Closed bttmly closed 8 years ago
quoting from http://savvastjortjoglou.com/nba-play-by-play-movements.html
[3,
1431486313010,
715.32,
19.0,
None,
[[-1, -1, 43.51745, 10.76997, 1.11823],
[1610612745, 1891, 43.21625, 12.9461, 0.0],
[1610612745, 2772, 90.84496, 7.79534, 0.0],
[1610612745, 2730, 77.19964, 34.36718, 0.0],
[1610612745, 2746, 46.24382, 21.14748, 0.0],
[1610612745, 201935, 81.0992, 48.10742, 0.0],
[1610612746, 2440, 88.12605, 11.23036, 0.0],
[1610612746, 200755, 84.41011, 43.47075, 0.0],
[1610612746, 101108, 46.18569, 16.49072, 0.0],
[1610612746, 201599, 78.64683, 31.87798, 0.0],
[1610612746, 201933, 65.89714, 25.57281, 0.0]]]
@brandly beautiful, thanks! Next up I'll be working on methods that will identify plays where a given player touches the ball, passes, shoots, turns it over, etc.
Any help or suggestions on API design or desired functionality much encouraged
I split this out into it's own package https://github.com/nickb1080/nba-movement
There is a JSON endpoint serving the player movement data that gets visualized in the stats.nba.com play-by-play area. To see what I'm talking about go here, then click on one of the plays and select "movement" (not "video").
I've done some initial work (on the movement branch) to pull the data for an entire game from the NBA's api, but as usual, the data is in unlabeled tuples to reduce data size, making it unclear what the data actually means. There is some example JSON in the /src/apis/movement directory, any help understanding this scheme would be much appreciated!