A method for monitoring an external plug-in of an MMO game and a terminal

Abstract

The application discloses an external hanging monitoring method and terminal of MMO game, based on normal click behavior model, click data is monitored to obtain first monitoring result, based on normal skill release behavior model, skill release data is monitored to obtain second monitoring result, based on normal task completion behavior model, task completion data is monitored to obtain third monitoring result, if three kinds of monitoring results are normal behavior, it is determined that the player is a normal player, otherwise, the player is added to a suspected external hanging list, through monitoring three kinds of behavior data of the player, the player's possible use of external hanging in the game is effectively captured from three aspects of click behavior, skill release behavior and task completion behavior, so that the external hanging behavior in the game is accurately and effectively monitored, and subsequent game operation team can further investigate and verify the suspected external hanging player according to the suspected external hanging list, so that the fairness and normal order of the game are maintained.

Description

Technical Field

[0001] This invention relates to the field of cheat detection, and more particularly to a method and terminal for detecting cheats in MMO games. Background Technology

[0002] In MMOs (Massively Multiplayer Online) games, cheating severely undermines the fairness and normal order of the game, negatively impacting the gaming experience for a large number of players. Therefore, game developers need to consider how to accurately monitor cheating behavior in order to effectively combat it and maintain a healthy and stable game environment. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a method and terminal for detecting cheats in MMO games, which can accurately and effectively detect cheat behavior in games.

[0004] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: A method for detecting cheats in MMO games, comprising the following steps: Collect player behavior data, including click data, skill release data, and task completion data; The click data is monitored based on a normal click behavior model to obtain a first monitoring result; The skill release data is monitored based on a normal skill release behavior model to obtain a second monitoring result; The task completion data is monitored based on a normal task completion behavior model to obtain a third monitoring result; If the first monitoring result, the second monitoring result, and the third monitoring result are all normal behavior, then the player is determined to be a normal player; otherwise, the player is added to the suspected cheater list.

[0005] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is as follows: A cheat detection terminal for an MMO game includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it performs the following steps: Collect player behavior data, including click data, skill release data, and task completion data; The click data is monitored based on a normal click behavior model to obtain a first monitoring result; The skill release data is monitored based on a normal skill release behavior model to obtain a second monitoring result; The task completion data is monitored based on a normal task completion behavior model to obtain a third monitoring result; If the first monitoring result, the second monitoring result, and the third monitoring result are all normal behavior, then the player is determined to be a normal player; otherwise, the player is added to the suspected cheater list.

[0006] The beneficial effects of this invention are as follows: It collects player behavior data, including click data, skill release data, and task completion data. Based on a normal click behavior model, it monitors the click data to obtain a first monitoring result; based on a normal skill release behavior model, it monitors the skill release data to obtain a second monitoring result; and based on a normal task completion behavior model, it monitors the task completion data to obtain a third monitoring result. If all three monitoring results are normal, the player is determined to be a normal player; otherwise, the player is added to a suspected cheater list. By monitoring these three types of player behavior data—click behavior, skill release behavior, and task completion behavior—it provides a more comprehensive monitoring of player game behavior, effectively capturing situations where players are likely to use cheats in the game. This allows for accurate and effective monitoring of cheating behavior in the game. Subsequently, the game operations team can further investigate and verify suspected cheating players based on the suspected cheater list, maintaining the fairness and normal order of the game. Attached Figure Description

[0007] Figure 1 This is a flowchart illustrating the steps of a method for detecting cheats in an MMO game according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of a cheat monitoring terminal for an MMO game according to an embodiment of the present invention. Detailed Implementation

[0008] To explain in detail the technical content, objectives, and effects of the present invention, the following description is provided in conjunction with the embodiments and accompanying drawings.

[0009] Please refer to Figure 1 A method for detecting cheats in MMO games, comprising the following steps: Collect player behavior data, including click data, skill release data, and task completion data; The click data is monitored based on a normal click behavior model to obtain a first monitoring result; The skill release data is monitored based on a normal skill release behavior model to obtain a second monitoring result; The task completion data is monitored based on a normal task completion behavior model to obtain a third monitoring result; If the first monitoring result, the second monitoring result, and the third monitoring result are all normal behavior, then the player is determined to be a normal player; otherwise, the player is added to the suspected cheater list.

[0010] As described above, the beneficial effects of this invention are as follows: It collects player behavior data, including click data, skill release data, and task completion data. Based on a normal click behavior model, it monitors the click data to obtain a first monitoring result; based on a normal skill release behavior model, it monitors the skill release data to obtain a second monitoring result; and based on a normal task completion behavior model, it monitors the task completion data to obtain a third monitoring result. If all three monitoring results are normal behavior, the player is determined to be a normal player; otherwise, the player is added to a suspected cheater list. By monitoring these three types of player behavior data—click behavior, skill release behavior, and task completion behavior—it comprehensively monitors player game behavior from three aspects, effectively capturing situations where players are likely to use cheats in the game. This allows for accurate and effective monitoring of cheating behavior in the game. Subsequently, the game operation team can further investigate and verify suspected cheating players based on the suspected cheater list, maintaining the fairness and normal order of the game.

[0011] Furthermore, it also includes: Acquire historical behavior data of normal players, including historical click data, historical skill release data, and historical task completion data; A normal click behavior model is established using machine learning algorithms based on the historical click data. A normal skill release behavior model is established using machine learning algorithms based on the historical skill release data. Based on the historical task completion data, a normal task completion behavior model is established using machine learning algorithms.

[0012] As described above, by collecting historical behavioral data from normal players and using machine learning algorithms to establish models of normal click behavior, normal skill release behavior, and normal task completion behavior, the model can better reflect the actual game scenarios and the actual behavior of game players, making subsequent monitoring more accurate and reliable.

[0013] Furthermore, the click data includes click frequency, click location, and click time interval; The first monitoring result obtained by monitoring the click data based on the normal click behavior model includes: If the click frequency exceeds the normal click frequency distribution range in the normal click behavior model, the first monitoring result is determined to be suspected cheating behavior. If it does not exceed the range, the randomness of the click position is determined to be lower than the randomness of the normal click position in the normal click behavior model. If it is lower, the first monitoring result is determined to be suspected cheating behavior. If it is not lower, the volatility of the click time interval is determined to be lower than the volatility of the normal click time interval in the normal click behavior model. If it is, the first monitoring result is determined to be suspected cheating behavior; if not, the first monitoring result is determined to be normal behavior.

[0014] As described above, when monitoring player click data, if a player is using cheats, their click frequency will be abnormally high, or the click location will be too precise and regular, lacking randomness, or the click time interval will be too fixed, unlike normal players who have certain fluctuations. Therefore, by comparing the click frequency, click location, and click time interval with the normal click behavior model, it is possible to effectively determine whether a player is using a click simulation cheat.

[0015] Furthermore, the skill release data includes the game scene and the skill release timing, skill release order, skill release frequency, and skill combination mode corresponding to the game scene; The second monitoring result obtained by monitoring the skill release data based on the normal skill release behavior model includes: The system determines whether the timing of skill release corresponding to the game scenario conforms to the normal skill release timing in the game scenario of the normal skill release behavior model. If it does not conform, the second monitoring result is determined to be suspected cheating behavior. If it does conform, the system determines whether the skill release order corresponding to the game scenario conforms to the normal skill release order in the game scenario of the normal skill release behavior model. If it does not conform, the second monitoring result is determined to be suspected cheating behavior. If it does conform, the system determines whether the skill release frequency corresponding to the game scenario exceeds the normal skill release frequency in the game scenario of the normal skill release behavior model. If it exceeds the frequency, the second monitoring result is determined to be suspected cheating behavior. If it does not exceed the frequency, the system determines whether the skill combination pattern corresponding to the game scenario conforms to the normal skill combination pattern in the game scenario of the normal skill release behavior model. If it conforms, the second monitoring result is determined to be normal behavior. If it does not conform, the second monitoring result is determined to be suspected cheating behavior.

[0016] As described above, if a player uses cheats when releasing skills, they will disregard normal skill release timing rules. This is because cheats may modify game data to ensure skills hit enemies without the player taking damage, or the player's skill release order may not conform to normal combat strategies, resulting in random skill releases. Furthermore, cheaters may exceed skill cooldown limits, using skills again within the theoretical cooldown time, exhibiting abnormal skill release frequency, and using unreasonable skill combinations, disregarding game fairness and normal combat strategies. Therefore, monitoring player skill release behavior by judging whether the timing, order, frequency, and combination patterns of skill release are normal within the current game scenario is more accurate and reasonable.

[0017] Furthermore, the task completion data includes the task completion time and the copy completion time; The third monitoring result obtained by monitoring the task completion data based on the normal task completion behavior model includes: If the task completion time is lower than the average task completion time or the fluctuation range of task completion time in the normal task completion behavior model, then the third monitoring result is determined to be suspected cheating behavior. If it is not lower than the average task completion time or the fluctuation range of task completion time in the normal task completion behavior model, then the third monitoring result is determined to be suspected cheating behavior. If it is not lower than the average task completion time, then the third monitoring result is determined to be normal behavior.

[0018] As described above, if a player completes a task or dungeon in a time significantly shorter than normal, such as completing a difficult task that usually takes a long time in a very short time, then there may be suspicion of using cheats to speed up the game. Therefore, by monitoring task completion time and dungeon completion time, we can determine whether a player has used cheats when completing tasks and dungeons. In addition to comparing the completion time with the average completion time in the normal task completion behavior model, we also compare it with the time fluctuation range, which is more in line with the differences among game players and makes cheat detection more reasonable and effective.

[0019] Please refer to Figure 2 A cheat detection terminal for an MMO game includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it performs the following steps: Collect player behavior data, including click data, skill release data, and task completion data; The click data is monitored based on a normal click behavior model to obtain a first monitoring result; The skill release data is monitored based on a normal skill release behavior model to obtain a second monitoring result; The task completion data is monitored based on a normal task completion behavior model to obtain a third monitoring result; If the first monitoring result, the second monitoring result, and the third monitoring result are all normal behavior, then the player is determined to be a normal player; otherwise, the player is added to the suspected cheater list.

[0020] As described above, the beneficial effects of this invention are as follows: It collects player behavior data, including click data, skill release data, and task completion data. Based on a normal click behavior model, it monitors the click data to obtain a first monitoring result; based on a normal skill release behavior model, it monitors the skill release data to obtain a second monitoring result; and based on a normal task completion behavior model, it monitors the task completion data to obtain a third monitoring result. If all three monitoring results are normal behavior, the player is determined to be a normal player; otherwise, the player is added to a suspected cheater list. By monitoring these three types of player behavior data—click behavior, skill release behavior, and task completion behavior—it comprehensively monitors player game behavior from three aspects, effectively capturing situations where players are likely to use cheats in the game. This allows for accurate and effective monitoring of cheating behavior in the game. Subsequently, the game operation team can further investigate and verify suspected cheating players based on the suspected cheater list, maintaining the fairness and normal order of the game.

[0021] Furthermore, it also includes: Acquire historical behavior data of normal players, including historical click data, historical skill release data, and historical task completion data; A normal click behavior model is established using machine learning algorithms based on the historical click data. A normal skill release behavior model is established using machine learning algorithms based on the historical skill release data. Based on the historical task completion data, a normal task completion behavior model is established using machine learning algorithms.

[0022] As described above, by collecting historical behavioral data from normal players and using machine learning algorithms to establish models of normal click behavior, normal skill release behavior, and normal task completion behavior, the model can better reflect the actual game scenarios and the actual behavior of game players, making subsequent monitoring more accurate and reliable.

[0023] Furthermore, the click data includes click frequency, click location, and click time interval; The first monitoring result obtained by monitoring the click data based on the normal click behavior model includes: If the click frequency exceeds the normal click frequency distribution range in the normal click behavior model, the first monitoring result is determined to be suspected cheating behavior. If it does not exceed the range, the randomness of the click position is determined to be lower than the randomness of the normal click position in the normal click behavior model. If it is lower, the first monitoring result is determined to be suspected cheating behavior. If it is not lower, the volatility of the click time interval is determined to be lower than the volatility of the normal click time interval in the normal click behavior model. If it is, the first monitoring result is determined to be suspected cheating behavior; if not, the first monitoring result is determined to be normal behavior.

[0024] As described above, when monitoring player click data, if a player is using cheats, their click frequency will be abnormally high, or the click location will be too precise and regular, lacking randomness, or the click time interval will be too fixed, unlike normal players who have certain fluctuations. Therefore, by comparing the click frequency, click location, and click time interval with the normal click behavior model, it is possible to effectively determine whether a player is using a click simulation cheat.

[0025] Furthermore, the skill release data includes the game scene and the skill release timing, skill release order, skill release frequency, and skill combination mode corresponding to the game scene; The second monitoring result obtained by monitoring the skill release data based on the normal skill release behavior model includes: The system determines whether the timing of skill release corresponding to the game scenario conforms to the normal skill release timing in the game scenario of the normal skill release behavior model. If it does not conform, the second monitoring result is determined to be suspected cheating behavior. If it does conform, the system determines whether the skill release order corresponding to the game scenario conforms to the normal skill release order in the game scenario of the normal skill release behavior model. If it does not conform, the second monitoring result is determined to be suspected cheating behavior. If it does conform, the system determines whether the skill release frequency corresponding to the game scenario exceeds the normal skill release frequency in the game scenario of the normal skill release behavior model. If it exceeds the frequency, the second monitoring result is determined to be suspected cheating behavior. If it does not exceed the frequency, the system determines whether the skill combination pattern corresponding to the game scenario conforms to the normal skill combination pattern in the game scenario of the normal skill release behavior model. If it conforms, the second monitoring result is determined to be normal behavior. If it does not conform, the second monitoring result is determined to be suspected cheating behavior.

[0026] As described above, if a player uses cheats when releasing skills, they will disregard normal skill release timing rules. This is because cheats may modify game data to ensure skills hit enemies without the player taking damage, or the player's skill release order may not conform to normal combat strategies, resulting in random skill releases. Furthermore, cheaters may exceed skill cooldown limits, using skills again within the theoretical cooldown time, exhibiting abnormal skill release frequency, and using unreasonable skill combinations, disregarding game fairness and normal combat strategies. Therefore, monitoring player skill release behavior by judging whether the timing, order, frequency, and combination patterns of skill release are normal within the current game scenario is more accurate and reasonable.

[0027] Furthermore, the task completion data includes the task completion time and the copy completion time; The third monitoring result obtained by monitoring the task completion data based on the normal task completion behavior model includes: If the task completion time is lower than the average task completion time or the fluctuation range of task completion time in the normal task completion behavior model, then the third monitoring result is determined to be suspected cheating behavior. If it is not lower than the average task completion time or the fluctuation range of task completion time in the normal task completion behavior model, then the third monitoring result is determined to be suspected cheating behavior. If it is not lower than the average task completion time, then the third monitoring result is determined to be normal behavior.

[0028] As described above, if a player completes a task or dungeon in a time significantly shorter than normal, such as completing a difficult task that usually takes a long time in a very short time, then there may be suspicion of using cheats to speed up the game. Therefore, by monitoring task completion time and dungeon completion time, we can determine whether a player has used cheats when completing tasks and dungeons. In addition to comparing the completion time with the average completion time in the normal task completion behavior model, we also compare it with the time fluctuation range, which is more in line with the differences among game players and makes cheat detection more reasonable and effective.

[0029] The above-described method and terminal for detecting cheats in MMO games are applicable to MMO games, and are described below through specific embodiments: Please refer to Figure 1 Embodiment 1 of the present invention is as follows: A method for detecting cheats in MMO games, comprising the following steps: S1. Obtain historical behavior data of normal players, including historical click data, historical skill release data, and historical task completion data.

[0030] The historical click data includes historical click frequency, historical click location, and historical click time interval; the historical skill release data includes game scenes and the corresponding historical skill release timing, historical skill release order, historical skill release frequency, and historical skill combination patterns; the task completion data includes historical task completion time and historical instance completion time. The more historical click data there is, the more accurate the model will be.

[0031] Specifically, this involves acquiring historical behavioral data from the entire game process of a normal player. This goes beyond a single, specific game progression point, comprehensively reflecting normal player behavior patterns. While game content can cause changes in player behavior—such as variations in click frequency and the randomness of location—collecting a large amount of long-term behavioral data across different scenarios can cover all possible situations. Of course, in practical applications, data from multiple detection points (such as different boss battle scenarios) can be used to further enrich and refine the model. However, this is a supplement to the large amount of overall game process data, not the sole data source.

[0032] S2. Based on the historical click data, a normal click behavior model is established using machine learning algorithms.

[0033] Specifically, machine learning algorithms are used to analyze the historical click frequency, historical click location, and historical click time interval to obtain the normal click frequency distribution range, the randomness of the normal click location, and the volatility of the normal click time interval. A normal click behavior model is then established based on the normal click frequency distribution range, the randomness of the normal click location, and the volatility of the normal click time interval.

[0034] This process involves counting the number of clicks each player makes within different time periods (e.g., by minute, by task stage, etc.), and then analyzing the data. This includes calculating statistics such as the mean and standard deviation. The mean serves as a typical click frequency, while the standard deviation reflects frequency fluctuations. By comprehensively considering these statistics, a reasonable and normal click frequency distribution range is determined. For each player's historical click locations, the degree of randomness can be measured by calculating the probability distribution of clicks in different areas of the screen. For example, if a player clicks with a high probability in a small area but almost none in other areas, the randomness is low. Conversely, if the probability of clicking is relatively even across different areas of the screen, the randomness is high. The degree of randomness can be quantified using mathematical models (such as probability distribution models). Similarly, the time interval between two consecutive clicks for each player is analyzed by calculating statistics such as the mean and standard deviation to assess the volatility of the time interval. A smaller standard deviation indicates a more stable time interval with less fluctuation, while a larger standard deviation indicates greater fluctuation.

[0035] S3. Based on the historical skill release data, a normal skill release behavior model is established using machine learning algorithms.

[0036] Specifically, machine learning algorithms are used to analyze the historical skill release timing, historical skill release order, historical skill release frequency, and historical skill combination patterns in the game scenario to obtain the normal skill release timing, normal skill release order, normal skill release frequency, and normal skill combination patterns in the game scenario. Based on the normal skill release timing, normal skill release order, normal skill release frequency, and normal skill combination patterns in the game scenario, a normal skill behavior model is established.

[0037] S4. Based on the historical task completion data, a normal task completion behavior model is established using machine learning algorithms.

[0038] Specifically, machine learning algorithms are used to analyze the completion time of historical tasks and the completion time of historical replicas to obtain the average task completion time, the fluctuation range of task completion time, the average completion time of replicas and the fluctuation range of replica completion time. A normal task completion behavior model is then established based on the average task completion time, the fluctuation range of task completion time, the average completion time of replicas and the fluctuation range of replica completion time.

[0039] The machine learning algorithms mentioned include random forest algorithm, linear regression, clustering algorithm, isolated forest algorithm, etc.

[0040] S5. Collect player behavior data, including click data, skill release data, and task completion data.

[0041] The click data includes click frequency, click location, and click time interval; the skill release data includes the game scene and the corresponding skill release timing, skill release order, skill release frequency, and skill combination mode; the task completion data includes task completion time and dungeon completion time.

[0042] Specifically, data scraping technology is used to collect player behavior data, and distributed data storage technology can be used to save the player behavior data.

[0043] S6. Monitor the click data based on the normal click behavior model to obtain the first monitoring result.

[0044] When monitoring players, their click data is compared with normal click behavior models to analyze the differences in click trajectories. Normal players' click behavior is random and diverse, while players using click-simulating cheats may click with abnormally high frequency and overly precise and regular click locations. Furthermore, this also helps identify users of bots and other cheats, as the operation trajectories of bots are usually mechanical and monotonous, significantly different from the flexibility and randomness of real players.

[0045] Specifically, it is determined whether the click frequency exceeds the normal click frequency distribution range in the normal click behavior model. If it does, the first monitoring result is determined to be suspected cheating behavior. If it does not exceed the range, it is determined whether the randomness of the click position is lower than the randomness of the normal click position in the normal click behavior model. If it is lower, the first monitoring result is determined to be suspected cheating behavior. If it is not lower, it is determined whether the volatility of the click time interval is lower than the volatility of the normal click time interval in the normal click behavior model. If it is, the first monitoring result is determined to be suspected cheating behavior. If it is not, the first monitoring result is determined to be normal behavior.

[0046] If a player's click frequency is abnormally high, exceeding the normal click frequency distribution range in the normal click behavior model; or the click location is too precise and regular, lacking the randomness of a normal player; or the time interval is too fixed, unlike a normal player who has certain fluctuations, then it can be preliminarily judged that the player may be using a click simulation cheat.

[0047] Additionally, the movement trajectory of a player's mouse or touch operation can be analyzed. If the trajectory is relatively mechanical and monotonous, and significantly different from the flexibility and randomness of a normal player, it can also serve as evidence of the use of cheats such as AFK scripts.

[0048] S7. Monitor the skill release data based on the normal skill release behavior model to obtain a second monitoring result.

[0049] Normal players in the game judge the timing, order, and frequency of skill release based on factors such as the actual combat situation and the opponent's behavior. The timing, order, and frequency of skill release have a certain degree of randomness and strategy. Cheating players, on the other hand, may unleash a large number of high-damage skills in a very short period, or frequently use specific skills at inappropriate times to gain an advantage.

[0050] Regarding the timing of skill release, in combat scenarios, such as when facing a melee enemy, a normal player will release a melee skill when the enemy gets close enough and they are in a suitable attack position. For example, in a role-playing game dungeon, when a monster enters within 3-5 meters of the player and the player is not disturbed by other monsters, the player might release a melee skill with a knockback effect to create distance and avoid damage. When facing ranged enemies, a normal player will choose the timing of skill release based on the enemy's attack rhythm and their own dodging ability. For example, after the enemy launches a ranged attack, there is a short skill cooldown or attack interval. A normal player will choose to release a skill to counterattack during this interval, or use the terrain to avoid enemy attacks and then release the skill at an appropriate time. However, cheaters may disregard normal skill timing rules. For example, in the scenario of facing melee enemies, cheaters may release skills before the enemy gets close enough, or release skills even when they are in a disadvantageous position (such as being surrounded by enemies and unable to dodge attacks). This is because cheaters may modify game data to ensure that skills hit enemies without causing damage. When facing ranged enemies, cheaters may release skills to counterattack as soon as the enemy launches a ranged attack, without considering whether they can dodge the enemy's attack. This is because cheaters may provide illegal dodge functions or directly modify skill effects, allowing skills to take effect regardless of enemy attacks.

[0051] Regarding skill release order, normal players typically determine it based on combat strategy and their own skill characteristics during a battle. For example, in a team battle with healers, DPS, and tanks, the tank might use aggro-drawing skills first to concentrate enemy attacks on themselves. DPS players, depending on the enemy's health and defense, might use low-damage skills that weaken the enemy's defense before unleashing high-damage attacks. Healers would use healing skills as needed when teammates' health is low. In solo combat, players might arrange their skill release order based on skill cooldowns and effects. For instance, a player with multiple attack skills and one control skill might use the control skill first to restrain the enemy before using attack skills in sequence. However, cheaters' skill release order may deviate from normal combat strategies. For example, in team battles, cheaters might not release skills in the normal tank-DPS-healer order, but rather randomly, perhaps using high-damage skills first without considering whether they will attract enemy aggro, leading to chaos in the team battle. In single-player combat, cheaters may not arrange the order of skill release according to the cooldown time and effect of their own skills, but instead frequently use high-damage skills without considering whether they have enough resources (mana or energy) to support the use of these skills.

[0052] Regarding skill usage frequency, a normal player's skill usage frequency is affected by their skill cooldown time, mana or energy (depending on the game settings), and the combat situation. For example, in a prolonged battle, if a player's skill has a 10-second cooldown, then under normal circumstances, that skill would be used approximately 6 times per minute (considering the possibility of missing some opportunities to use it due to dodging enemy attacks). Furthermore, different skills have different usage frequencies; some powerful but resource-intensive skills will have a lower usage frequency, while some support skills may have a relatively higher usage frequency. However, cheaters may bypass skill cooldown time limits, using skills again within the theoretical cooldown period. For example, a cheater might use a skill with a 10-second cooldown within 5 seconds, significantly increasing the skill usage frequency. Moreover, cheaters might use powerful but resource-intensive skills without restraint, disregarding their own resource situation, thus disrupting the game's resource balance.

[0053] Regarding skill combination modes, in dungeon scenarios, players might choose skill combinations based on the characteristics of the monsters. For example, in a dungeon with a large number of monsters, a player might use a skill that deals damage to the entire group, combined with a skill that increases their own defense, to survive and deal damage amidst the monster horde. In PvP (player versus player) scenarios, players will choose skill combinations based on the opponent's class and skill characteristics. For example, against a high-output mage, a player might choose a combination of skills with silence effects and skills that can interrupt the opponent's spellcasting to limit their damage output. However, in dungeons, cheaters might use unreasonable skill combinations. For instance, in dungeons with large groups of monsters, cheaters might only use high-damage skills without considering their own survival, as cheats may provide illegal survival assistance functions, such as infinite health or invincibility. In PvP scenarios, cheaters might use skill combinations that disrupt the game's balance. For example, against a high-output mage, a cheater might use a skill combination that can instantly kill the opponent, disregarding the game's fairness and normal combat strategies.

[0054] This invention monitors players' skill release behavior, analyzing the timing, sequence, frequency, and combination patterns of skill releases to determine if they conform to normal game logic. Furthermore, the analysis of skill release behavior needs to be closely linked to specific scenarios, as player skill release behaviors vary significantly across different game environments. For example, in dungeon scenarios, the player's goal is to defeat monsters and complete dungeon quests, so skill releases will focus on damaging monsters and ensuring the player's survival. In PvP scenarios, the player's goal is to defeat opponents, so skill releases will focus on limiting opponents and leveraging personal strengths. Only by analyzing skill release behavior in conjunction with specific game scenarios can the accuracy of player actions be determined, and whether cheating is being used can be accurately assessed. Therefore, this invention's skill release data includes game scenarios to achieve more accurate and reliable cheat detection.

[0055] Specifically, it is determined whether the timing of skill release corresponding to the game scene conforms to the normal skill release timing in the game scene of the normal skill release behavior model. If it does not conform, the second monitoring result is determined to be suspected cheating behavior. If it does conform, it is determined whether the skill release order corresponding to the game scene conforms to the normal skill release order in the game scene of the normal skill release behavior model. If not, the second monitoring result is determined to be suspected cheating behavior. If it does, it is determined whether the skill release frequency corresponding to the game scene exceeds the normal skill release frequency in the game scene of the normal skill release behavior model. If it exceeds, the second monitoring result is determined to be suspected cheating behavior. If it does not exceed, it is determined whether the skill combination pattern corresponding to the game scene conforms to the normal skill combination pattern in the game scene of the normal skill release behavior model. If it conforms, the second monitoring result is determined to be normal behavior. If it does not conform, the second monitoring result is determined to be suspected cheating behavior.

[0056] If a player unleashes a large number of high-damage skills in a very short period of time, or frequently uses specific skills at inappropriate times, they may be using cheats. Pattern recognition technology, such as support vector machines, can be used to determine whether the skill release is abnormal.

[0057] In terms of resource acquisition and usage, legitimate players and cheaters will also exhibit different behaviors. Cheating players may rapidly acquire large amounts of resources through illegal means, or exhibit abnormal behavior such as unreasonable resource consumption rates, which can be monitored and analyzed.

[0058] In one optional implementation, the skill release data further includes the resource acquisition speed and resource consumption speed corresponding to the game scene. The normal skill release behavior model also includes the normal resource acquisition speed and normal resource consumption speed under the game scene. It further includes: determining whether the resource acquisition speed corresponding to the game scene is greater than the normal resource acquisition speed under the game scene in the normal skill release behavior model; if so, determining that the second monitoring result is suspected cheating behavior; if not, determining whether the resource consumption speed corresponding to the game scene is greater than the normal resource consumption speed under the game scene in the normal skill release behavior model; if so, determining that the second monitoring result is suspected cheating behavior; if not, determining that the second monitoring result is normal behavior.

[0059] S8. Monitor the task completion data based on the normal task completion behavior model to obtain a third monitoring result.

[0060] Specifically, it is determined whether the task completion time is lower than the average task completion time or the fluctuation range of task completion time in the normal task completion behavior model. If it is lower, the third monitoring result is determined to be suspected cheating behavior. If it is not lower, it is determined whether the copy completion time is lower than the average copy completion time or the fluctuation range of copy completion time in the normal task completion behavior model. If it is, the third monitoring result is determined to be suspected cheating behavior. If not, the third monitoring result is determined to be normal behavior.

[0061] Different game tasks and dungeons have their normal completion time ranges. The normal task completion behavior model of this invention calculates average completion times and reasonable time fluctuation ranges based on data from a large number of players. If a player completes a task or dungeon much faster than the normal range—for example, completing a difficult task that usually takes a long time in a very short time—then there is a possibility of using cheats to speed up the game. In an optional implementation, statistical analysis methods, such as standard deviation analysis, can also be used to determine whether the task completion time is abnormal.

[0062] S9. If the first monitoring result, the second monitoring result, and the third monitoring result are all normal behavior, then the player is determined to be a normal player; otherwise, the player is added to the suspected cheater list.

[0063] Based on the list of suspected cheaters, the game operations team can conduct further investigations and verifications of suspected cheaters. This can be done through methods such as manually reviewing game recordings and checking player account data. If it is confirmed that a player has used cheats, appropriate penalties will be imposed, such as account bans or login restrictions, in order to maintain the fairness and normal order of the game.

[0064] Please refer to Figure 2 Embodiment two of the present invention is as follows: A cheat detection terminal for an MMO game includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the various steps of the cheat detection method for the MMO game in Embodiment 1.

[0065] In summary, this invention provides a method and terminal for detecting cheats in MMO games. It collects player behavior data, including click data, skill release data, and task completion data. Based on a normal click behavior model, it monitors the click data to obtain a first monitoring result; based on a normal skill release behavior model, it monitors the skill release data to obtain a second monitoring result; and based on a normal task completion behavior model, it monitors the task completion data to obtain a third monitoring result. If all three monitoring results are normal behavior, the player is determined to be a normal player; otherwise, the player is added to a suspected cheater list. By monitoring these three types of player behavior data, it can detect cheats in MMO games. This system monitors player behavior more comprehensively from three aspects: click behavior, skill release behavior, and task completion behavior. It effectively detects situations where players are likely to use cheats in the game, thus accurately and effectively monitoring cheating behavior. Subsequently, the game operations team can further investigate and verify suspected cheating players based on the suspected cheating list, maintaining the fairness and normal order of the game. Furthermore, by collecting historical behavior data from normal players and using machine learning algorithms to build normal click behavior models, normal skill release behavior models, and normal task completion behavior models, it more closely reflects the actual game scenarios and actual player behavior, making subsequent monitoring more accurate and reliable.

[0066] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent modifications made based on the content of the present invention specification and drawings, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for detecting cheats in an MMO game, characterized in that, Including the following steps: Collect player behavior data, including click data, skill release data, and task completion data; The click data is monitored based on a normal click behavior model to obtain a first monitoring result; The skill release data is monitored based on a normal skill release behavior model to obtain a second monitoring result; The task completion data is monitored based on a normal task completion behavior model to obtain a third monitoring result; If the first monitoring result, the second monitoring result, and the third monitoring result are all normal behavior, then the player is determined to be a normal player; otherwise, the player is added to the suspected cheater list.

2. The method for detecting cheats in an MMO game according to claim 1, characterized in that, Also includes: Acquire historical behavior data of normal players, including historical click data, historical skill release data, and historical task completion data; A normal click behavior model is established using machine learning algorithms based on the historical click data. A normal skill release behavior model is established using machine learning algorithms based on the historical skill release data. Based on the historical task completion data, a normal task completion behavior model is established using machine learning algorithms.

3. The method for detecting cheats in an MMO game according to claim 1, characterized in that, The click data includes click frequency, click location, and click time interval; The first monitoring result obtained by monitoring the click data based on the normal click behavior model includes: If the click frequency exceeds the normal click frequency distribution range in the normal click behavior model, the first monitoring result is determined to be suspected cheating behavior. If it does not exceed the range, the randomness of the click position is determined to be lower than the randomness of the normal click position in the normal click behavior model. If it is lower, the first monitoring result is determined to be suspected cheating behavior. If it is not lower, the volatility of the click time interval is determined to be lower than the volatility of the normal click time interval in the normal click behavior model. If it is, the first monitoring result is determined to be suspected cheating behavior; if not, the first monitoring result is determined to be normal behavior.

4. The method for detecting cheats in an MMO game according to claim 1, characterized in that, The skill release data includes the game scene and the skill release timing, skill release order, skill release frequency, and skill combination mode corresponding to the game scene; The second monitoring result obtained by monitoring the skill release data based on the normal skill release behavior model includes: The system determines whether the timing of skill release corresponding to the game scenario conforms to the normal skill release timing in the game scenario of the normal skill release behavior model. If it does not conform, the second monitoring result is determined to be suspected cheating behavior. If it does conform, the system determines whether the skill release order corresponding to the game scenario conforms to the normal skill release order in the game scenario of the normal skill release behavior model. If it does not conform, the second monitoring result is determined to be suspected cheating behavior. If it does conform, the system determines whether the skill release frequency corresponding to the game scenario exceeds the normal skill release frequency in the game scenario of the normal skill release behavior model. If it exceeds the frequency, the second monitoring result is determined to be suspected cheating behavior. If it does not exceed the frequency, the system determines whether the skill combination pattern corresponding to the game scenario conforms to the normal skill combination pattern in the game scenario of the normal skill release behavior model. If it conforms, the second monitoring result is determined to be normal behavior. If it does not conform, the second monitoring result is determined to be suspected cheating behavior.

5. The method for detecting cheats in an MMO game according to claim 1, characterized in that, The task completion data includes the task completion time and the copy completion time. The third monitoring result obtained by monitoring the task completion data based on the normal task completion behavior model includes: If the task completion time is lower than the average task completion time or the fluctuation range of task completion time in the normal task completion behavior model, then the third monitoring result is determined to be suspected cheating behavior. If it is not lower than the average task completion time or the fluctuation range of task completion time in the normal task completion behavior model, then the third monitoring result is determined to be suspected cheating behavior. If it is not lower than the average task completion time, then the third monitoring result is determined to be normal behavior.

6. A cheat detection terminal for an MMO game, 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, it performs the following steps: Collect player behavior data, including click data, skill release data, and task completion data; The click data is monitored based on a normal click behavior model to obtain a first monitoring result; The skill release data is monitored based on a normal skill release behavior model to obtain a second monitoring result; The task completion data is monitored based on a normal task completion behavior model to obtain a third monitoring result; If the first monitoring result, the second monitoring result, and the third monitoring result are all normal behavior, then the player is determined to be a normal player; otherwise, the player is added to the suspected cheater list.

7. The anti-cheat monitoring terminal for an MMO game according to claim 6, characterized in that, Also includes: Acquire historical behavior data of normal players, including historical click data, historical skill release data, and historical task completion data; A normal click behavior model is established using machine learning algorithms based on the historical click data. A normal skill release behavior model is established using machine learning algorithms based on the historical skill release data. Based on the historical task completion data, a normal task completion behavior model is established using machine learning algorithms.

8. A cheat detection terminal for MMO games according to claim 6, characterized in that, The click data includes click frequency, click location, and click time interval; The first monitoring result obtained by monitoring the click data based on the normal click behavior model includes: If the click frequency exceeds the normal click frequency distribution range in the normal click behavior model, the first monitoring result is determined to be suspected cheating behavior. If it does not exceed the range, the randomness of the click position is determined to be lower than the randomness of the normal click position in the normal click behavior model. If it is lower, the first monitoring result is determined to be suspected cheating behavior. If it is not lower, the volatility of the click time interval is determined to be lower than the volatility of the normal click time interval in the normal click behavior model. If it is, the first monitoring result is determined to be suspected cheating behavior; if not, the first monitoring result is determined to be normal behavior.

9. A cheat detection terminal for MMO games according to claim 6, characterized in that, The skill release data includes the game scene and the skill release timing, skill release order, skill release frequency, and skill combination mode corresponding to the game scene; The second monitoring result obtained by monitoring the skill release data based on the normal skill release behavior model includes: The system determines whether the timing of skill release corresponding to the game scenario conforms to the normal skill release timing in the game scenario of the normal skill release behavior model. If it does not conform, the second monitoring result is determined to be suspected cheating behavior. If it does conform, the system determines whether the skill release order corresponding to the game scenario conforms to the normal skill release order in the game scenario of the normal skill release behavior model. If it does not conform, the second monitoring result is determined to be suspected cheating behavior. If it does conform, the system determines whether the skill release frequency corresponding to the game scenario exceeds the normal skill release frequency in the game scenario of the normal skill release behavior model. If it exceeds the frequency, the second monitoring result is determined to be suspected cheating behavior. If it does not exceed the frequency, the system determines whether the skill combination pattern corresponding to the game scenario conforms to the normal skill combination pattern in the game scenario of the normal skill release behavior model. If it conforms, the second monitoring result is determined to be normal behavior. If it does not conform, the second monitoring result is determined to be suspected cheating behavior.

10. A cheat detection terminal for MMO games according to claim 6, characterized in that, The task completion data includes the task completion time and the copy completion time. The third monitoring result obtained by monitoring the task completion data based on the normal task completion behavior model includes: If the task completion time is lower than the average task completion time or the fluctuation range of task completion time in the normal task completion behavior model, then the third monitoring result is determined to be suspected cheating behavior. If it is not lower than the average task completion time or the fluctuation range of task completion time in the normal task completion behavior model, then the third monitoring result is determined to be suspected cheating behavior. If it is not lower than the average task completion time, then the third monitoring result is determined to be normal behavior.

Images