Game blind box extraction method, system and equipment based on player data and medium

Abstract

The invention is applicable to the field of game blind box extraction, and discloses a game blind box extraction method based on player data, which comprises the following steps: comprehensively analyzing historical game data of players, setting different time window lengths for different players to capture long-term game preference and short-term interest changes of the players, and adjusting the blind box extraction probability according to the changes; game behavior data of players in different time periods are recognized, false interests are eliminated, and improper influences on the extraction probability are avoided; for a plurality of accounts of the same player, the method combines the game data of the accounts for unified processing, and ensures the accuracy of extraction probability adjustment; a player satisfaction degree and game balance degree evaluation model is further constructed, and the blind box extraction probability is further finely adjusted; the method also has a monitoring function, can identify and prevent the abnormal behavior that the player operates the extraction probability by using the system vulnerability, monitors the blind box extraction condition, evaluates the influence of the adjusted probability distribution on the game economy, and ensures the fairness and stability of the game operation.

Description

Technical Field

[0001] The present invention relates to the field of game blind box extraction, and in particular to a game blind box extraction method, system, device and medium based on player data. Background Art

[0002] Game blind boxes are a popular commodity sales model, especially loved by young consumers. Consumers cannot predict the specific goods in the box when purchasing. This purchasing method brings a sense of mystery and expectation, which increases the fun of the consumption experience; game blind boxes usually contain small toys, character models or other commodities with collectible value. Consumers are curious about the goods inside before opening the game blind box.

[0003] However, traditional blind box extraction methods often adopt a fixed probability distribution, ignoring the personalized needs and changes in interests of players, resulting in poor player experience and decreased game participation.

[0004] In addition, some players may take advantage of system loopholes to perform improper operations and manipulate the probability distribution of blind box extraction, which not only undermines the fairness of the game, but also affects the stability of the game economy. Therefore, how to effectively identify and prevent such abnormal behaviors has become an urgent problem to be solved in the current game blind box extraction method. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes a game blind box extraction method, system, device and medium based on player data.

[0006] In the first aspect, the present application proposes a method for extracting blind boxes of games based on player data, including: obtaining the player's historical game data, setting different time window lengths for different players, obtaining the player's long-term game preferences and short-term game interest changes, and adjusting the probability of extracting blind boxes based on the player's long-term game preferences and short-term game interest changes;

[0007] Obtain the game behavior data of players in different time periods and identify the false interests of players. If the player has false interests, the probability of his / her blind box drawing will not be adjusted;

[0008] In the process of adjusting the probability of drawing the blind box, multiple accounts of the same player are identified, the game data of the associated accounts are merged, and the probability of drawing the blind box is adjusted according to the merged game data; or a player satisfaction evaluation model and a game balance evaluation index are constructed, and the probability of drawing the blind box is adjusted based on the player satisfaction evaluation model and the game balance evaluation index;

[0009] Monitor the blind box drawing behavior of the players and identify abnormal behaviors that exploit system loopholes to manipulate the probability distribution of blind box drawing;

[0010] After adjusting the probability of drawing the blind box, monitor the drawing of the blind box to determine whether the adjusted probability distribution will have an impact on the game economy.

[0011] More specifically, in the above technical solution, the historical game data of the player is obtained, and different time window lengths are set for different players to obtain the player's long-term game preferences and short-term game interest changes, and the probability of drawing the blind box is adjusted based on the player's long-term game preferences and short-term game interest changes, including:

[0012] Obtain the player's historical game data, which includes the player's game behavior, game duration, game frequency, and consumption records;

[0013] According to the historical game data, the players are divided into different player groups, each of the player groups having similar game preferences and behavior patterns;

[0014] Determine the length of the time window based on the historical game data of each player group, and predict the players' long-term game preferences and short-term interest changes for different blind boxes based on the game behaviors and preferences within the time window;

[0015] Adjust the probability distribution of drawing each blind box based on the long-term gaming preferences and short-term interest changes.

[0016] More specifically, in the above technical solution, the game behavior data of the player in different time periods is obtained, and the false interest of the player is identified. If the player has false interest, the probability of drawing the blind box is not adjusted, including:

[0017] Counting the frequency of the players drawing each type of blind box within a preset time period, and obtaining the players' preference distribution for different types of blind boxes based on the frequency;

[0018] Compare the player's preference distribution for blind boxes with his / her long-term game preferences. If there is a significant deviation, it is determined that the player's interest in blind boxes is abnormal;

[0019] Use anomaly detection algorithms such as Isolation Forest to detect anomalies in the blind box extraction behavior of players with abnormal interests and identify players who show false interests;

[0020] The false interest of the player is quantified, the false interest value of the player is obtained, and a corresponding false interest threshold is set; when the false interest value exceeds the false interest threshold, the player is marked as a false player, and when the blind box probability is adjusted, the blind box drawing probability of the false player is not adjusted.

[0021] More specifically, in the above technical solution, a player satisfaction evaluation model and a game balance evaluation index are constructed, and based on the player satisfaction evaluation model and the game balance evaluation index, the blind box extraction probability is adjusted, including:

[0022] Based on the historical game data, a player satisfaction evaluation model is constructed, wherein the player satisfaction evaluation model is used to quantify the player satisfaction under different extraction probabilities;

[0023] Based on the game balance, construct a game balance evaluation index, wherein the game balance evaluation index includes the player's sense of gain and the balance of game resource allocation;

[0024] Taking player satisfaction and game balance as optimization targets of the multi-objective optimization algorithm, and integrating the two optimization targets into a single optimization function through weighted summation;

[0025] Using a heuristic search method to obtain a combination of extraction probabilities that strikes a balance between the player satisfaction and the game balance;

[0026] Optimizing the extraction probability combination, generating a new extraction probability combination through a crossover mutation operation, and evaluating its quality according to a fitness function;

[0027] Adjust the blind box drawing probability according to the optimized drawing probability combination.

[0028] More specifically, in the above technical solution, monitoring the blind box drawing behavior of the player and identifying abnormal behaviors that manipulate the probability distribution of blind box drawing by exploiting system loopholes include:

[0029] Obtaining the player's behavior data during the blind box drawing process, the behavior data including the drawing time, the number of draws, and the information of the items obtained;

[0030] Based on the behavior data, extract the characteristics of the frequency of extraction, the number of consecutive extractions and the probability of obtaining rare items, and construct a player behavior pattern feature vector;

[0031] Dividing the players into different behavior pattern groups according to the player behavior pattern feature vectors;

[0032] Based on each of the behavior pattern groups, the probability distribution of blind box extraction is counted to obtain the extraction probability distribution characteristics of each group;

[0033] Comparing the extraction probability distribution characteristics of each group with a preset normal probability distribution, and calculating the deviation;

[0034] If the deviation exceeds a preset threshold, it is determined that the group has abnormal behavior;

[0035] For groups of players who are judged to have abnormal behavior, analyze their behavior sequences and identify the blind box extraction algorithm loopholes they exploit;

[0036] According to the identified blind box extraction algorithm vulnerability, patch repair measures are taken, including improving the server-side verification logic and increasing the random number seed update frequency;

[0037] Players who exploit loopholes in the blind box extraction algorithm will be warned or banned.

[0038] More specifically, in the above technical solution, in the process of adjusting the probability of drawing the blind box, multiple accounts of the same player are identified, the game data of the associated accounts are merged, and the probability of drawing the blind box is adjusted according to the merged game data, including:

[0039] Through account association analysis technology, multiple account information related to players is extracted from the game database;

[0040] Use data merging method to aggregate game data of multiple accounts into one main account;

[0041] Extracting the player's real game preference based on the merged game data through data analysis;

[0042] The drawing probability is adjusted according to the actual game preference.

[0043] More specifically, in the above technical solution, after adjusting the probability of drawing the blind box, monitoring the drawing of the blind box to determine whether the adjusted probability distribution will have an impact on the game economy includes:

[0044] Obtaining real-time data of blind box draws, including the number of draws and draw results;

[0045] According to the real-time data, the actual extraction frequency of each type of item is obtained;

[0046] Obtain the preset probability distribution model and calculate the theoretical extraction probability of each type of item;

[0047] Based on each category of items, the actual extraction frequency is divided by the total number of extractions to obtain the actual extraction probability of the item in that category;

[0048] Calculating the deviation between the actual extraction probability and the theoretical extraction probability;

[0049] If the deviation exceeds a preset threshold, a probability adjustment mechanism is triggered to adjust the theoretical extraction probability by increasing or decreasing the weight according to the deviation to obtain an adjusted probability distribution;

[0050] According to the adjusted probability distribution, by randomly generating a preset number of virtual drawing processes, the simulated drawing frequency and simulated drawing probability of each type of item are counted;

[0051] Comparing the simulated draw probability with the actual draw probability before adjustment, and evaluating the impact of the adjusted probability distribution on item rarity, player experience, and game economy;

[0052] Obtaining preset game indicators, including total spending amount, number of paying players, average spending amount per person, and number of daily active users;

[0053] Establishing a correlation model between the adjusted probability distribution and the game index, predicting the impact of the probability adjustment on each index, and obtaining a prediction result;

[0054] If there are indicators in the forecast results whose negative impact exceeds the preset acceptable threshold, the probability adjustment will be cancelled;

[0055] If there is no indicator in the prediction result whose negative impact exceeds the preset acceptable threshold, the adjusted probability distribution parameters are saved in the database.

[0056] In the second aspect, the present application also proposes a game blind box extraction system based on player data, including:

[0057] The first processing module: obtains the player's historical game data, sets different time window lengths for different players, obtains the player's long-term game preferences and short-term game interest changes, and adjusts the probability of drawing blind boxes based on the player's long-term game preferences and short-term game interest changes;

[0058] The second processing module: obtains the game behavior data of players in different time periods, and identifies the false interests of players. If the player has false interests, the probability of drawing the blind box will not be adjusted;

[0059] The third processing module: in the process of adjusting the probability of drawing the blind box, multiple accounts of the same player are identified, the game data of the associated accounts are merged, and the probability of drawing the blind box is adjusted according to the merged game data; or a player satisfaction evaluation model and a game balance evaluation index are constructed, and the probability of drawing the blind box is adjusted based on the player satisfaction evaluation model and the game balance evaluation index;

[0060] The fourth processing module: monitors the blind box drawing behavior of the player and identifies abnormal behaviors that use system loopholes to manipulate the probability distribution of blind box drawing;

[0061] The fifth processing module: after adjusting the probability of drawing the blind box, monitor the drawing situation of the blind box to determine whether the adjusted probability distribution will have an impact on the game economy.

[0062] In a third aspect, the present application further proposes a terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements any of the methods described above when executing the computer program.

[0063] In a fourth aspect, the present application further proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method as described in any one of the above items is implemented.

[0064] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0065] 1. The present invention can set different time window lengths for different players based on the players' historical game data, so as to accurately capture each player's long-term game preferences and short-term interest changes, so that the probability of blind box extraction is no longer fixed, but can be dynamically adjusted as the player's interests change, thereby greatly improving the player's gaming experience. Through personalized extraction probability, the present invention can meet the personalized needs of different players, so that each player can feel a sense of surprise that matches his or her interests when extracting a blind box, thereby increasing the player's game participation and loyalty.

[0066] 2. The present invention has the ability to identify players' false interests. By counting the frequency of players drawing various types of blind boxes within a preset time period and comparing them with the players' long-term game preferences, it can be identified whether the players have false interests. If the player's interest in the blind box is abnormal, the probability of drawing will not be adjusted, thereby avoiding resource mismatches caused by false interests. Preventing false interests not only protects the game's economic system from interference from improper operations, but also ensures the fairness and justice of blind box drawing, and maintains the overall order of the game and trust between players.

[0067] 3. The present invention constructs a player satisfaction evaluation model and a game balance evaluation index to quantify player satisfaction and the balance of game resources under different extraction probabilities. These two evaluation models provide a scientific basis for adjusting the probability of blind box extraction, ensuring that blind box extraction can meet the needs of players without destroying the balance of the game. Through satisfaction and balance evaluation, the present invention can continuously optimize the blind box extraction strategy, so that the game can maintain its fun while also having better sustainability and fairness.

[0068] 4. The present invention adopts a multi-objective optimization algorithm, takes player satisfaction and game balance as optimization goals, and integrates these two optimization goals into a single optimization function through weighted summation, so that when adjusting the probability of blind box extraction, the needs of players and the balance of the game can be taken into account, thereby achieving a global optimal solution. The application of the multi-objective optimization algorithm enables the present invention to more accurately control the probability distribution of blind box extraction, thereby achieving better balance and stability in the game economic system. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0070] Figure 1 It is a flowchart of a method for extracting a game blind box based on player data provided by an embodiment of the present invention;

[0071] Figure 2 It is a structural schematic diagram of a game blind box extraction system based on player data provided by one embodiment of the present invention;

[0072] Figure 3 It is a structural diagram of the terminal device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0073] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0074] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0075] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0076] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0077] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0078] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0079] See also Figure 1 The present application proposes a method for extracting blind boxes of games based on player data, including: S101 obtaining the player's historical game data, setting different time window lengths for different players, obtaining the player's long-term game preferences and short-term game interest changes, and adjusting the probability of extracting blind boxes based on the player's long-term game preferences and short-term game interest changes; S102 obtaining the player's game behavior data in different time periods, and identifying the player's false interests; if the player has false interests, the probability of extracting blind boxes will not be adjusted; S103 during the process of adjusting the probability of extracting blind boxes, identifying multiple players of the same player; For each account, the game data of the associated account is merged, and the probability of drawing the blind box is adjusted according to the merged game data; or a player satisfaction evaluation model and a game balance evaluation index are constructed, and the probability of drawing the blind box is adjusted based on the player satisfaction evaluation model and the game balance evaluation index. S104 monitors the blind box drawing behavior of the player, identifies abnormal behavior that uses system loopholes to manipulate the probability distribution of drawing blind boxes, and S105 adjusts the probability of drawing the blind box, monitors the drawing of the blind box, and determines whether the adjusted probability distribution will have an impact on the game economy. Including:

[0080] Obtain the player's historical game data, and set different time window lengths for different players, obtain the player's long-term game preferences and short-term game interest changes, and adjust the blind box extraction probability based on the player's long-term game preferences and short-term game interest changes; obtain the player's game behavior data in different time periods, and identify the player's false interests. If the player has false interests, do not adjust the blind box extraction probability; in the process of adjusting the blind box extraction probability, identify multiple accounts of the same player, merge the game data of the associated accounts, and adjust the blind box extraction probability according to the merged game data; or construct a player satisfaction evaluation model and a game balance evaluation index, and adjust the blind box extraction probability based on the player satisfaction evaluation model and the game balance evaluation index; monitor the player's blind box extraction behavior, and identify abnormal behavior that uses system loopholes to manipulate the blind box extraction probability distribution; after adjusting the blind box extraction probability, monitor the blind box extraction situation to determine whether the adjusted probability distribution will have an impact on the game economy.

[0081] Specifically, obtaining players' historical game data can provide a deep understanding of their game behaviors and preferences. To this end, different time window lengths need to be set. For example, the short-term window can be set to 7 days or 30 days to capture the player's recent interest changes; the long-term window can be set to 90 days or 180 days, or even longer, to reveal the player's stable preferences. For example, player A's short-term data shows that he has played racing games frequently recently, while long-term data shows that he has always preferred role-playing games. This shows that player A may have a short-term interest in racing games, but role-playing games are still his main preference. Based on these data, the probability of drawing blind boxes can be adjusted, such as increasing the probability of racing game-related props while retaining a certain proportion of role-playing game props to meet player A's long-term and short-term needs. In order to more accurately capture the player's true interests, it is necessary to identify the player's false interests. For example, in order to complete a time-limited task, player B has to play a strategy game that he does not like. In this case, if the probability of strategy game props is increased based only on short-term data, it may cause player B to be dissatisfied. Therefore, it is necessary to combine multi-dimensional data such as the player's game time, payment status, and social interaction to judge the player's true interests. If the system identifies that player B's strategy game behavior is a false interest, the probability of his blind box extraction will not be adjusted, and his original preference probability distribution will be maintained. In the process of adjusting the probability of blind box extraction, the problem of multiple accounts of the same player also needs to be considered. Some players may have multiple game accounts. For example, player C has both an iOS account and an Android account. If the data of these two accounts are processed separately, it may lead to a deviation in the understanding of player C's preferences. Therefore, it is necessary to identify and merge the game data of the associated accounts. For example, through information such as IP address, device ID, and game behavior, it can be determined that the two accounts of player C belong to the same person. After merging the data, the overall game preference of player C can be more accurately understood, and the probability of blind box extraction can be adjusted accordingly. In addition to adjusting the probability of blind box extraction based on the player's game data, a player satisfaction evaluation model and a game balance evaluation index can also be constructed. The player satisfaction evaluation model can evaluate the player's satisfaction with the current blind box probability distribution by collecting player feedback, evaluation, game time and other data. The game balance evaluation index can measure the value and scarcity of different props and their impact on the game economic system. For example, if an item is too rare or powerful, it may undermine the balance of the game and cause players to leave. Therefore, it is necessary to dynamically adjust the probability of blind box extraction based on player satisfaction and game balance. For example, if players are generally dissatisfied with the probability of obtaining a certain type of item, or the frequency of a certain item is too high, affecting the balance of the game, the corresponding probability needs to be adjusted. In addition, it is necessary to monitor the player's blind box extraction behavior and identify abnormal behavior. For example, some players may exploit system loopholes to manipulate the probability distribution of blind box extraction, or use plug-in programs to cheat. These behaviors will undermine the fairness and stability of the game.Therefore, it is necessary to establish an abnormal behavior detection mechanism. For example, by analyzing the player's extraction frequency, amount, time and other data, abnormal behavior can be identified and corresponding measures can be taken, such as banning accounts and rolling back data. Finally, after adjusting the probability of blind box extraction, it is also necessary to monitor the extraction of blind boxes to determine whether the adjusted probability distribution will have an impact on the game economy. For example, if the extraction probability of certain props is too high, the price of these props may fall, affecting the stability of the in-game economic system. Therefore, it is necessary to continuously monitor various indicators of the in-game economic system, such as prop prices, transaction volume, player payment, etc., and fine-tune the probability of blind box extraction based on the monitoring results to ensure the healthy development of the game economic system.

[0082] In some embodiments, the historical game data of players is obtained, and different time window lengths are set for different players to obtain the long-term game preferences and short-term game interest changes of players, and the probability of drawing blind boxes is adjusted based on the long-term game preferences and short-term game interest changes of the players. Including:

[0083] Obtain the player's historical game data, which includes the player's gaming behavior, gaming duration, gaming frequency, and consumption record; divide the players into different player groups based on the historical game data, each of the player groups having similar gaming preferences and behavior patterns; determine the length of the time window based on the historical game data of each of the player groups, and predict the player's long-term gaming preference and short-term interest changes for different blind boxes based on the gaming behavior and preference within the time window; adjust the probability distribution of drawing each blind box based on the long-term gaming preference and short-term interest changes.

[0084] Specifically, the historical game data of players is obtained, including the player's game behavior, game time, game frequency and consumption record. For example, the historical game data of player A shows that his game behavior is mainly concentrated in PVE copies, the average game time is 2 hours a day, the game frequency is 5 times a week, and he has never made any consumption in the game. The historical game data of player B shows that he mainly engages in PVP competition, with an average of 5 hours online per day, a game frequency of 1 time a day, and a small amount of recharge every month to purchase decorative props in the game. Player C's game behavior is relatively average, involving both PVE and PVP, with a daily game time of about 3 hours, a frequency of 4 times a week, and frequent purchases of blind boxes. Collecting this data can provide a basis for subsequent player grouping and probability adjustment. According to historical game data, players are divided into different groups, each of which has similar game preferences and behavior patterns. For example, based on the game behavior of the above three players, player A can be classified as a PVE casual player group, player B can be classified as a PVP competitive player group, and player C can be classified as a PVE / PVP mixed player group. Dividing player groups can more finely adjust the probability of blind boxes according to the needs of different player groups. For example, casual PVE players may pay more attention to collection and cultivation items, while PVP competitive players may prefer items that can enhance combat effectiveness. Determine the length of the time window based on the historical game data of each player group. The length of the time window is set to capture the short-term changes in players' interests. For casual PVE players, a longer time window, such as one month, can be set, because their game preferences are usually more stable. For PVP competitive players and mixed players, a shorter time window, such as one or two weeks, can be set to more quickly capture their changes in interest in different blind boxes. For example, for mixed players such as player C, who have focused more on PVP gameplay in the past two weeks, the probability of rare items in PVP-related blind boxes can be appropriately increased. Within a determined time window, analyze the player's game behavior and preferences, and predict the player's long-term game preferences and short-term interest changes for different blind boxes. For example, the PVE casual player group has long preferred collection items, but in the past week has shown a high interest in a newly launched PVE copy-related blind box. Then it can be predicted that the group's demand for this new blind box will increase in the short term. For the PVP competitive player group, they have long preferred props that enhance combat effectiveness, and in the past two weeks have shown interest in a cool-looking PVP mount blind box. It can be predicted that the group's demand for the mount blind box will increase in the short term. Adjust the probability distribution of each blind box according to long-term game preferences and short-term interest changes. For PVE casual players, since they have long preferred collection items, when adjusting the blind box probability, the probability of blind boxes containing rare collection items can be appropriately increased. At the same time, considering their recent interest in blind boxes related to new PVE copies, the probability of rare items in the blind box can be further increased.For PVP competitive players, the probability of blind boxes containing rare combat power-enhancing props can be appropriately increased. At the same time, considering their recent interest in cool PVP mount blind boxes, the probability of rare mounts in the blind box can also be increased. In this way, the needs of different player groups can be better met, and player satisfaction and gaming experience can be improved. For example, player C has focused more on PVP gameplay in the past two weeks, and the probability of rare props in PVP-related blind boxes can be increased. If player C recharges more in the game, the probability of obtaining rare props can be further increased based on his consumption record to increase his willingness to pay. Doing so can enable different types of players to obtain a gaming experience that meets their expectations, thereby increasing the overall activity and revenue of the game.

[0085] In some embodiments, the game behavior data of players in different time periods are obtained, and the false interests of players are identified. If the player has false interests, the probability of drawing the blind box is not adjusted. Including:

[0086] Count the frequency of the player's drawing of each type of blind box within a preset time period, and obtain the player's preference distribution for different types of blind boxes based on the frequency; compare the player's preference distribution for blind boxes with his long-term game preference, and if there is a significant deviation, determine that the player's interest in blind boxes is abnormal; use anomaly detection algorithms such as isolation forest to perform anomaly detection on the blind box drawing behavior of players with abnormal interests, and identify players who show false interests; quantify the player's false interests, obtain the player's false interest value, and set a corresponding false interest threshold; when the false interest value exceeds the false interest threshold, mark the player as a false player, and when adjusting the blind box probability, do not adjust the blind box drawing probability of the false player.

[0087] Specifically, players draw blind boxes in the game in order to obtain the desired virtual items. Only by understanding the real interests of players can we better operate the game and improve the player experience. However, some players may have false interests, such as pretending to be interested in certain blind boxes for some improper benefits. In order to distinguish these players, it is necessary to analyze the players' blind box extraction behavior. First, it is necessary to count the frequency of players drawing various blind boxes within a certain period of time (such as one month). Suppose there are three types of blind boxes in the game: A (weapons), B (clothing), and C (props). Player A drew Class A blind boxes 50 times, Class B blind boxes 10 times, and Class C blind boxes 5 times this month. This shows that Player A's preference for Class A blind boxes (weapons) is much higher than the other two categories. This is the distribution of players' preferences for blind boxes. Then, it is necessary to compare the distribution of players' preferences for blind boxes with their long-term game preferences. Long-term game preferences can be obtained from players' historical game data, including game behavior, game duration, game frequency, and consumption records. Suppose that the historical game data of player A shows that he is a player who focuses on combat, often participates in PVP activities, and spends more on weapons. This is consistent with his preference for Class A blind boxes (weapons). However, if the historical game data of another player B shows that he is a player who focuses on socializing, mainly participates in leisure activities, rarely participates in combat, and spends more on clothing. However, player B frequently draws Class A blind boxes (weapons) this month, far exceeding the number of times he draws Class B blind boxes (clothing), which is a clear deviation. This indicates that player B's interest in blind boxes may be abnormal. In order to further confirm, it is necessary to use anomaly detection algorithms, such as the isolation forest algorithm, to perform anomaly detection on the blind box extraction behavior of players with abnormal interests. The core idea of ​​the isolation forest algorithm is that abnormal data is easier to be isolated. If a player's blind box extraction behavior is very different from the behavior pattern of most players, then he is more likely to be isolated and is more likely to be abnormal behavior. Suppose that through the calculation of the isolation forest algorithm, player B's anomaly score is very high, exceeding the preset threshold. Next, the player's false interest needs to be quantified. A false interest value can be calculated based on factors such as abnormal scores and degree of deviation. Assume that the false interest value of player B is 0.8, and the false interest threshold is set to 0.7. Since the false interest value of player B exceeds the threshold, he is marked as a false player. Finally, when adjusting the probability of blind boxes, the data of these players marked as false will be ignored. In other words, the probability of blind box extraction will not be adjusted based on the extraction behavior of player B. Doing so can prevent the data of false players from interfering with the normal probability adjustment, thereby ensuring the fairness and healthy development of the game. Doing so can also prevent game operations from being misled by false player data and making wrong decisions.

[0088] In some embodiments, during the process of adjusting the probability of drawing the blind box, multiple accounts of the same player are identified, the game data of the associated accounts are merged, and the probability of drawing the blind box is adjusted according to the merged game data.

[0089] Through account association analysis technology, multiple account information related to the player is extracted from the game database; using a data merging method, the game data of multiple accounts are aggregated under a main account; through data analysis, the real game preference of the player is extracted based on the merged game data; and the extraction probability is adjusted according to the real game preference.

[0090] Specifically, account association analysis technology can help identify players' secondary accounts, so as to more accurately understand players' real game preferences. For example, it is possible to determine whether multiple accounts belong to the same player based on information such as IP address, device ID, and registered email address. For example, player A uses three accounts: account 1 is mainly used for PVE gameplay, account 2 is mainly used for PVP gameplay, and account 3 is mainly used for collection and cultivation gameplay. Through account association analysis, these three accounts can be associated with player A, so that it can be known that player A's real game preference is not a single PVE, PVP or collection and cultivation, but is interested in all three gameplays. This association analysis can avoid drawing one-sided conclusions based on the data of a single account and understand the player's game behavior more comprehensively. The data merging method can aggregate the game data of associated accounts into the main account for unified analysis. In the above example, the data of account 1, account 2, and account 3 can be merged into the main account of player A. The aggregated data includes the game time of PVE gameplay, the winning rate of PVP gameplay, the completion of collection and cultivation, etc. These data together constitute a complete behavioral portrait of player A, which can more accurately reflect player A's real game preferences. Data merging can avoid the problem of data fragmentation and provide a more complete data basis for subsequent analysis. Data analysis is a key step in extracting players' real game preferences based on the merged game data. For example, by analyzing the merged game data of player A, it is found that player A has invested the most time in PVE gameplay and completed the high-difficulty dungeon challenge; has a high win rate and ranks high in PVP gameplay; and has also collected a large number of rare props in the collection and development gameplay. These data show that player A has a high degree of participation and game level in these three gameplays, and his real game preferences are relatively balanced. Through data analysis, the player's various data can be converted into valuable information to reveal the player's real game preferences. After extracting the player's real game preferences, the probability of blind box extraction can be adjusted according to his preferences. For example, suppose there are three types of blind boxes in the game: PVE props blind box, PVP props blind box, and collection and development props blind box. If only based on the data of player A's account 1, it may be believed that player A only likes PVE gameplay, thereby increasing the probability of extracting PVE props blind boxes. However, through account association analysis and data merging, it is known that player A is interested in all three gameplays. Therefore, the probability of drawing the three blind boxes can be adjusted according to the player A's level of involvement and game level in the three modes of play. For example, the probability of a blind box for PVE props is 40%, the probability of a blind box for PVP props is 35%, and the probability of a blind box for collection and cultivation props is 25%. This adjustment method can better meet the needs of players, improve players' gaming experience and satisfaction, and effectively avoid resource mismatches caused by misjudgment of player preferences. More importantly, this adjustment based on real game preferences can encourage players to try more types of games, enrich players' gaming experience, and improve the overall activity of the game.Taking player B as an example, suppose that player B's account 4 is identified as a fake account. This account mainly conducts a large number of blind box extractions and focuses on a specific type of blind box. Through account association analysis, it is found that this account is associated with other accounts of player B (account 5 and account 6). However, the game behavior data of account 5 and account 6 show that player B does not show any particular preference for this particular type of game content. In this case, it can be determined that player B's account 4 has a false interest. When adjusting the blind box probability, no special adjustment will be made for account 4, but adjustments will be made based on the data of account 5 and account 6 to avoid negative impact on player B's overall gaming experience. Doing so can effectively identify and respond to the interference of fake accounts and maintain the fairness and healthy development of the game.

[0091] In some embodiments, the blind box drawing behavior of the player is monitored to identify abnormal behaviors that exploit system loopholes to manipulate the probability distribution of blind box drawing. This includes:

[0092] Obtain the behavioral data of players during the blind box drawing process, the behavioral data including the drawing time, the number of drawings and the information of the items obtained; preprocess the behavioral data to obtain the preprocessed behavioral data, the preprocessing including removing abnormal data and filling missing values; based on the preprocessed behavioral data, extract the key features of the drawing frequency, the number of consecutive drawings and the probability of obtaining rare items, and construct the player behavior pattern feature vector; cluster the player behavior pattern feature vector using the K-means clustering algorithm to obtain the player behavior pattern clustering result; divide the players into different behavior pattern groups according to the player behavior pattern clustering result; for For each of the behavior pattern groups, its blind box extraction probability distribution is counted to obtain the extraction probability distribution characteristics of each group; the extraction probability distribution characteristics of each group are compared with the preset normal probability distribution, and the deviation is calculated; if the deviation exceeds the preset threshold, the group is judged to have abnormal behavior; for the player group judged to have abnormal behavior, its behavior sequence is analyzed to identify the blind box extraction algorithm loopholes that are exploited; for the identified blind box extraction algorithm loopholes, patch repair measures are taken, and the patch repair measures include improving the server-side verification logic and increasing the random number seed update frequency; player accounts that exploit loopholes to cheat are warned or banned.

[0093] Specifically, obtain the behavioral data of players during the blind box extraction process. For example, player A draws once at 10:01 on March 1, 2024, and obtains the common item "Magic Wand"; draws again at 10:02 on the same day and obtains the rare item "Elf Wings". These data include the extraction time, the number of extractions, and the specific item information obtained. These data are recorded for subsequent analysis of player behavior patterns. Preprocess these behavioral data. For example, a certain extraction record of player B is lost due to network fluctuations, and the item information is missing. It needs to be filled with reasonable values, such as the item information obtained in his most recent extraction. If player C performs a large number of extractions in a short period of time, far exceeding the behavior of normal players, this may be abnormal data and needs to be eliminated. Data preprocessing is to ensure the accuracy of subsequent analysis results. Extract key features based on preprocessed behavioral data. For example, player A draws 5 times a day on average, the highest number of consecutive draws is 3 times, and the probability of obtaining rare items is 10%; player B draws 100 times a day on average, the highest number of consecutive draws is 50 times, and the probability of obtaining rare items is 50%. These key features include the frequency of draws, the number of consecutive draws, and the probability of obtaining rare items. These features are combined into a player behavior pattern feature vector. For example, the feature vector of player A is (5, 3, 0.1), and the feature vector of player B is (100, 50, 0.5). These features are extracted to quantify the player's behavior pattern for subsequent clustering analysis. The K-means clustering algorithm is used to cluster the player behavior pattern feature vector. For example, the feature vectors of all players are input into the K-means algorithm, and the number of clusters is set to 3. The algorithm will divide the players into 3 different groups according to the similarity of the feature vectors. The purpose of clustering is to group players with similar behavior patterns together for group analysis. According to the clustering results, the players are divided into different behavior pattern groups. For example, the first group is "casual players", whose characteristics are low extraction frequency, few consecutive extractions, and low probability of obtaining rare items; the second group is "active players", whose characteristics are moderate extraction frequency, medium consecutive extraction times, and moderate probability of obtaining rare items; the third group is "heavy players", whose characteristics are high extraction frequency, many consecutive extractions, and high probability of obtaining rare items. The purpose of dividing groups is to analyze the behavioral characteristics of different player groups more targeted. For each behavioral pattern group, the probability distribution of blind box extraction is counted. For example, in the "casual player" group, the probability of obtaining ordinary props is 90%, and the probability of obtaining rare props is 10%; in the "heavy player" group, the probability of obtaining ordinary props is 50%, and the probability of obtaining rare props is 50%. The statistical probability distribution is to understand the probability of different groups obtaining different items. Compare the extraction probability distribution characteristics of each group with the preset normal probability distribution, and calculate the deviation.Assume that the preset normal probability distribution is that the probability of obtaining common props is 70%, and the probability of obtaining rare props is 30%. For example, the probability distribution of the "casual player" group has a small deviation from the preset normal probability distribution, and the probability distribution of the "heavy player" group has a large deviation from the preset normal probability distribution. The purpose of calculating the deviation is to determine whether there is abnormal behavior. If the deviation exceeds the preset threshold, the group is judged to have abnormal behavior. For example, the preset threshold is set to 0.2. If the deviation of the "heavy player" group is 0.4, which exceeds the preset threshold, the group is judged to have abnormal behavior. The purpose of judging abnormal behavior is to promptly discover and deal with possible cheating behavior. For the player group judged to have abnormal behavior, analyze its behavior sequence. For example, it is found that some players in the "heavy player" group can always obtain rare items with a very high probability within a specific time period. The purpose of analyzing the behavior sequence is to identify possible vulnerabilities. For the identified blind box extraction algorithm vulnerabilities, patch repair measures are taken. For example, if the vulnerability is found to be caused by imperfect server-side verification logic, the server-side verification logic needs to be improved; if the vulnerability is found to be caused by too low a frequency of random number seed updates, the frequency of random number seed updates needs to be increased. The purpose of fixing vulnerabilities is to prevent players from continuing to exploit vulnerabilities to cheat. Player accounts that exploit vulnerabilities to cheat will be warned or banned. For example, players in the "heavy player" group who are proven to have exploited vulnerabilities to cheat will be warned or their accounts will be directly banned. The purpose of dealing with cheating players is to maintain the fairness of the game.

[0094] In some embodiments, after adjusting the probability of drawing the blind box, the drawing of the blind box is monitored to determine whether the adjusted probability distribution will have an impact on the game economy.

[0095] include:

[0096] Obtain real-time data of blind box drawing, the real-time data including the number of drawings and the results of the drawings; obtain the actual drawing frequency of each type of item based on the real-time data; obtain a preset probability distribution model to calculate the theoretical drawing probability of each type of item; based on each type of item, divide the actual drawing frequency by the total number of drawings to obtain the actual drawing probability of that type of item; calculate the deviation between the actual drawing probability and the theoretical drawing probability; if the deviation exceeds a preset threshold, trigger the probability adjustment mechanism, and adjust the theoretical drawing probability by increasing or decreasing the weight according to the deviation to obtain an adjusted probability distribution; according to the adjusted probability distribution, randomly generate a preset number of virtual drawing processes to statistically analyze the model probability of each type of item. Proposed drawing frequency and simulated drawing probability; compare the simulated drawing probability with the actual drawing probability before adjustment, and evaluate the impact of the adjusted probability distribution on item rarity, player experience and game economy; obtain preset game indicators, which include total consumption amount, number of paying players, average consumption amount per person and number of daily active users; establish a correlation model between the adjusted probability distribution and the game indicators, predict the impact of the probability adjustment on each indicator, and obtain the prediction result; if there are indicators in the prediction result whose negative impact exceeds the preset acceptable threshold, cancel this probability adjustment; if there are no indicators in the prediction result whose negative impact exceeds the preset acceptable threshold, save the adjusted probability distribution parameters to the database.

[0097] Specifically, real-time monitoring of blind box extraction data is to promptly detect probability deviations and make dynamic adjustments. For example, the game sets the theoretical probability of SSR-level items to 1%, and theoretically one SSR should appear every 100 extractions. The system will record the results of each player's extraction in real time. If the actual number of SSRs in a short period of time (for example, 1,000 extractions) far exceeds the theoretical value (for example, 10 times), the probability adjustment mechanism will be triggered. The actual extraction frequency is obtained dynamically. Suppose there are three types of items in the game: SSR, SR, and R, and their preset theoretical probabilities are 1%, 10%, and 89%, respectively. The system records the extraction behavior of all players in real time. For example, in a certain period of time, the total number of extractions is 10,000, of which SSR appears 80 times, SR appears 1,100 times, and R appears 8,820 times. To calculate the actual extraction probability, the actual extraction frequency is divided by the total number of extractions. Taking the above example, the actual extraction probability of SSR is 80 / 10000=0.8%, SR is 1100 / 10000=11%, and R is 8820 / 10000=88.2%. There are many ways to calculate the deviation, such as the absolute value of the difference between the actual probability and the theoretical probability, or the relative deviation. Taking the above SSR as an example, if the absolute value method is used, the deviation is |0.8%-1%|=0.2%. The probability adjustment mechanism is to make the actual probability closer to the theoretical probability. For example, if the actual probability of SSR is found to be lower than the theoretical probability, the probability of SSR needs to be increased. A common adjustment method is weight adjustment, which can increase the weight of SSR while reducing the weight of other items to ensure that the total probability is still 100%. The virtual extraction process is to simulate the effect after the probability adjustment. Suppose the probability of SSR is adjusted to 1.2%, SR is adjusted to 9.8%, and R is adjusted to 89%. The system can simulate 10,000 draws, randomly draw according to the adjusted probability, and count the simulated draw frequency and simulated draw probability of various items. There are many factors to consider when evaluating the impact of probability adjustment. For example, increasing the SSR probability may increase players' willingness to pay, thereby increasing the total consumption amount and the per capita consumption amount. However, if the SSR probability is too high, it may reduce players' fun of drawing, resulting in a decrease in the number of paying players and the number of daily active users. Therefore, it is necessary to consider these factors comprehensively and set acceptable thresholds. The establishment of an association model requires a lot of data and statistical analysis. For example, the impact of past probability adjustments on various game indicators can be analyzed, and a mathematical model can be established to predict future impacts. Assume that the prediction results show that increasing the SSR probability will cause the number of daily active users to decrease by 10%. If this decrease exceeds the preset threshold of 5%, the probability adjustment will be canceled. The database saves the adjusted probability distribution parameters to persist the adjustment results. When the adjusted probability distribution passes all evaluations, the system saves the new probability parameters to the database and applies them to the actual blind box draw.In this way, subsequent player draws will be carried out according to the new probability. This can ensure the continued effectiveness of the probability adjustment.

[0098] In some embodiments, a player satisfaction evaluation model and a game balance evaluation index are constructed, and the blind box extraction probability is adjusted based on the player satisfaction evaluation model and the game balance evaluation index.

[0099] According to the historical game data, a player satisfaction evaluation model is constructed, and the satisfaction evaluation model is used to quantify player satisfaction under different extraction probabilities; based on the game balance, a game balance evaluation index is constructed, and the game balance evaluation index includes the player's sense of gain and the balance of game resource allocation; player satisfaction and game balance are used as optimization targets of the multi-objective optimization algorithm, and the two optimization targets are integrated into a single optimization function through weighted summation; a heuristic search method is used to obtain an extraction probability combination that achieves a balance between the player satisfaction and the game balance; the extraction probability combination is optimized, and a new extraction probability combination is generated through crossover mutation operations, and its advantages and disadvantages are evaluated according to the fitness function; according to the optimized extraction probability combination, the blind box extraction probability is adjusted.

[0100] Specifically, player satisfaction and game balance are two crucial aspects in game design. By building an evaluation model and adopting an optimization algorithm, we can find a combination of blind box extraction probabilities that balances the two, thereby improving the overall quality of the game. First, build a player satisfaction evaluation model. Based on historical game data, such as players' extraction records, game time, recharge amount, etc., we can quantify players' satisfaction under different extraction probabilities. For example, we can count the average game time and recharge amount of players under different rarity explosion rates. If the explosion rate of high-rarity items is too low under a certain probability, it may lead to player loss, and the game time and recharge amount will decrease, and the satisfaction will be low. On the contrary, if the explosion rate is too high, although it may increase player satisfaction in the short term, it may reduce the fun of the game and willingness to pay in the long run, which will also lead to a decrease in satisfaction. Player satisfaction can be expressed on a scale of 1-10 points, and a satisfaction function can be formulated based on historical data to map factors such as game time and recharge amount to satisfaction scores. Secondly, build game balance evaluation indicators. Game balance includes players' sense of gain and the balance of game resource allocation. Players' sense of gain refers to the feeling of players getting in-game rewards through hard work. For example, if the burst rate is too low, players will spend a lot of time and money but get nothing, and the sense of gain is extremely low. The balance of game resource allocation means that the gap between game resources between different players is controlled within a reasonable range. For example, if the burst rate is too high, a few players will quickly obtain a large number of rare items, breaking the game balance and damaging the game experience of other players. The Gini coefficient can be used to measure the balance of game resource allocation. The lower the Gini coefficient, the more balanced the resource allocation. Then, player satisfaction and game balance are used as optimization targets of the multi-objective optimization algorithm. Since player satisfaction and game balance sometimes conflict, a balance point needs to be found. The two optimization targets can be integrated into a single optimization function by weighted summation. For example, the player satisfaction weight can be set to 0.7 and the game balance weight can be set to 0.3, indicating that player satisfaction is more valued. Next, a heuristic search method, such as a genetic algorithm, is used to obtain a combination of extraction probabilities that strikes a balance between player satisfaction and game balance. For example, a set of extraction probability combinations can be randomly generated as the initial population, and then new probability combinations can be generated through operations such as crossover mutation, and their pros and cons can be evaluated according to the fitness function (i.e., the optimization function after weighted summation). Select the probability combination with high fitness for the next round of iteration, and finally get an optimal probability combination. Take a blind box containing three rarity items as an example. The initial probability combination is: 80% explosion rate for low rarity items, 15% explosion rate for medium rarity items, and 5% explosion rate for high rarity items. After multiple rounds of iterative optimization, a new probability combination is obtained: 70% explosion rate for low rarity items, 25% explosion rate for medium rarity items, and 5% explosion rate for high rarity items. This new probability combination achieves a better balance between player satisfaction and game balance.Finally, based on the optimized extraction probability combination, adjust the blind box extraction probability, continuously monitor game data, evaluate the adjustment effect, and make fine adjustments based on actual conditions to ensure the continued healthy development of the game.

[0101] See also Figure 2 , this application also proposes a game blind box extraction system based on player data, including:

[0102] The first processing module 201: obtains the player's historical game data, sets different time window lengths for different players, obtains the player's long-term game preferences and short-term game interest changes, and adjusts the probability of drawing the blind box based on the player's long-term game preferences and short-term game interest changes;

[0103] The second processing module 202: obtaining the game behavior data of the player in different time periods, and identifying the false interest of the player. If the player has false interest, the probability of drawing the blind box is not adjusted;

[0104] The third processing module 203: in the process of adjusting the probability of drawing the blind box, multiple accounts of the same player are identified, the game data of the associated accounts are merged, and the probability of drawing the blind box is adjusted according to the merged game data; or a player satisfaction evaluation model and a game balance evaluation index are constructed, and the probability of drawing the blind box is adjusted based on the player satisfaction evaluation model and the game balance evaluation index;

[0105] Fourth processing module 204: monitoring the blind box drawing behavior of the player, and identifying abnormal behavior that uses system loopholes to manipulate the probability distribution of blind box drawing;

[0106] Fifth processing module 205: After adjusting the probability of drawing the blind box, monitor the drawing of the blind box to determine whether the adjusted probability distribution will have an impact on the game economy.

[0107] It is understandable that if Figure 1 The contents of the game blind box extraction method based on player data shown in the embodiment are all applicable to the game blind box extraction system based on player data in the embodiment. The functions specifically implemented by the game blind box extraction system based on player data are similar to those in the embodiment shown in the embodiment. Figure 1 The game blind box extraction method based on player data is the same as the embodiment shown in the figure, and the beneficial effects achieved are the same as those of the embodiment shown in the figure. Figure 1 The beneficial effects achieved by the embodiment of the game blind box extraction method based on player data shown are also the same.

[0108] It should be noted that the information interaction, execution process and other contents between the above-mentioned systems are based on the same concept as the embodiment of the method of the present invention. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0109] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0110] See also Figure 3 An embodiment of the present invention also provides a computer device 3, including: a memory 302 and a processor 301 and a computer program 303 stored on the memory 302. When the computer program 303 is executed on the processor 301, a game blind box extraction method based on player data as described in any one of the above methods is implemented.

[0111] The computer device 3 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The computer device 3 may include, but is not limited to, a processor 301 and a memory 302. Those skilled in the art will appreciate that Figure 3 It is only an example of computer device 3 and does not constitute a limitation on computer device 3. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components, for example, it may also include input and output devices, network access devices, etc.

[0112] The processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0113] In some embodiments, the memory 302 may be an internal storage unit of the computer device 3, such as a hard disk or memory of the computer device 3. In other embodiments, the memory 302 may also be an external storage device of the computer device 3, such as a plug-in hard disk, a smart memory card (SmartMed iaCard, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. equipped on the computer device 3. Further, the memory 302 may also include both an internal storage unit of the computer device 3 and an external storage device. The memory 302 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program. The memory 302 may also be used to temporarily store data that has been output or is to be output.

[0114] An embodiment of the present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the game blind box extraction method based on player data as described in any of the above methods is implemented.

[0115] In this embodiment, if the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the camera / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0116] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for extracting game blind boxes based on player data, characterized in that: include: Obtain the player's historical game data, set different time window lengths for different players, obtain the player's long-term game preferences and short-term game interest changes, and adjust the probability of drawing blind boxes based on the player's long-term game preferences and short-term game interest changes; Obtain the game behavior data of players in different time periods and identify the false interests of players. If the player has false interests, the probability of his / her blind box drawing will not be adjusted; In the process of adjusting the probability of drawing the blind box, multiple accounts of the same player are identified, the game data of the associated accounts are merged, and the probability of drawing the blind box is adjusted according to the merged game data; or a player satisfaction evaluation model and a game balance evaluation index are constructed, and the probability of drawing the blind box is adjusted based on the player satisfaction evaluation model and the game balance evaluation index; Monitor the blind box drawing behavior of the players and identify abnormal behaviors that exploit system loopholes to manipulate the probability distribution of blind box drawing; After adjusting the probability of drawing the blind box, monitor the drawing of the blind box to determine whether the adjusted probability distribution will have an impact on the game economy.

2. The method for extracting a blind game box according to claim 1, characterized in that: Obtain the player's historical game data, set different time window lengths for different players, obtain the player's long-term game preferences and short-term game interest changes, and adjust the probability of drawing the blind box based on the player's long-term game preferences and short-term game interest changes, including: Obtain the player's historical game data, which includes the player's game behavior, game duration, game frequency, and consumption records; According to the historical game data, the players are divided into different player groups, each of the player groups having similar game preferences and behavior patterns; Determine the length of the time window based on the historical game data of each player group, and predict the players' long-term game preferences and short-term interest changes for different blind boxes based on the game behaviors and preferences within the time window; Adjust the probability distribution of drawing each blind box based on the long-term gaming preferences and short-term interest changes.

3. The method for extracting a blind game box according to claim 1, characterized in that: Obtain the game behavior data of players in different time periods and identify the false interests of players. If the players have false interests, do not adjust their blind box drawing probability, including: Counting the frequency of the players drawing each type of blind box within a preset time period, and obtaining the players' preference distribution for different types of blind boxes based on the frequency; Compare the player's preference distribution for blind boxes with his / her long-term game preferences. If there is a significant deviation, it is determined that the player's interest in blind boxes is abnormal; Use anomaly detection algorithms such as Isolation Forest to detect anomalies in the blind box extraction behavior of players with abnormal interests and identify players who show false interests; The false interest of the player is quantified, the false interest value of the player is obtained, and a corresponding false interest threshold is set; when the false interest value exceeds the false interest threshold, the player is marked as a false player, and when the blind box probability is adjusted, the blind box drawing probability of the false player is not adjusted.

4. The method for extracting a blind game box according to claim 1, characterized in that: Constructing a player satisfaction evaluation model and a game balance evaluation index, and adjusting the blind box extraction probability based on the player satisfaction evaluation model and the game balance evaluation index, including: Based on the historical game data, a player satisfaction evaluation model is constructed, wherein the player satisfaction evaluation model is used to quantify the player satisfaction under different extraction probabilities; Based on the game balance, construct a game balance evaluation index, wherein the game balance evaluation index includes the player's sense of gain and the balance of game resource allocation; Taking player satisfaction and game balance as optimization targets of the multi-objective optimization algorithm, and integrating the two optimization targets into a single optimization function through weighted summation; Using a heuristic search method to obtain a combination of extraction probabilities that strikes a balance between the player satisfaction and the game balance; Optimizing the extraction probability combination, generating a new extraction probability combination through a crossover mutation operation, and evaluating its quality according to a fitness function; Adjust the blind box drawing probability according to the optimized drawing probability combination.

5. The method for extracting a blind game box according to claim 1, characterized in that: Monitor the blind box drawing behavior of the players and identify abnormal behaviors that use system loopholes to manipulate the probability distribution of blind box drawing, including: Obtaining the player's behavior data during the blind box drawing process, the behavior data including the drawing time, the number of draws, and the information of the items obtained; Based on the behavior data, extract the characteristics of the frequency of extraction, the number of consecutive extractions and the probability of obtaining rare items, and construct a player behavior pattern feature vector; Dividing the players into different behavior pattern groups according to the player behavior pattern feature vectors; Based on each of the behavior pattern groups, the probability distribution of blind box extraction is counted to obtain the extraction probability distribution characteristics of each group; Comparing the extraction probability distribution characteristics of each group with a preset normal probability distribution, and calculating the deviation; If the deviation exceeds a preset threshold, it is determined that the group has abnormal behavior; For groups of players who are judged to have abnormal behavior, analyze their behavior sequences and identify the blind box extraction algorithm loopholes they exploit; According to the identified blind box extraction algorithm vulnerability, patch repair measures are taken, including improving the server-side verification logic and increasing the random number seed update frequency; Players who exploit loopholes in the blind box extraction algorithm will be warned or banned.

6. The method for extracting a blind game box according to claim 1, characterized in that: In the process of adjusting the probability of drawing the blind box, multiple accounts of the same player are identified, game data of the associated accounts are merged, and the probability of drawing the blind box is adjusted according to the merged game data, including: Through account association analysis technology, multiple account information related to players is extracted from the game database; Use data merging method to aggregate game data of multiple accounts into one main account; Extracting the player's real game preference based on the merged game data through data analysis; The drawing probability is adjusted according to the actual game preference.

7. The method for extracting a blind game box according to claim 1, characterized in that: After adjusting the probability of drawing the blind box, monitor the drawing of the blind box to determine whether the adjusted probability distribution will have an impact on the game economy, including: Obtaining real-time data of blind box draws, including the number of draws and draw results; According to the real-time data, the actual extraction frequency of each type of item is obtained; Obtain the preset probability distribution model and calculate the theoretical extraction probability of each type of item; Based on each category of items, the actual extraction frequency is divided by the total number of extractions to obtain the actual extraction probability of the item in that category; Calculating the deviation between the actual extraction probability and the theoretical extraction probability; If the deviation exceeds a preset threshold, a probability adjustment mechanism is triggered to adjust the theoretical extraction probability by increasing or decreasing the weight according to the deviation to obtain an adjusted probability distribution; According to the adjusted probability distribution, by randomly generating a preset number of virtual drawing processes, the simulated drawing frequency and simulated drawing probability of each type of item are counted; Comparing the simulated draw probability with the actual draw probability before adjustment, and evaluating the impact of the adjusted probability distribution on item rarity, player experience, and game economy; Obtaining preset game indicators, including total spending amount, number of paying players, average spending amount per person, and number of daily active users; Establishing a correlation model between the adjusted probability distribution and the game index, predicting the impact of the probability adjustment on each index, and obtaining a prediction result; If there are indicators in the forecast results whose negative impact exceeds the preset acceptable threshold, the probability adjustment will be cancelled; If there is no indicator in the prediction result whose negative impact exceeds the preset acceptable threshold, the adjusted probability distribution parameters are saved in the database.

8. A game blind box extraction system based on player data, characterized in that: include: The first processing module: obtains the player's historical game data, sets different time window lengths for different players, obtains the player's long-term game preferences and short-term game interest changes, and adjusts the probability of drawing blind boxes based on the player's long-term game preferences and short-term game interest changes; The second processing module: obtains the game behavior data of players in different time periods, and identifies the false interests of players. If the player has false interests, the probability of drawing the blind box will not be adjusted; The third processing module: in the process of adjusting the probability of drawing the blind box, multiple accounts of the same player are identified, the game data of the associated accounts are merged, and the probability of drawing the blind box is adjusted according to the merged game data; or a player satisfaction evaluation model and a game balance evaluation index are constructed, and the probability of drawing the blind box is adjusted based on the player satisfaction evaluation model and the game balance evaluation index; The fourth processing module: monitors the blind box drawing behavior of the player and identifies abnormal behaviors that use system loopholes to manipulate the probability distribution of blind box drawing; The fifth processing module: after adjusting the probability of drawing the blind box, monitor the drawing situation of the blind box to determine whether the adjusted probability distribution will have an impact on the game economy.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images