WO2019085823A1 - Procédé, dispositif et support de stockage pour déterminer des informations de jeu, et dispositif électronique - Google Patents

Procédé, dispositif et support de stockage pour déterminer des informations de jeu, et dispositif électronique Download PDF

Info

Publication number
WO2019085823A1
WO2019085823A1 PCT/CN2018/111950 CN2018111950W WO2019085823A1 WO 2019085823 A1 WO2019085823 A1 WO 2019085823A1 CN 2018111950 W CN2018111950 W CN 2018111950W WO 2019085823 A1 WO2019085823 A1 WO 2019085823A1
Authority
WO
WIPO (PCT)
Prior art keywords
game
party
item
running
target
Prior art date
Application number
PCT/CN2018/111950
Other languages
English (en)
Chinese (zh)
Inventor
李立
邓大付
殷俊
苏博览
万乐
张银锋
卢伟平
王晓曦
Original Assignee
腾讯科技（深圳）有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 腾讯科技（深圳）有限公司 filed Critical 腾讯科技（深圳）有限公司
Publication of WO2019085823A1 publication Critical patent/WO2019085823A1/fr

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63F—CARD, BOARD, OR ROULETTE GAMES; INDOOR GAMES USING SMALL MOVING PLAYING BODIES; VIDEO GAMES; GAMES NOT OTHERWISE PROVIDED FOR
    • A63F13/00—Video games, i.e. games using an electronically generated display having two or more dimensions
    • A63F13/70—Game security or game management aspects
    • A63F13/79—Game security or game management aspects involving player-related data, e.g. identities, accounts, preferences or play histories
    • A63F13/798—Game security or game management aspects involving player-related data, e.g. identities, accounts, preferences or play histories for assessing skills or for ranking players, e.g. for generating a hall of fame

Definitions

  • the present application relates to the field of the Internet, and in particular, to a method and device for determining game information, a storage medium, and an electronic device.
  • the embodiment of the present application provides a method and device for determining game information, a storage medium, and an electronic device, so as to at least solve the technical problem that the accuracy of winning the game in the related art is low.
  • a method for determining game information includes: the terminal performs a simulation operation on a game of a target game, in a game running process, in a game of a game.
  • the first party uses the props in a game according to the target mode, and the second party in the game uses the props in a game according to a predetermined manner, and the target mode is the props usage mode of the first account in the target game;
  • the terminal acquires the pair
  • the running result of the simulation running of the game is used to indicate whether the first party wins in the simulation running of the game;
  • the terminal determines the game information through the plurality of running results, and the game information is used to indicate the first account as In the case of the first party, the probability of winning in a game is obtained, and the multiple running results are obtained by performing simulation operations on one game in multiple times.
  • a device for determining game information which is applied to a terminal, the device comprising: an operation unit configured to perform a simulation operation on a game of the target game, in a game During the simulation running, the first party in a game uses the props in a game according to the target mode, and the second party in the game uses the props in a game in a predetermined manner, and the target mode is in the target game.
  • the terminal performs a simulation operation on a game of the target game.
  • the first party in the game uses the props in one game according to the target mode, one game.
  • the second party in the game uses the props in a game in a predetermined manner; obtains the running result of the simulation running of the one game; and then determines the game information through the plurality of running results, the game information is used to indicate the first account as the first
  • the probability of winning in a game is obtained.
  • the multiple running results are obtained by simulating the running of a game in multiple times, which can solve the accuracy of the winning rate of the chess game in the related art.
  • the low problem achieves the technical effect of improving the accuracy of the score of the chess situation.
  • FIG. 1 is a schematic diagram of a hardware environment of a method of determining game information according to an embodiment of the present application
  • FIG. 2 is a flowchart of a method for determining optional game information according to an embodiment of the present application
  • FIG. 3 is a schematic diagram of an optional game level interface in accordance with an embodiment of the present application.
  • FIG. 4 is a schematic diagram of a method of determining game information according to an embodiment of the present application.
  • FIG. 5 is a schematic diagram of an alternative depth model in accordance with an embodiment of the present application.
  • FIG. 6 is a schematic diagram of an alternative simulated winning rate according to an embodiment of the present application.
  • FIG. 7 is a schematic diagram of an optional depth model training according to an embodiment of the present application.
  • FIG. 8 is a schematic diagram of an optional device for determining game information according to an embodiment of the present application.
  • Chess Game There are mainly Chinese chess, Chinese checkers, black and white chess and backgammon. Chess games are different from card games. Chess games are complete information games, and card games are incomplete information games. If the information of the card game is made public, the card game can be seen in the card game. If the landlord player knows the opponent's card, the landlord can also be regarded as a board game.
  • a method embodiment of a method for determining game information is provided.
  • the method for determining the game information may be applied to a hardware environment composed of the server 102 and the terminal 104 as shown in FIG. 1.
  • the server 102 is connected to the terminal 104 through a network.
  • the network includes but is not limited to a wide area network, a metropolitan area network, or a local area network.
  • the terminal 104 is not limited to a PC, a mobile phone, a tablet, or the like.
  • the method for determining the game information in the embodiment of the present application may be executed by the server 102, may be executed by the terminal 104, or may be performed by the server 102 and the terminal 104 in common.
  • the determining method of the game information by the terminal 104 in the embodiment of the present application may also be performed by a client installed thereon.
  • FIG. 2 is a flowchart of a method for determining optional game information according to an embodiment of the present application. As shown in FIG. 2, the method may include the following steps:
  • Step S202 the terminal performs a simulation running on a game of the target game.
  • the first party in the game uses the props in a game according to the target mode, and the first in the game.
  • the two parties use the props in a game according to a predetermined method, and the target mode is the props usage mode of the first account in the target game.
  • the target mode is the props usage mode of the first account in the target game.
  • the above target game is a game in which the game participants are two parties (ie, the first party and the second party described above), including but not limited to a board game or a card game.
  • the props can be game props used in the corresponding types of games, such as in a board game, the props are pawns; in a card game, the props are cards.
  • the second party in a game uses the props in a game in a predetermined manner, that is, for the second party, the props are used in advance, such as the second party is pre-edited (set). Game logic robot.
  • the props are used to simulate the use of the first account, and the second party uses the props according to the preset game logic to the first party. To respond to the props used by the first party.
  • the above account number can be a specific account or a general term for a type of account.
  • One type of account here refers to a type of account with the same or close game level.
  • step S204 the terminal acquires a running result of performing a simulation running on a game, and the running result is used to indicate whether the first party wins in the simulation running of the game.
  • the above running result is a game result corresponding to the game type, for example, in the case of a board game, when the first party or the second party's pieces satisfy the game rule stipulated victory (if one party eats or knocks the other party's pieces) , the result of the operation is obtained; for the card game, when the card of the first party or the second party satisfies the victory specified by the game rules (if the card is played), the running result is obtained.
  • Step S206 the terminal determines game information by using a plurality of running results, where the game information is used to indicate or indicate the probability of winning in a game in the case where the first account is the first party, and the multiple running results are multiple times. A game of the game was simulated.
  • the above game information is used to indicate the probability that the first party wins when the first account is the first party.
  • the terminal performs a simulation operation on a game of the target game.
  • the first party in the game uses the props in a game according to the target mode.
  • the second party in the game plays the props in a game in a predetermined manner; obtains the running result of the simulation running of the game; and then determines the game information through the plurality of running results, the game information is used to indicate the first account as In the case of the first party, the probability of winning in a game is obtained.
  • the multiple running results are obtained by simulating the running of the game in multiple times, which can solve the accuracy of the winning rate of the chess game in the related art.
  • the lower technical problems achieve the technical effect of improving the accuracy of the evaluation of the winning rate of the chess situation.
  • the terminal performs a simulation running on a game of the target game.
  • the first party in the game uses the props in a game according to the target mode.
  • the second party in the game uses the props in a game in a predetermined manner, and the target mode is the props usage mode of the first account in the target game.
  • the difficulty of the game is calculated by the characteristics of the game.
  • the characteristics of the game are selected. For example, for the chess game of Chinese chess, the remaining number of cars, the remaining number of horses, the number of remaining pieces, etc. can be selected as features, and then, the game is played.
  • the feature acts as a feature to be processed, and a mapping between features and difficulty assessments is established using rules or deep learning methods.
  • the simulated winning rate is used as the difficulty index of the game.
  • some heuristic rules are used as a gameplay.
  • the above two methods are combined, and the deep learning model is used to mine the relationship between the various features in the game, and the skills in the game are deeply explored, thereby avoiding the influence of the intuitive steps on the evaluation. , reducing the requirements for human and material resources, and improving the accuracy of the assessment.
  • the deep learning model of different levels of human player gameplay can be characterized, and the deep learning representing different human levels can be performed.
  • the model and the highest level of deep learning model are used to simulate the battle.
  • the simulated winning rate is used as the difficulty of the game played by different levels of human players, which can improve the accuracy of the evaluation.
  • a plurality of deep learning models can be used to mine the chess game processing features of different levels of human players to be able to assess the difficulty perceived by different levels of human players through the deep learning model.
  • Method 1 Train with a preset training set
  • Step S2021 The terminal acquires a plurality of training sets for training the second depth model, where the second depth model includes parameters to be initialized, and each training set stores a plurality of game data under a game level level, each of which The game data is used to indicate that in the course of a game, the second account of the first party playing the game with the second party uses the information of the first item in each game round, and the second account is the account in the target game. .
  • the game data of the second account with the same or similar game level level as the first account is obtained, and the game data includes the case of the item used by the second account in one game (eg, which item is specifically used and Using the number of items, the commonality of the accounts at the level of the game (ie, initialization parameters) is mined by learning a large amount of this type of data.
  • Step S2022 The terminal trains the second depth model by using the training set to initialize the parameters in the second depth model, and uses the second depth model obtained after the initialization parameter as the first depth model, and one training set is used for training.
  • a first depth model that mimics the first party of a game level.
  • Method 2 Training with a preset training set
  • the terminal will obtain the behavior data (the data related to the item usage in each round of the game, such as the specific used props and the number of items) as a metadata into the training set, the element The data is annotated with the results of the run.
  • the terminal After obtaining a plurality of metadata, the terminal re-uses the training set for training to update the specific values of the parameters in the depth model, thereby gradually perfecting the depth model.
  • a game round means that both sides of the game have completed the round of the game, first out According to the rules of the game, the latter player also plays the game according to the rules of the game.
  • a game round means that both parties have completed one round of the card (one party plays the card according to the rules of the card, and the other party according to the previous party) The card is followed up, or one party completes the card and the other party gives up the card, or one party completes the card and the other party does not match.
  • the above game can be a complete game (that is, both items are not used yet), or it can be an endgame (there are already items used compared to the complete situation), this application Especially suitable for the case where the game is a mess.
  • Simulating the running of a game of the target game using the first depth model described above can be achieved by the following steps:
  • step S2023 the terminal identifies the item belonging to the first party and the item belonging to the second party in the game, and the second party may be the robot.
  • step S2024 the terminal uses the first item as an input of the first depth model.
  • Step S2025 During the simulation running of the game in one game, each game round is performed according to a predetermined game mode, and the predetermined game mode includes: in the first depth model as the first party, in the target mode to the second game in a game.
  • the terminal allows the second party to respond to the first party using the corresponding second prop according to the first prop used by the first party, or to the first party in the game according to the predetermined manner in the second party.
  • the first depth model is allowed to respond according to the second prop used by the second party, the first prop belongs to the first party, and the second prop belongs to the second party.
  • each game round is performed according to a predetermined game mode, including but not limited to the following three cases:
  • the first depth model as the first party and the second party (such as the black or white player in the game, the main card in the game)
  • the landlords in both sides of the bureau use the props first, and allow the other party in the first party and the second party to use the corresponding props according to the props used by one party, such as the straights used by one of the landlords (eg, 3. 4, 5, 6, 7), then the other party can only produce five straights and the smallest card in the straight should be greater than the smallest card of the other straight (ie greater than 3).
  • the second is: in any game round of a game, if the first party first uses the first item in the second party in one game, and the second party does not respond with the corresponding second item, then Determining that the first party wins in the current game round, and the first party first uses the first item in one game in the next game round, if the second party takes the lead in using the first party in the game If the second item does not use the corresponding first item to respond, it is determined that the second party wins in the current game round, and the second party first uses the second item in one game in the next game round. If one side uses a straight (such as 3, 4, 5, 6, 7) and the other side does not have a straight, the game round ends, and the next round continues to be played by the side of the straight.
  • a straight such as 3, 4, 5, 6, 7
  • the third is: in the last game round of a game, the first item in the first party has been used (it may be the first to use the item or follow the second party to use the item, causing the item to be used), and If the second prop of the second party is not used, it is determined that the first party wins in one game, and the second prop in the second party has been used (may be used first when using the props or following the first party) When the props cause the props to be used, and the first props of the first party are not used up, it is determined that the second party wins in one game, such as a straight used by one party (such as 3, 4, 5, 6) After 7), the card in the hand has already been finished, and the party that has the straight card wins.
  • a straight used by one party such as 3, 4, 5, 6) After 7
  • the first depth model is multiple, Each first depth model is used to use the props according to a target manner corresponding to a game level level, and when the first prop is used as the input of the first depth model, the first depth model corresponding to the game level level of the first account may be selected.
  • the first item is used as the input to the first depth model. That is, it is possible to predict the winning rate of the level level one by one.
  • step S204 the terminal acquires an operation result of performing a simulation operation on a game, and the operation result is used to indicate whether the first party wins in the simulation operation of the game.
  • the terminal determines game information by using a plurality of operation results, and the game information is used to indicate a probability of winning in a game in the case where the first account is the first party, and the plurality of running results are It was obtained by performing a simulation run on a game in succession.
  • An alternative implementation is as follows:
  • Step S2061 For each first account level of the game level, the terminal acquires multiple running results of the corresponding first depth model;
  • step S2062 the terminal uses, as a probability, a ratio of the number of running results used by the first party to indicate that the first party wins in one game and the number of the plurality of running results.
  • Chess game difficulty assessment is a complex task.
  • a difficulty assessor needs to find out the structure and level of the game by trying different gameplays. It takes a lot of time and effort, and the difficulty evaluation values given by different difficulty assessors are very different, and there are great defects.
  • the difficulty of the game has certain subjectivity: the difficulty of the first level player, the middle level player and the high level player to the same game should be different.
  • the current manual evaluation can not describe the difficulty of the game perceived by different levels of players.
  • the technical solution proposed by the present application is based on the game learning difficulty evaluation scheme of deep learning, and simulates different levels of human player behavior through deep learning, and repeatedly performs game simulation to obtain the probability that different levels of human players win opponents in the target game, thereby obtaining different levels of players. Feeling the difficulty.
  • the landlord is a card game, but the landlord's endgame is a clear card, so it can be seen as a board game.
  • the landlord game is a three-party match, one landlord and the other two farmers.
  • the landlord's endgame is usually a landlord and a farmer.
  • the landlord is usually a human player.
  • the farmer is played by a computer robot.
  • the human gamer is the landlord.
  • the design process of the landlord endgame level includes:
  • Step S11 generating a candidate battle landlord situation randomly or according to a certain rule (ie, a game situation of a game in which a game board item of both sides of the game has been determined);
  • Step S12 using the minimax search to find a situation in which the landlord must follow the best method to win, and use it as an endgame;
  • step S13 the difficulty assessment is performed on the endgame using the difficulty assessment technique, and the residuals are assigned to different levels according to the difficulty.
  • the final level interface is shown in Figure 3 (where the 0th level is unlocked and the remaining levels are unlocked).
  • the depth learning-based game difficulty evaluation scheme proposed by the present application includes the following main modules: (1) a human player game record library, which records the game of different levels of players in the board game. Recording; (2) a deep learning training module for training initialization of the original model (second depth model) according to the game record; (3) a board game environment module for performing a simulated game. With these data and modules, a game-level difficulty assessment scheme based on deep learning can be realized.
  • the process of the game difficulty assessment scheme based on deep learning is as follows:
  • the deep learning training module obtains the game of different levels of players from the human player game record database, and uses the deep learning algorithm to train and obtain different levels of models (such as a high level model, a medium level model, and a low level model).
  • models such as a high level model, a medium level model, and a low level model.
  • the second depth model selected a five-layer fully connected deep neural network model, which was trained using a stochastic gradient descent algorithm.
  • the input of the deep learning model is the current landlord situation (all cards), and the current landlord situation may include the player (ie, the first account) hand and the opponent's hand, and the input may be processed into A 30-dimensional vector, the first 15 digits of the vector represent the current player's hand number, and the playing card's points are 3, 4, 5, 6, 7, 8, 9, 10, J, Q, K, A, 2, Xiao Wang And the king, corresponding to 1 to 15 of the 15 hand vectors, the current hand has n points of playing cards, then the hand vector corresponding position value is n. For example, if the current player has 3 3, the first digit of the hand vector is 3.
  • the deep learning model input consists of a 30-dimensional state vector composed of the current player's hand vector and the opponent's hand vector. The last 15 digits of the vector represent the opponent's hand vector. Then, the scheme connects three intermediate layers through a full connection, and the intermediate layer node may be 500. The final output layer has 13,552 nodes, each of which represents a card-out action. Different depth learning models and learning algorithms can and should be used for different board games.
  • step S23 the deep learning model performs multiple game simulations for a situation in the game environment, and generates behavior data and a match win rate.
  • the situation is first entered as the model of the current player, and the probability that the corresponding level of the human player selects the legal action in this situation is output. According to this probability, a legal action is selected as the current player play; then as the model input of the opponent.
  • each circle black solid or hollow indicates a card situation
  • the black circle indicates that the landlord and the peasant are successively selected.
  • the card There are two possibilities for the card to be played for the third time. There are three outcomes after the two possibilities, two of which are the winners of the peasants (in squares) and one of which is the winner of the landlord (indicated by an ellipse).
  • step S24 the simulated generated behavior data is added to the game record library.
  • the simulated winning rate is used as the difficulty index of the game. The higher the simulated winning rate, the lower the difficulty of the game; the lower the simulated winning rate, the higher the difficulty of the game.
  • heuristic rules are used as the player's gameplay. For example, in the landlord game, a commonly used heuristic gameplay "plays as many cards as possible; when the number of cards is the same, a small card is played.” As a result, some levels of the game can be found, making the difficulty assessment more accurate.
  • the main goal is to model the human player gameplay through deep learning.
  • deep learning can model different levels of player play.
  • the first is: having a full game record of different levels of players, this time directly using the supervised learning training deep learning model, supervised learning, by learning the marked data (ie directly marking the difficulty), imitate Labeling behavior;
  • the second is that the player's game record is not enough or insufficient to train the deep learning model.
  • the enhanced learning is to explore in a given environment, through the given target and environmental feedback, to imitate the most valuable behavior in the current environment.
  • the first round of the model is obtained from the zero-data reinforcement learning training.
  • the first round of the model is simulated in the game environment to get the first round of the game record; using the first round of the game record, using the enhanced learning training.
  • the second round model is obtained, and the second round model performs a third round of game matching in the simulated game in the game environment; several rounds of models are obtained by looping back and forth. According to the different win models and the last round model (ie, the strongest model), the models representing different levels are selected.
  • Level layout as shown in Figure 3, it is possible to estimate the difficulty of different endgames, and arrange different endgames into different levels according to the difficulty;
  • a game learning difficulty evaluation method based on deep learning is proposed.
  • deep learning By applying deep learning to simulate the human player gameplay, the human game players of different levels are quickly simulated and the winning rate is obtained, and the simulated winning rate is used as the difficulty index of the game. .
  • this scheme does not need manual labeling, requires less manpower and material resources; can describe the difficulty of different levels of players' feelings, and provides more perspectives on the difficulty of the game; using the deep learning model to improve the accuracy of the difficulty assessment degree.
  • the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, by hardware, but in many cases, the former is A better implementation.
  • the technical solution of the present application which is essential or contributes to the related art, may be embodied in the form of a software product stored in a storage medium (such as ROM/RAM, disk, CD-ROM).
  • the instructions include a number of instructions for causing a terminal device (which may be a cell phone, a computer, a server, or a network device, etc.) to perform the methods described in various embodiments of the present application.
  • FIG. 8 is a schematic diagram of an apparatus for determining game information according to an embodiment of the present application. As shown in FIG. 8, the apparatus may include an operation unit 81, an acquisition unit 83, and a determination unit 85.
  • the running unit 81 is configured to perform a simulation running on a game of the target game.
  • the first party in the game uses the props in a game according to the target mode, and the game is in one game.
  • the second party uses the props in a game in a predetermined manner, and the target mode is the props usage mode of the first account in the target game.
  • the above target game is a game in which the game participants are two parties (ie, the first party and the second party described above), including but not limited to a board game or a card game.
  • the props can be game props used in the corresponding types of games, such as in a board game, the props are pawns; in a card game, the props are cards.
  • the second party in a game uses the props in a game in a predetermined manner, that is, for the second party, the props are used in advance, such as the second party is pre-edited (set). Game logic robot.
  • the props are used to simulate the use of the first account, and the second party uses the props according to the preset game logic to the first party. To respond to the props used by the first party.
  • the above account number can be a specific account or a general term for a type of account.
  • One type of account here refers to a type of account with the same or close game level.
  • the obtaining unit 83 is configured to obtain a running result of performing a simulation running on a game, and the running result is used to indicate whether the first party wins in the simulation running of the game.
  • the above running result is a game result corresponding to the game type, for example, in the case of a board game, when the first party or the second party's pieces satisfy the game rule stipulated victory (if one party eats or knocks the other party's pieces) , the result of the operation is obtained; for the card game, when the card of the first party or the second party satisfies the victory specified by the game rules (if the card is played), the running result is obtained.
  • the determining unit 85 is configured to determine game information by using a plurality of running results, wherein the game information is used to indicate a probability of winning in a game in the case where the first account is the first party, and the plurality of running results are through multiple times. A game of the game was simulated.
  • the operation unit 81 in this embodiment may be used to perform step S202 in the embodiment of the present application.
  • the obtaining unit 83 in this embodiment may be used to perform step S204 in the embodiment of the present application.
  • the determining unit 85 can be used to perform step S206 in the embodiment of the present application.
  • the foregoing modules are the same as the examples and application scenarios implemented by the corresponding steps, but are not limited to the contents disclosed in the foregoing embodiments. It should be noted that the foregoing module may be implemented in a hardware environment as shown in FIG. 1 as part of the device, and may be implemented by software or by hardware.
  • the game running in the game of the target game is simulated.
  • the first party in the game uses the props in a game according to the target mode, and the first in the game.
  • the two parties use the props in a game in a predetermined manner; obtain the running result of the simulation running of the one game; then determine the game information through the plurality of running results, and the game information is used to indicate that the first account is the first party.
  • the probability of winning in the next game the multiple running results are obtained by simulating the running of the game in multiple games, which can solve the technical problem of lower accuracy of the winning rate in the related art. In turn, the technical effect of improving the accuracy of the winning rate of the chess situation is achieved.
  • the running unit simulates a game of the game of the target game, and during the simulation running of the game, the first party in the game uses the props in a game according to the target mode, and the game is in one game.
  • the second party uses the props in a game in a predetermined manner, and the target mode is the props usage mode of the first account in the target game.
  • the difficulty of the game is calculated by the characteristics of the game.
  • the characteristics of the game are selected. For example, for the chess game of Chinese chess, the remaining number of cars, the remaining number of horses, the number of remaining pieces, etc. can be selected as features, and then, the game is played.
  • the feature acts as a feature to be processed, and a mapping between features and difficulty assessments is established using rules or deep learning methods.
  • the simulated winning rate is used as the difficulty index of the game.
  • some heuristic rules are used as a gameplay.
  • the above two methods are combined, and the deep learning model is used to mine the relationship between the various features in the game, and the skills in the game are deeply explored, thereby avoiding the influence of the intuitive steps on the evaluation. , reducing the requirements for human and material resources, and improving the accuracy of the assessment.
  • the deep learning model of different levels of human player gameplay can be characterized, and the deep learning representing different human levels can be performed.
  • the model and the highest level of deep learning model are used to simulate the battle.
  • the simulated winning rate is used as the difficulty of the game played by different levels of human players, which can improve the accuracy of the evaluation.
  • a plurality of deep learning models can be used to mine the chess game processing features of different levels of human players to be able to assess the difficulty perceived by different levels of human players through the deep learning model.
  • Step S2021 Acquire a plurality of training sets for training the second depth model, where the second depth model includes parameters to be initialized, and each training set stores a plurality of game data under a game level level, and each game The data is used to indicate that in the course of a game, the second account of the first party playing the game with the second party uses the information of the first item in each game round, and the second account is the account number in the target game.
  • the game data of the second account with the same or similar game level level as the first account is obtained, and the game data includes the case of the item used by the second account in one game (eg, which item is specifically used and Using the number of items, the commonality of the accounts at the level of the game (ie, initialization parameters) is mined by learning a large amount of this type of data.
  • Step S2022 The second depth model is trained by using the training set to initialize the parameters in the second depth model, and the second depth model obtained after the initialization parameter is used as the first depth model, and one training set is trained for simulation.
  • the above-mentioned game includes at least one game round, wherein the running unit may include: an input module configured to input the first item as the input of the first depth model; and the running module is set to be in a simulation operation of the game in a game,
  • Each game round is performed according to a predetermined game mode, wherein the predetermined game mode includes: when the first depth model as the first party uses the props in a game in a target manner, the second party is allowed to use according to the first party.
  • the first item is used to use the corresponding second item, or when the second party uses the second item in a game in a predetermined manner, the first depth model is allowed to use the corresponding item according to the second item used by the second party.
  • a prop the first prop belongs to the first party, and the second prop belongs to the second party.
  • the first depth model is multiple, and each of the first depth models is used to use the props according to a target manner corresponding to a game level level, wherein the input module is further configured to select a game level with the first account.
  • the first depth model corresponding to the level; the first item is used as the input of the first depth model.
  • the above running module is further configured to perform the following functions:
  • One is: in the first game round of a game, the first depth model as the first party and the second party use the props first, and allow the other party in the first party and the second party Use the corresponding item according to the item used by one party.
  • the second is: in any game round of a game, if the first party uses the first item in a game and the second party does not use the corresponding second item, then the first party is determined to be in the current Winning in the game round, the first party first uses the first item in a game in the next game round, if the second party uses the second item in a game, and the first party does not use the corresponding item An item determines that the second party wins in the current game round, and the second party first uses the second item in a game in the next game round.
  • the third is: in the last game round of a game, in the case where the first item of the first party has been used and the second item of the second party is not used, the first party is determined to play in one game. To win, in the case where the second prop of the second party has been used and the first prop of the first party has not been used, it is determined that the second party wins in one game.
  • the determining unit is further configured to acquire, for the first account of each game level level, a plurality of running results of the corresponding first depth model; and use the plurality of running results to indicate the first party in one round
  • the ratio of the number of running results that are won in the game to the number of running results is taken as the probability.
  • the target game of the present application includes a board game
  • the apparatus may further include: an identification unit configured to recognize that the game belongs to the first party in a game of the board game before performing a simulation run on the game of the target game
  • the chess board and the chess board belonging to the second party wherein the AI robot is the second party, and the props include the chess board.
  • Chess game difficulty assessment is a complex task.
  • a difficulty assessor needs to find out the structure and level of the game by trying different gameplays. It takes a lot of time and effort, and the difficulty evaluation values given by different difficulty assessors are very different, and there are great defects.
  • the difficulty of the game has certain subjectivity: the difficulty of the first level player, the middle level player and the high level player to the same game should be different.
  • the current manual evaluation can not describe the difficulty of the game perceived by different levels of players.
  • the technical solution proposed by the present application is based on the game learning difficulty evaluation scheme of deep learning, and simulates different levels of human player behavior through deep learning, and repeatedly performs game simulation to obtain the probability that different levels of human players win opponents in the target game, thereby obtaining different levels of players. Feeling the difficulty.
  • the foregoing modules are the same as the examples and application scenarios implemented by the corresponding steps, but are not limited to the contents disclosed in the foregoing embodiments. It should be noted that the foregoing module may be implemented in a hardware environment as shown in FIG. 1 as part of the device, and may be implemented by software or by hardware, where the hardware environment includes a network environment.
  • a server or terminal for implementing the above determination method of game information.
  • the terminal may include: one or more (only one shown in FIG. 9) processor 901, memory 903, and transmission device. 905 (such as the transmitting device in the above embodiment), as shown in FIG. 9, the terminal may further include an input/output device 907.
  • the memory 903 can be used to store the software program and the module, such as the method for determining the game information and the program instruction/module corresponding to the device in the embodiment of the present application.
  • the processor 901 runs the software program and the module stored in the memory 903, thereby The various function applications and data processing are performed, that is, the method of determining the game information described above is implemented.
  • Memory 903 can include high speed random access memory, and can also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid state memory.
  • memory 903 can include memory remotely located relative to processor 901, which can be connected to the terminal over a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
  • the transmission device 905 described above is for receiving or transmitting data via a network, and can also be used for data transmission between the processor and the memory.
  • Alternative examples of the above networks may include wired networks and wireless networks.
  • the transmission device 905 includes a Network Interface Controller (NIC) that can be connected to other network devices and routers through a network cable to communicate with the Internet or a local area network.
  • the transmission device 905 is a Radio Frequency (RF) module for communicating with the Internet wirelessly.
  • NIC Network Interface Controller
  • RF Radio Frequency
  • the memory 903 is configured to store an application.
  • the processor 901 can call the application stored in the memory 903 through the transmission device 905 to perform the following steps:
  • the first party in the game uses the props in a game according to the target mode, and the second party in the game is scheduled according to the schedule.
  • the way to use the props in a game the target way is the way the props of the first account in the target game are used;
  • the game information is determined by a plurality of running results, the game information is used to indicate the probability of winning in a game in the case where the first account is the first party, and the plurality of running results are simulated running through the game multiple times. owned.
  • the processor 901 is further configured to perform the following steps:
  • the first depth model as the first party and the second party use the props first, and allow the other party of the first party and the second party to use according to one party. Use the corresponding props;
  • any game round of a game if the first party uses the first item in one game and the second party does not use the corresponding second item, it is determined that the first party obtains in the current game round. Victory, in the next game round, the first party first uses the first item in a game. If the second party uses the second item in a game, and the first party does not use the corresponding first item, then Determining that the second party wins in the current game round, and the second party first uses the second item in one game in the next game round;
  • a game of the game of the target game is simulated, and during the simulation running of the game, the first party in the game uses the props in a game according to the target mode, and the game is in one game.
  • the second party uses the props in a game in a predetermined manner; obtains the running result of the simulation running of the one game; and then determines the game information through the plurality of running results, the game information is used to indicate that the first account is the first party
  • multiple running results are obtained by simulating the running of a game in multiple times, which can solve the low accuracy of the winning rate of the chess game in the related art.
  • the technical problem achieves the technical effect of improving the accuracy of the evaluation of the winning rate of the chess situation.
  • the terminal can be a smart phone (such as an Android mobile phone, an iOS mobile phone, etc.), a tablet computer, a palmtop computer, and a mobile Internet device (Mobile Internet Devices, MID for short). ), PAD and other terminal devices.
  • FIG. 9 does not limit the structure of the above electronic device.
  • the terminal may also include more or fewer components (such as a network interface, display device, etc.) than shown in FIG. 9, or have a different configuration than that shown in FIG.
  • a person of ordinary skill in the art may understand that all or part of the steps of the foregoing embodiments may be completed by a program to instruct terminal device related hardware, and the program may be stored in a computer readable storage medium, and the storage medium may be Including: flash disk, read-only memory (Read-Only Memory, ROM for short), random access memory (Random Access Memory, RAM), disk or optical disk.
  • the storage medium may be Including: flash disk, read-only memory (Read-Only Memory, ROM for short), random access memory (Random Access Memory, RAM), disk or optical disk.
  • Embodiments of the present application also provide a storage medium.
  • the storage medium may be used to execute program code of the method for determining game information.
  • the foregoing storage medium may be located on at least one of the plurality of network devices in the network shown in the foregoing embodiment.
  • the storage medium is arranged to store program code for performing the following steps:
  • the first party in the game uses the props in a game according to the target mode, and the second party in the game.
  • the target mode is the way the props of the first account in the target game are used;
  • S33 Determine game information by using a plurality of running results, where the game information is used to indicate a probability of winning in a game in a case where the first account is the first party, and the plurality of running results are performed by playing the game multiple times.
  • the simulation runs out.
  • the storage medium is further arranged to store program code for performing the following steps:
  • the first depth model as the first party and the second party use the props first, and allow the other party of the first party and the second party to Use the corresponding props for the props used;
  • the foregoing storage medium may include, but is not limited to, a U disk, a ROM, a RAM, a mobile hard disk, a magnetic disk, or an optical disk, and the like, which can store program codes.
  • the integrated unit in the above embodiment if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in the above-described computer readable storage medium.
  • the technical solution of the present application may be embodied in the form of a software product, or the whole or part of the technical solution, which is stored in the storage medium, including
  • the instructions are used to cause one or more computer devices (which may be a personal computer, server or network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application.
  • the disclosed client may be implemented in other manners.
  • the device embodiments described above are merely illustrative.
  • the division of the unit is only a logical function division.
  • multiple units or components may be combined or may be Integrate into another system, or some features can be ignored or not executed.
  • the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, unit or module, and may be electrical or otherwise.
  • the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.
  • each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
  • the above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Computer Security & Cryptography (AREA)
  • General Business, Economics & Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

L'invention concerne un procédé, un dispositif et un milieu de stockage pour déterminer des informations de jeu, ainsi qu'un dispositif électronique. Le procédé comprend les étapes suivantes : un terminal exécute une simulation d'un tour d'un jeu cible, une première partie utilisant un élément d'une manière ciblée pendant le tour dans la simulation d'un tour du jeu, une seconde partie dans le tour utilisant un élément d'une manière prédéterminée pendant le tour du jeu, et la manière ciblée étant la manière dont un premier compte utilise un élément dans le jeu cible ; le terminal acquiert un résultat de simulation consistant à exécuter la simulation du tour du jeu, le résultat indiquant si la première partie a gagné dans la simulation du tour du jeu ; et le terminal détermine des informations de jeu sur la base de multiples résultats de simulation, les informations de jeu indiquant une probabilité qu'a le premier compte de gagner le tour du jeu, et les multiples résultats de simulation étant acquis au moyen de l'exécution de la simulation du tour du jeu plusieurs fois.
PCT/CN2018/111950 2017-10-31 2018-10-25 Procédé, dispositif et support de stockage pour déterminer des informations de jeu, et dispositif électronique WO2019085823A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711051192.9 2017-10-31
CN201711051192.9A CN109718558B (zh) 2017-10-31 2017-10-31 游戏信息的确定方法和装置、存储介质、电子装置

Publications (1)

Publication Number Publication Date
WO2019085823A1 true WO2019085823A1 (fr) 2019-05-09

Family

ID=66294514

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/111950 WO2019085823A1 (fr) 2017-10-31 2018-10-25 Procédé, dispositif et support de stockage pour déterminer des informations de jeu, et dispositif électronique

Country Status (2)

Country Link
CN (1) CN109718558B (fr)
WO (1) WO2019085823A1 (fr)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110180187B (zh) * 2019-06-05 2022-06-14 苏州派维斯信息科技有限公司 面向竞赛的智能对阵选择方法和系统
CN110263937B (zh) * 2019-06-18 2021-09-28 深圳市腾讯网域计算机网络有限公司 一种数据处理方法、设备及存储介质
CN110457534A (zh) * 2019-07-30 2019-11-15 深圳市腾讯网域计算机网络有限公司 一种基于人工智能的数据处理方法、装置、终端及介质
CN110659023B (zh) * 2019-09-11 2020-10-23 腾讯科技（深圳）有限公司 一种程序化内容生成的方法以及相关装置
CN111679879B (zh) * 2020-06-05 2021-09-14 腾讯科技（深圳）有限公司 帐号段位信息的显示方法、装置、终端及可读存储介质
CN112494938B (zh) * 2020-12-07 2024-01-12 北京达佳互联信息技术有限公司 游戏资源分发方法、装置、电子设备及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120252570A1 (en) * 2011-03-29 2012-10-04 Ganz Method for virtual friendship and accessing restricted portions of virtual worlds
CN103838982A (zh) * 2014-03-27 2014-06-04 优视科技有限公司 一种虚拟游戏对象生成方法及装置
CN106339582A (zh) * 2016-08-19 2017-01-18 北京大学深圳研究生院 一种基于机器博弈技术的棋牌类游戏自动化残局生成方法
CN106512407A (zh) * 2016-11-18 2017-03-22 网易（杭州）网络有限公司 匹配处理方法及装置
CN106582025A (zh) * 2016-11-24 2017-04-26 北京暴风魔镜科技有限公司 游戏作品制作系统和方法

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101377798A (zh) * 2008-09-23 2009-03-04 上海久游网络科技有限公司 游戏道具统一处理系统及其处理方法
CN105260171B (zh) * 2015-09-10 2019-01-29 深圳市创梦天地科技有限公司 一种虚拟道具的生成方法及装置
US10201752B2 (en) * 2015-12-14 2019-02-12 Stats Llc System for interactive sports analytics using multi-template alignment and discriminative clustering
CN106201877B (zh) * 2016-07-08 2018-09-14 福州市鼓楼区森林创数文化传播有限公司 一种fps游戏道具的测试方法
CN106919790A (zh) * 2017-02-16 2017-07-04 网易（杭州）网络有限公司 游戏的角色推荐、阵容组建方法和装置、游戏方法和装置
CN107291232A (zh) * 2017-06-20 2017-10-24 深圳市泽科科技有限公司 一种基于深度学习与大数据的体感游戏交互方法及系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120252570A1 (en) * 2011-03-29 2012-10-04 Ganz Method for virtual friendship and accessing restricted portions of virtual worlds
CN103838982A (zh) * 2014-03-27 2014-06-04 优视科技有限公司 一种虚拟游戏对象生成方法及装置
CN106339582A (zh) * 2016-08-19 2017-01-18 北京大学深圳研究生院 一种基于机器博弈技术的棋牌类游戏自动化残局生成方法
CN106512407A (zh) * 2016-11-18 2017-03-22 网易（杭州）网络有限公司 匹配处理方法及装置
CN106582025A (zh) * 2016-11-24 2017-04-26 北京暴风魔镜科技有限公司 游戏作品制作系统和方法

Also Published As

Publication number Publication date
CN109718558A (zh) 2019-05-07
CN109718558B (zh) 2021-05-11

Similar Documents

Publication Publication Date Title
WO2019085823A1 (fr) Procédé, dispositif et support de stockage pour déterminer des informations de jeu, et dispositif électronique
CN107970608B (zh) 关卡游戏的设置方法和装置、存储介质、电子装置
CN106390456B (zh) 游戏中角色行为的生成方法和装置
CN109499068B (zh) 对象的控制方法和装置、存储介质、电子装置
CN109814955B (zh) 阵容确定的方法、装置及电子设备
CN109513215B (zh) 一种对象匹配方法、模型训练方法以及服务器
US11325048B2 (en) User experience modeling for gaming applications
CN113660987A (zh) 生成适用于游戏应用程序的游戏机器人
CN111569429B (zh) 模型训练方法、模型使用方法、计算机设备及存储介质
CN111738294B (zh) Ai模型的训练方法、使用方法、计算机设备及存储介质
CN108776944B (zh) 一种用于网络竞赛式学习的数据处理系统和方法
Holmgård et al. Personas versus clones for player decision modeling
CN112704882B (zh) 基于模型的棋牌游戏策略更新的方法、系统、介质及装置
CN111506514B (zh) 一种应用于消除类游戏的智能测试方法及系统
Muñoz-Avila et al. Learning and game AI
US7980931B1 (en) Social networking methods using color pair personality profiles
CN110263937B (zh) 一种数据处理方法、设备及存储介质
US11596867B2 (en) AI-based content generation for gaming applications
CN115944921A (zh) 游戏数据处理方法、装置、设备及介质
Chen et al. Challenging artificial intelligence with multiopponent and multimovement prediction for the card game Big2
CN114870403A (zh) 一种游戏中的对战匹配方法、装置、设备及存储介质
EP4100139A1 (fr) Génération de contenu basée sur l'ia pour des applications de jeu
CN114681924A (zh) 虚拟对象的推荐方法、装置和电子设备
Gugerell et al. Daredevil or socialiser?: exploring the relations between intrinsic motivation, game experience and player types in serious games with environmental narratives
KR102054261B1 (ko) 인공지능의 데이터분류방법 체득용 보드게임 프로그램이 기록된 컴퓨터 판독이 가능한 기록매체 및 그 시스템과 이를 이용하는 보드게임방법

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 18874082

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18874082

Country of ref document: EP

Kind code of ref document: A1