Method and system for interactive, interpretable, and improved match and player performance predictions in team sports
A method of generating an outcome for a sporting event is disclosed herein. A computing system retrieves tracking data from a data store. The computing system generates a predictive model using a deep neural network. The one or more neural networks of the deep neural network generates one or more embeddings comprising team-specific information and agent-specific information based on the tracking data. The computing system selects, from the tracking data, one or more features related to a current context of the sporting event. The computing system learns, by the deep neural network, one or more likely outcomes of one or more sporting events. The computing system receives a pre-match lineup for the sporting event. The computing system generates, via the predictive model, a likely outcome of the sporting event based on historical information of each agent for the home team, each agent for the away team, and team-specific features.
1 . A method of generating an outcome for a sporting event, comprising:
receiving, by a computing system, lineup information for a sporting event, the lineup information comprising a first lineup comprising a first plurality of players for a home team and a second lineup comprising a second plurality of players for an away team;
identifying, by the computing system, first team embeddings corresponding to the home team and first player-specific embeddings corresponding to the first plurality of players in the first lineup, the first team embeddings and the first player-specific embeddings having a lower dimension than the lineup information;
identifying, by the computing system, second team embeddings corresponding to the away team and second player-specific embeddings corresponding to the second plurality of players in the second lineup, the second team embeddings and the second player-specific embeddings having the lower dimension than the lineup information;
determining, by the computing system, a current context of the sporting event, the current context comprising a current score and a current time remaining; and
generating, by the computing system via a predictive model, a predicted outcome of the sporting event based on the current context of the sporting event, the first team embeddings, the first player-specific embeddings, the second team embeddings, and the second player-specific embeddings.
2 . The method of claim 1 , further comprising:
generating, by the computing system, a first graphical user interface comprising the first plurality of players for the home team; and
receiving, by the computing system via the first graphical user interface, a user input substituting a first player of the first plurality of players for the home team for a second player of the first plurality of players for the home team.
3 . The method of claim 2 , further comprising:
updating, by the computing system, the first player-specific embeddings by substituting existing player embeddings corresponding to the first player with new player embeddings corresponding to the second player; and
generating, by the computing system via the predictive model, an updated predicted outcome of the sporting event based on the current context of the sporting event, the first team embeddings, the updated first player-specific embeddings, the second team embeddings, and the second player-specific embeddings.
4 . The method of claim 1 , further comprising:
determining, by the computing system, that the current context of the sporting event comprises a current time remaining that is less than a total length of the sporting event; and
based on the determining, projecting, by the computing system, a current lineup for the home team and the away team based on the first lineup of the home team and the second lineup of the away team.
5 . The method of claim 4 , further comprising:
generating, by the computing system via the predictive model, an updated predicted outcome of the sporting event based on the current context of the sporting event, the first team embeddings, updated first player-specific embeddings based on the projected current lineup for the home team, the second team embeddings, and updated second player-specific embeddings based on the projected current lineup for the away team.
6 . The method of claim 1 , further comprising:
generating, by the computing system, a graphical user interface comprising a graphical representation of the predicted outcome.
7 . The method of claim 1 , wherein the current time remaining is a total length of the sporting event and the first lineup for the home team and the second lineup for the away team are starting lineups for the home team and the away team.
8 . The method of claim 1 , where the current context of the sporting event further includes an injury corresponding to at least one of the first plurality of players or at least one of the second plurality of players.
9 . A non-transitory computer readable medium comprising one or more sequences of instructions, which, when executed by a processor, causes a computing system to perform operations comprising:
receiving, by the computing system, lineup information for a sporting event, the lineup information comprising a first lineup comprising a first plurality of players for a home team and a second lineup comprising a second plurality of players for an away team;
identifying, by the computing system, first team embeddings corresponding to the home team and first player-specific embeddings corresponding to the first plurality of players in the first lineup, the first team embeddings and the first player-specific embeddings having a lower dimension than the lineup information;
identifying, by the computing system, second team embeddings corresponding to the away team and second player-specific embeddings corresponding to the second plurality of players in the second lineup, the second team embeddings and the second player-specific embeddings having the lower dimension than the lineup information;
determining, by the computing system, a current context of the sporting event, the current context comprising a current score and a current time remaining; and
generating, by the computing system via a predictive model, a predicted outcome of the sporting event based on the current context of the sporting event, the first team embeddings, the first player-specific embeddings, the second team embeddings, and the second player-specific embeddings.
10 . The non-transitory computer readable medium of claim 9 , further comprising:
generating, by the computing system, a first graphical user interface comprising the first plurality of players for the home team; and
receiving, by the computing system via the first graphical user interface, a user input substituting a first player of the first plurality of players for the home team for a second player of the first plurality of players for the home team.
11 . The non-transitory computer readable medium of claim 10 , further comprising:
updating, by the computing system, the first player-specific embeddings by substituting existing player embeddings corresponding to the first player with new player embeddings corresponding to the second player; and
generating, by the computing system via the predictive model, an updated predicted outcome of the sporting event based on the current context of the sporting event, the first team embeddings, the updated first player-specific embeddings, the second team embeddings, and the second player-specific embeddings.
12 . The non-transitory computer readable medium of claim 9 , further comprising:
determining, by the computing system, that the current context of the sporting event comprises a current time remaining that is less than a total length of the sporting event; and
based on the determining, projecting, by the computing system, a current lineup for the home team and the away team based on the first lineup of the home team and the second lineup of the away team.
13 . The non-transitory computer readable medium of claim 12 , further comprising:
generating, by the computing system via the predictive model, an updated predicted outcome of the sporting event based on the current context of the sporting event, the first team embeddings, updated first player-specific embeddings based on the projected current lineup for the home team, the second team embeddings, and updated second player-specific embeddings based on the projected current lineup for the away team.
14 . The non-transitory computer readable medium of claim 9 , further comprising:
generating, by the computing system, a graphical user interface comprising a graphical representation of the predicted outcome.
15 . The non-transitory computer readable medium of claim 9 , wherein the current time remaining is a total length of the sporting event and the first lineup for the home team and the second lineup for the away team are starting lineups for the home team and the away team.
16 . 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:
receiving lineup information for a sporting event, the lineup information comprising a first lineup comprising a first plurality of players for a home team and a second lineup comprising a second plurality of players for an away team;
identifying first team embeddings corresponding to the home team and first player-specific embeddings corresponding to the first plurality of players in the first lineup, the first team embeddings and the first player-specific embeddings having a lower dimension than the lineup information;
identifying second team embeddings corresponding to the away team and second player-specific embeddings corresponding to the second plurality of players in the second lineup, the second team embeddings and the second player-specific embeddings having the lower dimension than the lineup information;
determining a current context of the sporting event, the current context comprising a current score and a current time remaining; and
generating, via a predictive model, a predicted outcome of the sporting event based on the current context of the sporting event, the first team embeddings, the first player-specific embeddings, the second team embeddings, and the second player-specific embeddings.
17 . The system of claim 16 , wherein the operations further comprise:
generating a first graphical user interface comprising the first plurality of players for the home team; and
receiving, via the first graphical user interface, a user input substituting a first player of the first plurality of players for the home team for a second player of the first plurality of players for the home team.
18 . The system of claim 17 , wherein the operations further comprise:
updating the first player-specific embeddings by substituting existing player embeddings corresponding to the first player with new player embeddings corresponding to the second player; and
generating, via the predictive model, an updated predicted outcome of the sporting event based on the current context of the sporting event, the first team embeddings, the updated first player-specific embeddings, the second team embeddings, and the second player-specific embeddings.
19 . The system of claim 16 , wherein the operations further comprise:
determining that the current context of the sporting event comprises a current time remaining that is less than a total length of the sporting event; and
based on the determining, projecting a current lineup for the home team and the away team based on the first lineup of the home team and the second lineup of the away team.
20 . The system of claim 19 , wherein the operations further comprise:
generating, via the predictive model, an updated predicted outcome of the sporting event based on the current context of the sporting event, the first team embeddings, updated first player-specific embeddings based on the projected current lineup for the home team, the second team embeddings, and updated second player-specific embeddings based on the projected current lineup for the away team.