Interactive formation analysis in sports utilizing semi-supervised methods

1. A method, comprising:

identifying, by a computing system, player tracking data and event data corresponding to a match that includes a first team and a second team, wherein the player tracking data includes one or more coordinate positions of each player during the match, and wherein the event data defines one or more events that occur during the match;

dividing, by the computing system, the player tracking data into a plurality of segments based on the event data;

for each segment of the plurality of segments, learning, by the computing system, a first formation associated with a respective team in possession;

for each segment of the plurality of segments, learning, by the computing system, a second formation associated with a respective team not in possession;

mapping, by the computing system, the first formation to a first semantic label;

mapping, by the computing system, the second formation to a second semantic label;

detecting, by the computing system, a formation change based on the mapped first formation or the mapped second formation;

outputting, by the computing system, a flag corresponding to the detected formation change.

2. The method of claim 1 , further comprising:

receiving, by the computing system from a user device, a request to view a formation of a target team across a target game; and

based on the request, generating, by the computing system, a graphical output that visually indicates a change in the formation of the target team across the target game.

3. The method of claim 1 , further comprising:

receiving, by the computing system from a user device, a request to view a formation of a target team across a season;

identifying, by the computing system, a plurality of target games associated with the season; and

based on the request, generating, by the computing system, a graphical output that visually indicates one or more changes in the formation of the target team across the season.

4. The method of claim 1 , further comprising:

receiving, by the computing system from a user device, a request to view a response of a target team to a trajectory of a ball path; and

based on the request, generating, by the computing system, a graphical output that visually indicates a formation of the target team responsive to the trajectory of the ball path.

5. The method of claim 4 , further comprising:

receiving, by the computing system from the user device, a second request to view a second response of a second target team to the trajectory of the ball path; and

based on the request, generating, by the computing system, a second graphical output that visually indicates a second target formation of the second target team responsive to the trajectory of the ball path and compares the second target formation of the second target team to the formation of the target team.

6. The method of claim 1 , wherein learning, by the computing system, the first formation associated with the respective team in possession comprises:

generating a vector representation of the player tracking data, wherein the vector representation comprises a total number of frames, a total number of players, and a dimensionality of the player tracking data.

7. The method of claim 1 , further comprising:

initializing, by the computing system, one or more cluster centers of the player tracking data with one or more average player positions.

8. A non-transitory computer readable medium comprising one or more sequences of instructions, which, when executed by one or more processors, causes a computing system to perform operations comprising:

identifying, by a computing system, player tracking data and event data corresponding to a match that includes a first team and a second team, wherein the player tracking data includes one or more coordinate positions of each player during the match, and wherein the event data defines one or more events that occur during the match;

dividing, by the computing system, the player tracking data into a plurality of segments based on the event data;

for each segment of the plurality of segments, learning, by the computing system, a first formation associated with a respective team in possession;

for each segment of the plurality of segments, learning, by the computing system, a second formation associated with a respective team not in possession;

mapping, by the computing system, the first formation to a first semantic label;

mapping, by the computing system, the second formation to a second semantic label;

detecting, by the computing system, a formation change based on the mapped first formation or the mapped second formation;

outputting, by the computing system, a flag corresponding to the detected formation change.

9. The non-transitory computer readable medium of claim 8 , further comprising:

receiving, by the computing system from a user device, a request to view a formation of a target team across a target game; and

based on the request, generating, by the computing system, a graphical output that visually indicates a change in the formation of the target team across the target game.

10. The non-transitory computer readable medium of claim 8 , further comprising:

receiving, by the computing system from a user device, a request to view a formation of a target team across a season;

identifying, by the computing system, a plurality of target games associated with the season; and

based on the request, generating, by the computing system, a graphical output that visually indicates one or more changes in the formation of the target team across the season.

11. The non-transitory computer readable medium of claim 8 , further comprising:

receiving, by the computing system from a user device, a request to view a response of a target team to a trajectory of a ball path; and

based on the request, generating, by the computing system, a graphical output that visually indicates a formation of the target team responsive to the trajectory of the ball path.

12. The non-transitory computer readable medium of claim 11 , further comprising:

receiving, by the computing system from the user device, a second request to view a response of a second target team to the trajectory of the ball path; and

based on the request, generating, by the computing system, a second graphical output that visually indicates a second target formation of the second target team responsive to the trajectory of the ball path and compares the second target formation of the second target team to the formation of the target team.

13. The non-transitory computer readable medium of claim 8 , wherein learning, by the computing system, the first formation associated with the respective team in possession comprises:

generating a vector representation of the player tracking data, wherein the vector representation comprises a total number of frames, a total number of players, and a dimensionality of the player tracking data.

14. The non-transitory computer readable medium of claim 8 , further comprising:

initializing, by the computing system, one or more cluster centers of the player tracking data with one or more average player positions.

15. A system, comprising:

a processor; and

a memory having programming instructions stored thereon, which, when executed by the processor, causes the system to perform operations comprising:

identifying player tracking data and event data corresponding to a match that includes a first team and a second team, wherein the player tracking data includes one or more coordinate positions of each player during the match, and wherein the event data defines one or more events that occur during the match;

dividing the player tracking data into a plurality of segments based on the event data;

for each segment of the plurality of segments, learning a first formation associated with a respective team in possession;

for each segment of the plurality of segments, learning a second formation associated with a respective team not in possession;

mapping the first formation to a first semantic label;

mapping the second formation to a second semantic label;

detecting a formation change based on the mapped first formation or the mapped second formation;

outputting a flag corresponding to the detected formation change.

16. The system of claim 15 , wherein the operations further comprise:

receiving, from a user device, a request to view a formation of a target team across a target game; and

based on the request, generating a graphical output that visually indicates a change in the formation of the target team across the target game.

17. The system of claim 15 , wherein the operations further comprise:

receiving, from a user device, a request to view a formation of a target team across a season;

identifying a plurality of target games associated with the season; and

based on the request, generating a graphical output that visually indicates one or more changes in the formation of the target team across a target season.

18. The system of claim 15 , wherein the operations further comprise:

receiving, from a user device, a request to view a response of a target team to a trajectory of a ball path; and

based on the request, generating a graphical output that visually indicates a formation of the target team formation responsive to the trajectory of the ball path.

19. The system of claim 18 , wherein the operations further comprise:

receiving, from the user device, a second request to view a second response of a second target team to the trajectory of the ball path; and

based on the request, generating a second graphical output that visually indicates a second target formation of the second target team responsive to the trajectory of the ball path and compares the second target formation of the second target team to the formation of the target team.

20. The system of claim 15 , wherein learning the first formation associated with the respective team in possession comprises:

generating a vector representation of the player tracking data, wherein the vector representation comprises a total number of frames, a total number of players, and a dimensionality of the player tracking data.