System and methods for implementing a feature set of high-dimensional spatial data in sports predictions
A method for generating a probability for a first action of a sporting event by implementing a feature set, the method including: obtaining an initial set of data relating to the first action of a sporting event, the initial set of data including at least a position of a first player on a surface and a position of a target area on the surface; generating, by a machine learning model, an initial projected scoring probability based on the initial set of data; generating a feature set relating to the sporting event; and modifying, by the machine learning model, the initial projected scoring probability to an updated scoring probability using the feature set.
1. A method for generating a probability for a first action of a sporting event by implementing a feature set, the method comprising:
obtaining an initial set of data relating to the first action of a sporting event, the initial set of data including at least a position of a first player on a surface and a position of a target area on the surface;
generating, by a machine learning model, an initial projected scoring probability based on the initial set of data, wherein the machine learning model includes a generative adversarial network (GAN) model, and wherein the GAN model includes at least one of a set of monotonic constraints or a weight for each of the monotonic constraints;
generating a feature set relating to the sporting event, the feature set being derived from any of a position of a second player on the surface, at least one of a distance or an angle between any of the first player, the second player, and the target area of the surface, wherein the first player and the second player are on opposing teams in the sporting event; and
modifying, by the machine learning model, the initial projected scoring probability to an updated scoring probability using the feature set;
wherein the feature set includes: a virtual line directed between the position of the first player and the position of the target area and a distance between the second player and an area of the virtual line;
wherein the feature set further includes the second player's save probability permutation modifying a save probability for each of a set of shot locations at different angles from the virtual line between the position of the first player and the position of the target area.
2. The method of claim 1 , wherein the feature set includes a previous action type specifying a previous action performed prior to an occurrence of the first action.
3. The method of claim 1 , wherein any of the initial set of data and the feature set includes historical save probabilities for the second player.
4. A method for generating a projected save probability for a first action of a sporting event using a feature set, the method comprising:
obtaining an initial set of data relating to the first action of a sporting event, the initial set of data including at least a position of an offensive player on a field and a position of a goal on the field;
generating, by a machine learning model, an initial projected save probability based on the initial set of data, wherein the machine learning model includes a generative adversarial network (GAN) model, and wherein the GAN model comprises at least one of a set of monotonic constraints or a weight for each of the monotonic constraints;
generating a feature set relating to the sporting event, the feature set derived from any of a position of a defending player on the field, at least one of a distance or an angle between any of the defending player, the offensive player, and the goal, or a presence of one or more additional players within a proximity of the defending player, the offensive player, and the goal; and
modifying, by the machine learning model, the initial projected save probability to an updated projected save probability using the feature set;
wherein the feature set includes: a virtual line directed between the position of the offensive player and the position of the goal; and a distance between the position of the defending player and the virtual line between the position of the offensive player and the position of the goal;
wherein the feature set further includes a goalkeeper save probability permutation modifying a goalkeeper save probability for each of a set of shot locations at different angles from the virtual line between the position of the offensive player and the position of the goal.
5. The method of claim 4 , wherein the feature set includes a clarity value specifying a number of intervening players positioned within a proximity of the virtual line between the position of the offensive player and the position of the goal.
6. The method of claim 4 , wherein the feature set includes a previous action type specifying a previous action performed prior to an occurrence of the first action.
7. The method of claim 4 , wherein the first action comprises a shot on the goal, the sporting event includes a soccer match, and the defending player is a goalkeeper.
8. The method of claim 7 , wherein the feature set includes a shot archetype specifying one or more of a shot type, whether the shot is contested, a distance range of the shot, and/or whether the shot is from a central portion of the field or a lateral portion of the field.
9. The method of claim 4 , wherein any of the initial set of data or the feature set includes historical save probabilities for the defending player.
10. A method for generating a projected point winner probability for a shot of a tennis match using a feature set, the method comprising:
obtaining an initial set of data relating to an action of a sporting event, the initial set of data including at least a position of a first player on a surface and a position of a target area on a field;
generating, by a machine learning model, an initial projected point winner probability based on the initial set of data, wherein the machine learning model includes a generative adversarial network (GAN) model, and wherein the GAN model comprises at least one of a set of monotonic constraints or a weight for each of the monotonic constraints;
generating a feature set relating to the sporting event, the feature set derived from any of a position of a second player on the field, a distance and/or an angle between any of the second player, the first player, and the target area, or a presence of any additional players within a proximity of the second player, the first player, and the target area, wherein the first player and the second player are on opposing teams in the sporting event; and
modifying, by the machine learning model, the initial projected point winner probability to an updated projected point winner probability using the feature set;
wherein the feature set includes: a virtual line directed between the position of the first player and the position of the target area; and a distance between the position of the second player and the virtual line between the position of the first player and the target area,
wherein the feature set further includes data of a shot trajectory, 3D body-pose information of the first player and second player, a current shot count, velocity and acceleration information of the first player and second player, a court surface type, and/or a shot type.