The world is being quietly rearranged by people who write very long documents.


The title they went with Decoding Defensive Coverage Responsibilities in American Football Using Factorized Attention Based Transformer Models Noisy translates that to

AI now predicts individual defensive assignments in real NFL games


Researchers built a machine learning model that watches NFL play-by-play video and predicts what defensive coverage scheme each player is running and which receiver they're supposed to cover — reaching 89% accuracy. This moves defensive analysis from post-game classification of what happened to real-time prediction of individual player assignments as plays unfold.
For the first time, coaches and broadcasters can see frame-by-frame what each defender is supposed to be doing during a play, not just categorize the overall scheme after the fact — which changes how teams can study opponent strategy and evaluate whether individual players executed their assignments correctly.

If you insist
Read the original →