two bodies at once: how fast each is going, which way, and how quickly each can change it. Nothing was set to make drives succeed more or less often. In our test games drivers chose to attack about one time in eight more often than before, and an attacker and a defender end up inside each other less often. Pick any drive in a replay and watch the defender's first step: it carries on from where they were already going.