# US 2023/0056715 A1 · Video game with coaching session

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2023/0056715 A1
- Document ID: US20230056715A1
- Status: Published application
- Kind: A1
- Application number: 17/887,868
- Filed: 2022-08-15
- Published: 2023-02-23
- Earliest claimed: 2021-08-23
- Assignees: Square Enix Co., Ltd.
- Applicants: Square Enix Ltd.
- Inventors: Carle CÔTÉ; Ugo LOUCHE; Eric MARTEL; Romain TRACHEL
- Web page: https://patentgrep.com/patent/US20230056715A1

## CPC

- A63F 13/5375: for graphically or textually suggesting an action, e.g. by displaying an arrow indicating a turn in a driving game
- A63F 13/67: adaptively or by learning from player actions, e.g. skill level adjustment or by storing successful combat sequences for re-use
- A63F 13/533: for prompting the player, e.g. by displaying a game menu
- A63F 13/79: involving player-related data, e.g. identities, accounts, preferences or play histories
- G06N 20/00: Machine learning
- A63F: CARD, BOARD, OR ROULETTE GAMES; INDOOR GAMES USING SMALL MOVING PLAYING BODIES; VIDEO GAMES; GAMES NOT OTHERWISE PROVIDED FOR
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A method of operating a computer for carrying out a video game, a computer-readable storage medium and a video game apparatus for a video game. The method includes maintaining a gaming environment in a non-transitory memory, the gaming environment defining an autonomous agent and a behavioral model associated with the agent. The method also includes implementing a gaming session of the video game during which the agent carries out actions in accordance with the behavioral model. The method also includes implementing a coaching session of the video game by receiving instructions from a player of the video game, the instructions indicative of a gameplay action to be performed by the agent in the gaming environment. The method also includes implementing the coaching session of the video game by receiving contextual feedback from the player regarding said gameplay action. The method also includes implementing the coaching session of the video game.

## Claims

### Claim 1

A method of operating a computer for carrying out a video game, the method comprising:
maintaining a gaming environment in a non-transitory memory of the computer, the gaming environment defining an autonomous agent and a behavioral model associated with the agent;
implementing a gaming session of the video game during which the autonomous agent carries out actions in accordance with the behavioral model; and
implementing a coaching session of the video game by:
a. receiving instructions from a player of the video game, the instructions indicative of a gameplay action that the player wants performed by the autonomous agent in the gaming environment;
b. receiving contextual feedback from the player regarding said gameplay action; and
c. adapting the behavioral model based on the instructions and the contextual feedback.

### Claim 2 (depends on 1)

The method of claim 1, the gaming environment further defining current game data, wherein the behavioral model includes inputs and outputs, wherein the inputs of the behavioral model comprise the current game data and wherein the outputs of the behavioral model comprise an indicator for each performable gameplay action of a plurality of performable gameplay actions.

### Claim 3 (depends on 2)

The method of claim 2, wherein the behavioral model is configured to compute the indicator for each of the performable gameplay actions based on the current game data, wherein the indicator for a particular performable gameplay action includes a preference score associated with the particular performable gameplay action.

### Claim 4 (depends on 3)

The method of claim 3, wherein the preference score is indicative of the player’s preference for the particular performable gameplay action given the current game data.

### Claim 5 (depends on 3)

The method of claim 3, wherein the method further comprises, during the gaming session, identifying the performable gameplay action having the highest preference score, wherein at least one of the actions carried out by the autonomous agent in accordance with the behavioral model comprises the performable gameplay action having the highest preference score.

### Claim 6 (depends on 4)

The method of claim 4, wherein the method further comprises, during the gaming session, identifying a subset of performable gameplay actions each having the highest preference score, wherein at least one of the actions carried out by the autonomous agent in accordance with the behavioral model comprises a randomly selected one of the performable gameplay actions selected in the subset of performable gameplay actions.

### Claim 7 (depends on 5)

The method of claim 5, wherein said identifying the performable gameplay action having the highest preference score is carried out continuously, periodically or based on occurrence of an event.

### Claim 8 (depends on 5)

The method of claim 5, wherein the performable gameplay action having the highest preference score changes dynamically.

### Claim 9 (depends on 3)

The method of claim 3, wherein the preference score associated with the particular performable gameplay action is computed based on a set of contextual parameters associated with the particular performable gameplay action.

### Claim 10 (depends on 9)

The method of claim 9, wherein each contextual parameter in the set of contextual parameters reflects a parameter to be assessed by the autonomous agent to determine if a particular performable gameplay action should be carried out during a gaming session given the current game data.

### Claim 11 (depends on 10)

The method of claim 10, wherein the current game data comprises a plurality of game state variables, wherein each of the contextual parameters in the set of contextual parameters is associated with a set of contextual factors and wherein each contextual factor in the set of contextual factors is associated with one or more of the game state variables.

### Claim 12 (depends on 11)

The method of claim 11, further comprising:
a. generating a contextual parameter aggregate for each given contextual parameter in the set of contextual parameters associated with the particular performable gameplay action based on (i) a contextual factor value for each given contextual factor in the set of contextual factors associated with the given contextual parameter and (ii) a weight associated with each given contextual factor, the weight being indicative of a contribution of the contextual factor value to the contextual parameter aggregate, wherein the contextual factor value for each given contextual factor in the set of contextual factors associated with the given contextual parameter is based on values of the game state variables associated with the given contextual factor;
b. computing the preference score associated with the performable gameplay action by combining each said contextual parameter aggregate for each given contextual parameter in the set of contextual parameters associated with the particular performable gameplay action.

### Claim 13 (depends on 12)

The method of claim 12, wherein said combining comprises summing, averaging or linearly combining.

### Claim 14 (depends on 13)

The method of claim 13, wherein said linearly combining includes associating a coefficient with each said contextual parameter aggregate, the coefficient being indicative of relative importance of the contextual parameters in the set of contextual parameters.

### Claim 15 (depends on 12)

The method of claim 12, wherein implementing a coaching session of the video game by adapting the behavioral model comprises:
based on the instructions indicative of a gameplay action that the player wants performed by the autonomous agent in the gaming environment, identifying the set of contextual parameters associated with said gameplay action.

### Claim 16 (depends on 15)

The method of claim 15, wherein said adapting the behavioral model further comprises, based on the contextual feedback:
a. identifying a particular contextual parameter from the set of contextual parameters associated with said gameplay action that is a personalizable contextual parameter;
b. identifying a particular contextual factor to be associated with the personalizable contextual parameter; and
c. personalizing the personalizable contextual parameter by creating an association between the particular contextual factor and the personalizable contextual parameter, the association being stored in memory.

### Claim 17 (depends on 16)

The method of claim 16, wherein the particular contextual factor is indicative of the player’s interpretation of the associated personalizable contextual parameter.

### Claim 18 (depends on 16)

The method of claim 16, wherein personalizing the personalizable contextual parameter comprises modifying the weight associated with the particular contextual factor.

### Claim 19 (depends on 16)

The method of claim 16, wherein the contextual feedback comprises player input related to (i) a contextual parameter that is personalizable or (ii) a particular contextual factor to be associated with a contextual parameter that is personalizable.

### Claim 20 (depends on 1)

The method of claim 1, wherein implementing the coaching session further comprises, prior to the receiving instructions, providing the player with an indication of the performable gameplay actions that the autonomous agent can be instructed to carry out.

### Claim 21 (depends on 1)

The method of claim 1, wherein the coaching session is a first coaching session, the player is a first player, the instructions are first instructions, the gameplay action is a first gameplay action and the contextual feedback is first contextual feedback, the method further comprising implementing a second coaching session of the video game by:
receiving second instructions from a second player of the video game, the second instructions indicative of a second gameplay action that the second player wants performed by the autonomous agent in the gaming environment;
receiving second contextual feedback from the second player regarding said second gameplay action; and
adapting the behavioral model based on the second instructions and the second contextual feedback,
wherein the first and the second gameplay actions are the same and wherein the first and the second contextual feedback are different.

### Claim 22 (depends on 1)

The method of claim 1, wherein implementing the coaching session further comprises providing the player with an opportunity to select the contextual feedback from a set of contextual feedback options.

### Claim 23 (depends on 22)

The method of claim 22, wherein the coaching session is a current coaching session preceded by at least one previous coaching session in which previous instructions indicative of a previous gameplay action that the player wants performed by the autonomous agent in the gaming environment and previous contextual feedback from the player regarding the previous gameplay action have been received, the previous instructions and contextual feedback being part of a set of past player recommendations, and
wherein the method further comprises generating one or more markers of inconsistencies in the set of contextual factors, and wherein generating the one or more markers of inconsistencies comprises:
a. identifying the set of contextual factors associated with a given personalizable contextual parameter;
b. generating the one or more markers of inconsistencies for each contextual factor in the set of contextual factors based at least on a statistical analysis of the set of past player recommendations;
c. storing the one or more markers of inconsistencies in memory.

### Claim 24 (depends on 23)

The method of claim 23, wherein the one or more markers of inconsistencies comprise one or more of a marker of conflict, a marker of importance or a marker of independence.

### Claim 25 (depends on 24)

The method of claim 24, wherein further to generating the one or more markers of inconsistencies, the method further comprises requesting player input related to the one or more markers of inconsistencies and in response to the player input related to the one or more markers of inconsistences, modifying the set of contextual factors associated with the given personalizable contextual parameter.

### Claim 26 (depends on 25)

The method of claim 25, wherein modifying the set of contextual factors associated with the given personalizable contextual parameter comprises removing a contextual factor from the set of contextual factors or adding a contextual factor to the set of contextual factors.

### Claim 27 (depends on 1)

The method of claim 1, wherein implementing the coaching session comprises executing a graphical user interface configured to provide the player with an opportunity to enter the instructions and the contextual feedback.

### Claim 28 (depends on 27)

The method of claim 27, wherein the graphical user interface is configured to interact with the player by way of a structured dialogue.

### Claim 29

A non-transitory computer-readable storage medium comprising computer-readable instructions which, when read and executed by at least one processor, cause the at least one processor to execute a method of operating a computer for carrying out a video game, wherein the method comprises:
maintaining a gaming environment in a non-transitory memory of the computer, the gaming environment defining an autonomous agent and a behavioral model associated with the agent;
implementing a gaming session of the video game during which the autonomous agent carries out actions in accordance with the behavioral model; and
implementing a coaching session of the video game by:
a. receiving instructions from a player of the video game, the instructions indicative of a gameplay action that the player wants performed by the autonomous agent in the gaming environment;
b. receiving contextual feedback from the player regarding said gameplay action; and
c. adapting the behavioral model based on the instructions and the contextual feedback.

### Claim 30

A video game apparatus comprising:
at least one processor;
a memory storing instructions for execution by the at least one processor; and
at least one input device configured to receive input from a player, at least one output device configured for providing output to the player,
wherein the at least one processor is configured to execute the instructions in the memory for implementing an interactive computer program that generates the output in response to the received input and, the interactive computer program including at least one process that comprises:
maintaining a gaming environment in a non-transitory memory of the computer, the gaming environment defining an autonomous agent and a behavioral model associated with the agent;
implementing a gaming session of the video game during which the autonomous agent carries out actions in accordance with the behavioral model; and
implementing a coaching session of the video game by:
a. receiving instructions from a player of the video game, the instructions indicative of a gameplay action that the player wants performed by the autonomous agent in the gaming environment;
b. receiving contextual feedback from the player regarding said gameplay action; and
c. adapting the behavioral model based on the instructions and the contextual feedback.
