Method and system for interactive, interpretable, and improved match and player performance predictions in team sports

Patent metadata

Field	Value
Publication number	US-11577145-B2
Application number	US-201916254108-A
Country	US
Kind code	B2
Filing date	Jan 22, 2019
Priority date	Jan 21, 2018
Publication date	Feb 14, 2023
Grant date	Feb 14, 2023

Abstract

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.

First claim

What is claimed:

1. A method of generating an outcome for a sporting event, comprising:
retrieving, by a computing system, tracking data from a data store, the tracking data comprising event data for a plurality of events across a plurality of seasons;
generating, by the computing system, a predictive model using a deep neural network, by:
learning, by a neural network, one or more players likely to be in each event at each time t, based on lineup features of each team, current state of each event at each time t, and current box score at each time t;
generating a data set comprising the one or more players likely to be in each event at each time t; and
learning, by a mixture density network, a score difference at each time t, based on the lineup features of each team, the current state of each event at each time t, the current box score at each time t, and the data set comprising the one or more players likely to be in each event at each time t;

receiving, by the computing system, an indication to generate a predicted outcome of the sporting event at a time T; and
generating, by the computing system via the predictive model, a final score differential for the sporting event based on lineup features of each team to the sporting event, current state of the sporting event at the time T, current box score at the time T, and current lineup in the sporting event at time T.

2. The method of claim 1 , wherein receiving, by the computing system, the indication to generate the predicted outcome of the sporting event at the time T comprises:
receiving, from a client device, a request to predict the outcome of the sporting event at the time T.

3. The method of claim 1 , wherein learning, by the mixture density network, the score difference at each time t, comprises:
comparing the score difference at each time t to an actual score difference at each time t; and
minimizing an error between the score difference and the actual score difference using a negative log likelihood of finding an optimal set of parameters.

4. The method of claim 1 , wherein learning, by the neural network, the one or more players likely to be in each event at each time t, given time based on the lineup features of each team, the current state of each event at each time t, and the current box score at each time t comprises:
learning one or more lineup encoding features.

5. The method of claim 4 , further comprising:
training a set of random forest classifiers with the one or more lineup encoding features.

6. The method of claim 1 , wherein the lineup features of each team are represented as a union of a plurality of lineup vectors constructed for each player on each team.

7. The method of claim 1 , wherein the current state of each event at each time t is represented by a vector comprising each play-by-play event up to the time t and at each time t.

8. A system for predicting an outcome of a sporting event, comprising:
a processor; and
a memory having programming instructions stored thereon, which, when executed by the processor, performs one or more operations comprising:
retrieving tracking data from a data store, the tracking data comprising event data for a plurality of events across a plurality of seasons;
generating a predictive model using a deep neural network, by:
learning, by a neural network, one or more players likely to be in each event at each time t, based on lineup features of each team, current state of each event at each time t, and current box score at each time t;
generating a data set comprising the one or more players likely to be in each event at each time t; and
learning, by a mixture density network, a score difference at each time t, based on the lineup features of each team, the current state of each event at each time t, the current box score at each time t, and the data set comprising the one or more players likely to be in each event at each time t;

receiving an indication to generate a predicted outcome of the sporting event at a time T; and
generating, via the predictive model, a final score differential for the sporting event based on lineup features of each team to the sporting event, current state of the sporting event at the time T, current box score at the time T, and current lineup in the sporting event at time T.

9. The system of claim 8 , wherein receiving the indication to generate the predicted outcome of the sporting event at the time T comprises:
receiving, from a client device, a request to predict the outcome of the sporting event at the time T.

10. The system of claim 8 , wherein learning, by the mixture density network, the score difference at each time t, comprises:
comparing the score difference at each time t to an actual score difference at each time t; and
minimizing an error between the score difference and the actual score difference using a negative log likelihood of finding an optimal set of parameters.

11. The system of claim 8 , wherein learning, by the neural network, the one or more players likely to be in each event at each time t, given time based on the lineup features of each team, the current state of each event at each time t, and the current box score at each time t comprises:
learning one or more lineup encoding features.

12. The system of claim 8 , wherein the one or more operations further comprise:
training a set of random forest classifiers with one or more lineup encoding features.

13. The system of claim 8 , wherein the lineup features of each team are represented as a union of a plurality of lineup vectors constructed for each player on each team.

14. The system of claim 8 , wherein the current state of each event at each time t is represented by a vector comprising each play-by-play event up to the time t and at each time t.

15. A non-transitory computer readable medium including one or more sequences of instructions that, when executed by a processor, causes a computing system to perform operations comprising:
retrieving, by the computing system, tracking data from a data store, the tracking data comprising event data for a plurality of events across a plurality of seasons;
generating, by the computing system, a predictive model using a deep neural network, by:
learning, by a neural network, one or more players likely to be in each event at each time t, based on lineup features of each team, current state of each event at each time t, and current box score at each time t;
generating a data set comprising the one or more players likely to be in each event at each time t; and
learning, by a mixture density network, a score difference at each time t, based on the lineup features of each team, the current state of each event at each time t, the current box score at each time t, and the data set comprising the one or more players likely to be in each event at each time t;

receiving, by the computing system, an indication to generate a predicted outcome of a sporting event at a time T; and
generating, by the computing system via the predictive model, a final score differential for the sporting event based on lineup features of each team to the sporting event, current state of the sporting event at the time T, current box score at the time T, and current lineup in the sporting event at time T.

16. The non-transitory computer readable medium of claim 15 , wherein receiving, by the computing system, the indication to generate the predicted outcome of the sporting event at the time T comprises:
receiving, from a client device, a request to predict the outcome of the sporting event at the time T.

17.

Assignees

• Stats Llc

Inventors

• Ruiz Hector
• Ganguly Sujoy
• Frank Nathan
• Lucey Patrick

Classifications

• G06N3/0499

  Feedforward networks · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06F2123/02

  in the time domain, e.g. time-series data · CPC title

• G06N3/084

  Backpropagation, e.g. using gradient descent · CPC title

• G06N3/047

  Probabilistic or stochastic networks · CPC title