# US 2024/0198237 A1 · Training method of model, game confrontation player matching method, medium and device

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

- Number: US 2024/0198237 A1
- Document ID: US20240198237A1
- Status: Published application
- Kind: A1
- Application number: 18/533,160
- Filed: 2023-12-07
- Published: 2024-06-20
- Earliest claimed: 2022-12-15
- Assignees: Beijing Zitiao Network Technology Co., Ltd.
- Applicants: Beijing Zitiao Network Technology Co., Ltd.
- Inventors: Hanye LI
- Web page: https://patentgrep.com/patent/US20240198237A1

## CPC

- A63F 13/798: for assessing skills or for ranking players, e.g. for generating a hall of fame
- 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 13/795: for finding other players; for building a team; for providing a buddy list
- A63F 13/847: Cooperative playing, e.g. requiring coordinated actions from several players to achieve a common goal
- A63F 13/35: Details of game servers
- A63F: CARD, BOARD, OR ROULETTE GAMES; INDOOR GAMES USING SMALL MOVING PLAYING BODIES; VIDEO GAMES; GAMES NOT OTHERWISE PROVIDED FOR

## Abstract

The present disclosure relates to a training method of a model, a game confrontation player matching method, a medium, and a device. The training method of the advantage prediction model includes: acquiring an advantage value of each confrontation among a plurality of confrontations within a preset historical period and historical confrontation positive performance sample data of both game players before each confrontation; and training a first preset initial model by taking the advantage value of each confrontation among the plurality of confrontations and the historical confrontation positive performance sample data of both game players before each confrontation as first training data, so as to obtain the advantage prediction model.

## Claims

### Claim 1

A training method of an advantage prediction model, comprising:
acquiring an advantage value of each confrontation among a plurality of confrontations within a preset historical period and historical confrontation positive performance sample data of both game players before each confrontation; and
training a first preset initial model by taking the advantage value of each confrontation among the plurality of confrontations and the historical confrontation positive performance sample data of both game players before each confrontation as first training data, so as to obtain the advantage prediction model.

### Claim 2 (depends on 1)

The method according to claim 1, wherein acquiring the advantage value of each confrontation among the plurality of confrontations within the preset historical period comprises:
acquiring respective game tiers of both game players before each confrontation and game data corresponding to each time bucket during confrontation process, wherein the game data comprises a score difference of each game scoring dimension;
determining an average tier and a tier difference of the confrontation according to respective game tiers of both game players before each confrontation; and
determining the advantage value of each confrontation according to the average tier, the tier difference, and the game data.

### Claim 3 (depends on 2)

The method according to claim 2, wherein determining the advantage value of each confrontation according to the average tier, the tier difference, and the game data comprises:
determining a game win rate of a designated opponent at each time point according to the average tier, the tier difference, and the game data corresponding to each time bucket;
calculating an advantage period of the designated opponent and an advantage level at each time point according to the game win rate of the designated opponent at each time point; and
determining the advantage value of each confrontation according to the advantage period and the advantage level of each time point.

### Claim 4 (depends on 3)

The method according to claim 3, wherein determining the game win rate of the designated opponent at each time point according to the average tier, the tier difference, and the game data corresponding to each time bucket comprises:
determining a target win rate prediction model from a target set according to the average tier and a target identifier of each time bucket, wherein the target set comprises a win rate prediction model for each time bucket under different average tiers; and
taking the average tier, the tier difference, and the score difference of each game scoring dimension as inputs of the target prediction model, so as to acquire the game win rate output by the target prediction model.

### Claim 5 (depends on 4)

The method according to claim 4, wherein a training method of the win rate prediction model for each time bucket under different average tiers comprises:
acquiring historical game sample data, wherein the historical game sample data comprises a tier difference sample feature of both game players under each time bucket in each confrontation among the plurality of confrontations and a score difference sample feature of each game scoring dimension; and
determining the win rate prediction model corresponding to each time bucket under each average tier according to the tier difference sample feature of each time bucket under each average tier in the historical game sample data and the score difference sample feature.

### Claim 6 (depends on 1)

A game confrontation player matching method, applied to a controller, wherein the controller comprises the advantage prediction model according to claim 1, and the method comprises:
in response to a game opponent matching request from a target player, determining a predicted advantage value of the target player over each player to be matched through the advantage prediction model, wherein the predicted advantage value is used to represent an advantage that the target player has when the player to be matched plays against the target player; and
matching the target player with an opponent player from a plurality of players to be matched according to the predicted advantage value.

### Claim 7 (depends on 6)

The method according to claim 6, wherein matching the target player with the opponent player from the plurality of players to be matched according to the predicted advantage value comprises:
determining a target confrontation type corresponding to the target player, wherein the target confrontation type is used to represent different degrees of game experience needs; and
taking the player to be matched as the opponent player in a case where it is determined that the predicted advantage value belongs to the target confrontation type.

### Claim 8 (depends on 7)

The method according to claim 7, wherein determining that the predicted advantage value belongs to the target confrontation type comprises:
determining an advantage value interval corresponding to the target confrontation type; and
determining that the predicted advantage value belongs to the target confrontation type in a case where the predicted advantage value belongs to the advantage value interval.

### Claim 9 (depends on 6)

The method according to claim 6, wherein determining the predicted advantage value of the target player over each player to be matched comprises:
acquiring first confrontation positive performance data of the player to be matched within a designated historical period and second confrontation positive performance data of the target player; and
inputting the first confrontation positive performance data and the second confrontation positive performance data into the advantage prediction model that is preset, so as to acquire the predicted advantage value output by the advantage prediction model.

### Claim 10 (depends on 7)

The method according to claim 7, wherein determining the target confrontation type corresponding to the target player comprises:
acquiring user attribute data of the target player, wherein the user attribute data comprises a user identity feature and a user game attribute feature; and
inputting the user attribute data into a confrontation type prediction model that is preset, so as to acquire the target confrontation type output by the confrontation type prediction model.

### Claim 11 (depends on 10)

The method according to claim 10, wherein a training mode of the confrontation type prediction model comprises:
acquiring user attribute sample data of a plurality of players, wherein the user attribute sample data comprises confrontation type annotation data; and
training a second preset initial model by taking the user attribute sample data as second training data, so as to obtain the confrontation type prediction model.

### Claim 12

A training apparatus of an advantage prediction model, comprising:
an acquiring module, configured to acquire an advantage value of each confrontation among a plurality of confrontations within a preset historical period and historical confrontation positive performance sample data of both game players before each confrontation; and
a training module, configured to train a first preset initial model by taking the advantage value of each confrontation among the plurality of confrontations and the historical confrontation positive performance sample data of both game players before each confrontation as first training data, so as to obtain the advantage prediction model.

### Claim 13 (depends on 1)

A game confrontation player matching apparatus, applied to a controller, wherein the controller comprises the advantage prediction model according to claim 1, and the apparatus comprises:
a first determining module, configured to, in response to a game opponent matching request from a target player, determine a predicted advantage value of the target player over each player to be matched through the advantage prediction model, wherein the predicted advantage value is used to represent an advantage that the target player has when the player to be matched plays against the target player; and
a second determining module, configured to match the target player with an opponent player from a plurality of players to be matched according to the predicted advantage value.

### Claim 14 (depends on 1)

A computer readable medium, wherein a computer program is stored on the computer readable medium, and the computer program, when executed by a processing apparatus, implements steps of the method according to claim 1.

### Claim 15 (depends on 1)

An electronic device, comprising:
a storage apparatus, storing a computer program; and
a processing apparatus, configured to execute the computer program in the storage apparatus, so as to implement steps of the method according to claim 1.

### Claim 16 (depends on 6)

A computer readable medium, wherein a computer program is stored on the computer readable medium, and the computer program, when executed by a processing apparatus, implements steps of the method according to claim 6.

### Claim 17 (depends on 6)

An electronic device, comprising:
a storage apparatus, storing a computer program; and
a processing apparatus, configured to execute the computer program in the storage apparatus, so as to implement steps of the method according to claim 6.

### Claim 18 (depends on 2)

A game confrontation player matching method, applied to a controller, wherein the controller comprises the advantage prediction model according to claim 2, and the method comprises:
in response to a game opponent matching request from a target player, determining a predicted advantage value of the target player over each player to be matched through the advantage prediction model, wherein the predicted advantage value is used to represent an advantage that the target player has when the player to be matched plays against the target player; and
matching the target player with an opponent player from a plurality of players to be matched according to the predicted advantage value.

### Claim 19 (depends on 3)

A game confrontation player matching method, applied to a controller, wherein the controller comprises the advantage prediction model according to claim 3, and the method comprises:
in response to a game opponent matching request from a target player, determining a predicted advantage value of the target player over each player to be matched through the advantage prediction model, wherein the predicted advantage value is used to represent an advantage that the target player has when the player to be matched plays against the target player; and
matching the target player with an opponent player from a plurality of players to be matched according to the predicted advantage value.

### Claim 20 (depends on 4)

A game confrontation player matching method, applied to a controller, wherein the controller comprises the advantage prediction model according to claim 4, and the method comprises:
in response to a game opponent matching request from a target player, determining a predicted advantage value of the target player over each player to be matched through the advantage prediction model, wherein the predicted advantage value is used to represent an advantage that the target player has when the player to be matched plays against the target player; and
matching the target player with an opponent player from a plurality of players to be matched according to the predicted advantage value.
