Game replaying method, device and equipment, storage medium and program product

Abstract

The embodiment of the invention discloses a game replaying method and device, equipment, a storage medium and a program product, and the method comprises the steps: obtaining event associated data corresponding to a game key event, the event associated data at least comprising respective role state data of at least two key virtual roles in an occurrence time period of the game key event; according to role state data of a target virtual role in the event association data, determining a view field range of the target virtual role; according to the view field of the target virtual character and the character state data of the other key virtual characters, determining the observation state of the target virtual character to the other key virtual characters in the game key event; and determining view replay information corresponding to the game key event based on the observation state. By means of the method, fine-grained game replaying can be achieved, replaying efficiency is improved, and meanwhile it can be guaranteed that the replaying result has high accuracy and reliability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, and particularly relates to a game review method and device, equipment, a storage medium and a program product. BACKGROUND

[0002] The game review function is a function provided for players to review and analyze their game processes. Through the game review function, players can learn game experience, and game experience can be extended from short-term win or lose pleasure to a complete closed loop that can be traced, analyzed and learned.

[0003] In the related art, the game review function can display a data statistics panel for the player, which includes the statistical results of a plurality of key data indicators (such as virtual damage value, virtual kill amount, virtual death times, and winning rate) in the game process. The player can macroscopically understand his performance in the game process according to the data statistics panel. Alternatively, the game review function can provide a video recording of the game process for the player. The player can review the performance of himself and other players in the game process by playing back the video recording.

[0004] However, the above mechanism only supports the player to macroscopically review the game, such as reviewing the game by combining the data statistics results or the video recording of the whole game. The review granularity is coarse, and the review efficiency is low. In addition, the above mechanism often needs to rely on artificial subjective experience to further review in depth. Due to the instability of artificial subjective experience, the accuracy and reliability of the review results are often low. SUMMARY

[0005] The embodiments of the present application provide a game review method, device, equipment, storage medium and program product, which can realize fine-grained game review, improve the review efficiency, and ensure that the review results have high accuracy and reliability.

[0006] The first aspect of the present application provides a game review method, which comprises:

[0007] Obtaining event association data corresponding to a game key event, the game key event being a preset type event occurring between at least two key virtual characters, and the event association data at least including role state data of the at least two key virtual characters in an occurrence time period of the game key event;

[0008] Determining a field of view range of a target virtual character according to role state data of the target virtual character in the event association data, the target virtual character being any one of the key virtual characters;

[0009] determine an observation state of the target virtual character to other key virtual characters in the game key event according to the target virtual character's field of view range and the role state data of other key virtual characters;

[0010] determine the field of view review information corresponding to the game key event based on the observation state.

[0011] The second aspect of the application provides a game review method, the method comprising:

[0012] display a virtual game map, and mark target positions of at least two key virtual characters involved in a game key event in the virtual game map, the target positions being positions of the key virtual characters when the game key event occurs;

[0013] in response to a first operation triggered based on a target position of a target virtual character in the virtual game map, display event review information corresponding to the game key event, the target virtual character being any one of the key virtual characters, the event review information at least including field of view review information, the field of view review information being used to indicate an observation state of the target virtual character to other key virtual characters in the game key event.

[0014] The third aspect of the application provides a game review device, the device comprising:

[0015] a data acquisition module configured to acquire event association data corresponding to a game key event, the game key event being a preset type event occurring between at least two key virtual characters, the event association data at least including role state data of the at least two key virtual characters respectively within an occurrence time period of the game key event;

[0016] a field of view determination module configured to determine a field of view range of a target virtual character according to role state data of the target virtual character in the event association data, the target virtual character being any one of the key virtual characters;

[0017] an observation state determination module configured to determine an observation state of the target virtual character to other key virtual characters in the game key event according to the target virtual character's field of view range and the role state data of other key virtual characters;

[0018] a review information determination module configured to determine field of view review information corresponding to the game key event based on the observation state.

[0019] The fourth aspect of the application provides a game review device, the device comprising:

[0020] A map display module is used to display a virtual game map and mark the target locations of at least two key virtual characters involved in key game events on the virtual game map. The target locations are the positions of the key virtual characters when the key game events occur.

[0021] The recap information display module is used to respond to a first operation triggered based on the target position of the target virtual character in the virtual game map, and to display the event recap information corresponding to the key game event. The target virtual character is any one of the key virtual characters. The event recap information includes at least vision recap information, which is used to indicate the observation status of the target virtual character towards other key virtual characters in the key game event.

[0022] A fifth aspect of this application provides a computer device, the device comprising a processor and a memory:

[0023] The memory is used to store computer programs;

[0024] The processor is configured to perform the steps of the game replay method as described in the first or second aspect above, according to the computer program.

[0025] The sixth aspect of this application provides a computer-readable storage medium for storing a computer program for performing the steps of the game replay method described in the first or second aspect above.

[0026] A seventh aspect of this application provides a computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps of the game replay method described in the first or second aspect.

[0027] As can be seen from the above technical solutions, the embodiments of this application have the following advantages:

[0028] This application provides a game replay method that innovatively proposes a fine-grained replay scheme based on key game events. This scheme can help players objectively replay the visual observation of virtual characters during key game events. Specifically, in this method, event association data corresponding to key game events occurring between at least two key virtual characters can be obtained first. This data includes at least the character state data of each of the at least two key virtual characters during the occurrence of the key game event. The character state data can indicate, for example, the position and orientation of the key virtual characters. Then, based on the character state data of the target virtual character (i.e., any one of the key virtual characters), the visual field of the target virtual character can be determined. Based on the visual field and the character state data of other key virtual characters, the observation state of the target virtual character towards other key virtual characters during the key game event can be determined. Furthermore, based on this observation state, visual replay information corresponding to the key game event can be determined. The aforementioned method, on the one hand, focuses the review on key game events during the game process. It determines review information based on the character status data of key virtual characters during the occurrence of these key events, thus providing players with fine-grained event-level review information. This allows players to efficiently review the game using review information related to key game events, without wasting time and energy focusing on other unimportant game events. On the other hand, the method objectively constructs the target virtual character's field of vision and accurately determines the target virtual character's observation status of other key virtual characters during key game events. The resulting field of vision review information reliably reflects the target virtual character's field of vision observation during key game events. Compared to related technologies where players rely on their subjective experience to judge field of vision observation, this application provides players with objective and accurate field of vision review information, ensuring the accuracy and reliability of the review results. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is a schematic diagram of a data statistics panel in related technologies;

[0031] Figure 2 A schematic diagram of the structure of a computer system provided in an embodiment of this application;

[0032] Figure 3 This is a schematic diagram illustrating an application scenario of the game replay method provided in the embodiments of this application;

[0033] Figure 4 A flowchart illustrating the game replay method on the backend side provided in this application embodiment;

[0034] Figure 5 A schematic diagram illustrating the field of view of the target virtual character provided in an embodiment of this application;

[0035] Figure 6 This is a flowchart illustrating the front-end game replay method provided in an embodiment of this application.

[0036] Figure 7 A schematic diagram of a virtual game map provided in an embodiment of this application;

[0037] Figure 8 A schematic diagram of another virtual game map provided in an embodiment of this application;

[0038] Figure 9 A schematic diagram of a review information interface provided in an embodiment of this application;

[0039] Figure 10 A schematic diagram of a death thermal layer provided in an embodiment of this application;

[0040] Figure 11 A schematic diagram illustrating an event recording video provided in an embodiment of this application;

[0041] Figure 12 A schematic diagram of a support trajectory provided for an embodiment of this application;

[0042] Figure 13 This is a schematic diagram of the structure of a game replay device provided in an embodiment of this application;

[0043] Figure 14 This is a schematic diagram of another game replay device provided in an embodiment of this application;

[0044] Figure 15 This is a schematic diagram of the structure of the terminal device provided in the embodiments of this application;

[0045] Figure 16 This is a schematic diagram of the server structure provided in an embodiment of this application. Detailed Implementation

[0046] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0047] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0048] It should be noted that this application may display prompt interfaces, pop-ups, or output voice prompts before and during the collection of user data. These prompt interfaces, pop-ups, or voice prompts are used to inform the user that their data is being collected. This ensures that the application only begins the steps for collecting user data after receiving confirmation from the user regarding the prompt interface or pop-up; otherwise (i.e., without user confirmation), the steps for collecting user data end, meaning no user data is collected. In other words, all user data collected in this application is collected with the user's consent and authorization, and the collection, use, and processing of related user data must comply with the relevant laws, regulations, and standards of the relevant countries and regions.

[0049] In related technologies, data statistics panels or game recording videos are primarily used to help players review their gameplay. See also Figure 1 , Figure 1 This is a schematic diagram of a data statistics panel in related technologies, such as... Figure 1 As shown, this includes statistical results for various key data indicators, specifically including season score, highest score, average performance parameters (kill / death ratio, kill / death / assist ratio, damage per game), number of wins, number of kills, number of deaths, win rate, top ten rate, and total number of games participated in, all based on statistics from multiple games the player has recently participated in. It also includes individual performance parameters for each game the player has recently participated in, such as ranking, number of kills, damage output, and survival time. Alternatively, players can request a recording of their gameplay. After the game, the recorded video can be replayed, and key moments in the video can be located and analyzed to perform a game review.

[0050] The virtual character's field of vision is a crucial piece of information during game replays. The virtual character's field of vision during significant game events (such as kills) severely impacts player strategy and ultimately determines the game's trajectory and outcome. Currently, the data analytics panel does not provide this dimension of replay information. Therefore, replaying the virtual character's field of vision requires manually replaying the game video, locating the key events within the video, and analyzing the corresponding video content, combined with the player's subjective experience, to determine and replay the virtual character's field of vision.

[0051] The aforementioned debriefing methods have two main drawbacks. First, they require manually locating key game events during gameplay replays, necessitating the viewing of a large amount of irrelevant video content, resulting in low efficiency and wasting players' time and energy. Second, players need to combine their subjective experience to determine the debriefing results regarding their observations of the virtual character's field of vision. Given the instability of players' subjective experience, the accuracy and reliability of the final debriefing results are often low.

[0052] To address the aforementioned issues, this application provides a game replay method. This method innovatively proposes a fine-grained replay scheme based on key game events. This scheme helps players objectively replay the virtual character's field of vision during key game events. Specifically, this method focuses the replay on key game events during the game process. Based on the character status data of the key virtual character (the virtual character participating in the key game event) during the occurrence of the key game event, replay information is determined, thereby providing players with event-level fine-grained replay information. This allows players to efficiently replay the game using replay information related to key game events, without wasting time and energy focusing on other unimportant game events. Furthermore, this method can objectively construct the field of vision of any key virtual character and accurately determine the observation status of the key virtual character towards other key virtual characters during key game events, thereby obtaining field of vision recap information. This field of vision recap information can reliably reflect the field of vision observation of the key virtual character during key game events. Compared with related technologies where players judge the field of vision observation based on their personal subjective experience, this application can provide players with objective and accurate field of vision recap information, thus ensuring that the recap results have high accuracy and reliability.

[0053] To facilitate understanding of the game replay method provided in this application embodiment, the system architecture used to implement the game replay method will be introduced below.

[0054] See Figure 2 ,Figure 2 This is a schematic diagram of the structure of a computer system provided in an embodiment of this application. The computer system 200 can serve as a system architecture for implementing the game replay method in the embodiment of this application. The computer system 200 includes a terminal device 220 and a server 240.

[0055] The terminal device 220 has a client application for the target application installed and running. This target application can be any program that provides game replay functionality. Specifically, it can be a dedicated game replay application, or a game application that integrates game replay functionality. Examples include, but are not limited to, strategy games (SLG), first-person shooter / third-person shooter (FPS / TPS), multiplayer online battle arena (MOBA), action games (ACT), and role-playing games (RPG). This embodiment does not limit the type of the target application. The client application for the target application can be, for example, an application (App) or app on the terminal device 220, or a webpage. This embodiment does not limit the form of the client application for the target application.

[0056] Terminal device 220 is a user-facing terminal device, which can be any user of the target application. The front-end game replay method provided in this application embodiment can be implemented through terminal device 220. Specifically, the client of the target application running in terminal device 220 can display a virtual game map and mark the target positions of at least two key virtual characters involved in a key game event on the virtual game map. The target position here refers to the location of the key virtual character when the key game event occurs. In response to a first operation triggered based on the target position of the target virtual character (any key virtual character) on the virtual game map, the client of the target application will display event replay information corresponding to the key game event. This event replay information includes at least field-of-view replay information, which indicates the observation status of the target virtual character towards other key virtual characters during the key game event.

[0057] Terminal device 220 can refer to one of a plurality of terminal devices, and this embodiment only uses terminal device 220 as an example. The device type of terminal device 220 may include, but is not limited to, at least one of the following: smartphone, tablet computer, wearable device, personal computer (PC), laptop computer, and desktop computer. Those skilled in the art will know that the number of the above-mentioned terminal devices 220 may be more or less. For example, the above-mentioned terminal device 220 may be only one, or there may be multiple or more terminal devices 220. This application embodiment does not limit the number or device type of terminal devices 220 in any way.

[0058] Terminal device 220 can be connected to server 240 via wireless network or wired network.

[0059] Server 240 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides cloud computing services such as cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and basic cloud computing services such as big data and artificial intelligence platforms.

[0060] Server 240 provides background services to the client of the target application running on terminal device 220. Server 240 includes processor 244 and memory 242. Memory 242 stores related computer programs, such as a computer program for executing the background game replay method provided in the embodiments of this application. Processor 244 executes the computer program stored in memory 242 to execute the background game replay method provided in the embodiments of this application, providing game replay information to the client of the target application running on terminal device 220. Specifically, the processor 244 can acquire event association data corresponding to key game events. Key game events are preset type events that occur between at least two key virtual characters. The event association data includes at least the character status data of each of the at least two key virtual characters during the occurrence period of the key game event. Then, the processor 244 can determine the field of view of the target virtual character (any key virtual character) based on the character status data of the target virtual character, and determine the observation state of the target virtual character towards other key virtual characters during the key game event based on the field of view of the target virtual character and the character status data of other key virtual characters. Furthermore, the processor 244 can determine the field of view recap information corresponding to the key game event based on this observation state.

[0061] Optionally, server 240 may undertake the primary computing task, while terminal device 220 may undertake the secondary computing task; alternatively, server 240 may undertake the secondary computing task, while terminal device 220 may undertake the primary computing task; or, server 240 and terminal device 220 may collaborate using a distributed computing architecture. This embodiment does not impose any limitations on the cooperation method between terminal device 220 and server 240.

[0062] See Figure 3 , Figure 3 This is a schematic diagram illustrating an application scenario of the game replay method provided in this application embodiment. For example... Figure 3 As shown, the game replay method on the backend side provided in this application embodiment can be generated by... Figure 2 The server 240 shown in this application executes the front-end game replay method provided in this embodiment, which can be executed by... Figure 2 The terminal device 220 shown is executed.

[0063] Specifically, when server 240 determines the replay information for a game, it can obtain event-related data corresponding to key game events. These key game events are preset-type events occurring between at least two key virtual characters, such as kill events, attack events, and healing events. The event-related data includes at least the character state data of these two key virtual characters during the time period of the key game event. This character state data can, for example, indicate the position and posture of the key virtual characters. Figure 3 As shown, taking the killing event of key virtual character a against key virtual character b as an example, the acquired event-related data 310 may include the position data and orientation angle data of key virtual character a and key virtual character b during the time period of the killing event.

[0064] Next, server 240 can select any one of the at least two key virtual characters mentioned above as the target virtual character, and determine the target virtual character's field of vision during the period in which the key game event occurs based on the target virtual character's character status data. For example, as shown... Figure 3 As shown, key virtual character a and key virtual character b can be taken as target virtual characters, and their field of vision ranges 321 and 322 can be determined. Specifically, when determining the field of vision range of the target virtual character, a field of vision cone can be constructed based on the orientation angle data and field of vision distance threshold of the target virtual character. The range covered by the field of vision cone is the field of vision range of the target virtual character.

[0065] Furthermore, server 240 can determine the target virtual character's observation status towards the key virtual character during a key game event based on the target virtual character's field of vision and the character status data of other key virtual characters (any key virtual character other than the target virtual character). This observation status can indicate whether the target virtual character has detected the key virtual character. For example, such as... Figure 3 As shown, the observation state 331 of key virtual character a towards key virtual character b can be determined based on the field of view 321 of key virtual character a and the character status data of key virtual character b. Alternatively, the observation state 332 of key virtual character b towards key virtual character a can be determined based on the field of view 322 of key virtual character b and the character status data of key virtual character a.

[0066] Ultimately, server 240 can determine the visual recap information corresponding to the key game event based on the observation status of the target virtual character towards other key virtual characters during the key game event. For example, such as... Figure 3 As shown, taking the kill of key virtual character b by key virtual character a as an example, if the observation state 331 of key virtual character a towards key virtual character b and the observation state 332 of key virtual character b towards key virtual character a have been determined when the kill event occurs, the vision recap information 340 corresponding to the kill event can be determined based on these two observation states. The vision recap information 340 can indicate, for example, whether key virtual character a and key virtual character b discovered each other in a two-way manner, or discovered each other in a one-way manner, or did not discover each other in the kill event.

[0067] The replay information determined through the above method can be viewed by the player using the client of the target application; that is, the player can trigger the viewing of the replay information corresponding to the game they participated in through the client of the target application running on terminal device 220. When the target application displays the replay information corresponding to the game to the player, it can first display a virtual game map, which is a virtual scene map corresponding to the game the player participated in. Furthermore, it can mark the target locations of at least two key virtual characters involved in key game events on this virtual game map. These target locations are the positions of the key virtual characters when the key game events occurred. For example, as shown... Figure 3As shown, taking kill events as a key game event as an example, the virtual game map 350 marks the target locations of key virtual characters involved in each kill event that occurs during the game. For example, the black circle and white circle numbered 1 in the virtual game map 350 indicate the target locations of the key virtual character who is the victim and the key virtual character who is the killer in the first kill event that occurs in the game, respectively. The black circle and white circle numbered 2 in the virtual game map 350 indicate the target locations of the key virtual character who is the victim and the key virtual character who is the killer in the second kill event that occurs in the game, respectively, and so on.

[0068] The client of the target application in the terminal device 220 can respond to the first operation triggered by the player based on the target location of the target virtual character (any key virtual character in any key game event) in the virtual game map, and display the event recap information corresponding to the key game event to which the target virtual character belongs. The event recap information includes at least the field of view recap information, which is the field of view recap information determined by the server 240 as described above. It can indicate the observation status of the target virtual character towards other key virtual characters in the key game event. For example, a player can click on the black circle marked 1 on the virtual scene map 350 to trigger a first operation based on the target location indicated by the black circle (corresponding to the target virtual character). In response to this first operation, the client of the target application will display a replay information interface 360, which includes event replay information corresponding to the first kill event that occurred in the game. This event replay information includes vision replay information 361 corresponding to the first kill event, which is used to indicate the observation status of the two key virtual characters, as the killer and the killed, towards each other in the first kill event. Optionally, the event replay information may also include basic information 362 of the two key virtual characters, as the killer and the killed (such as nickname, location, health changes, virtual equipment used, stance type, whether they are inside virtual buildings, etc.), and may also include the team evaluation results 363 of the teams to which the two key virtual characters belong at the time of the first kill event (such as team formation, overall score, team cohesion, coverage, tactical effectiveness, etc.), as well as corresponding tactical suggestions 364.

[0069] It should be understood that Figure 3 The application scenarios shown are merely examples. In practical applications, the game replay method provided in this application embodiment can also be applied to other scenarios. No limitation is made here on the application scenarios of the game replay method provided in this application embodiment.

[0070] The game replay method provided in this application will be described in detail below through method embodiments.

[0071] See Figure 4 , Figure 4 This is a flowchart illustrating a background-side game replay method provided in an embodiment of this application. This method is executed by a computer device, such as a terminal device or a server. Figure 4 As shown, the game replay method includes the following steps:

[0072] S401: Obtain event-related data corresponding to key game events.

[0073] A key game event is a preset type of event that occurs between at least two key virtual characters, where these at least two key virtual characters are the virtual characters participating in the key game event. For example, the key game event can be determined according to the game type. For instance, in SLG games, FPS / TPS games, or MOBA games, a kill event can be considered a key game event, and correspondingly, the virtual character who kills and the virtual character who is killed in the kill event are the aforementioned key virtual characters. Of course, in practical applications, other types of game events, such as attack events and healing events, can also be considered key game events. This application embodiment does not limit the type of key game event in any way.

[0074] Event-related data refers to game data of a preset dimension related to key game events. It can include, at a minimum, the character status data of at least two key virtual characters during the occurrence period of the key game event. The occurrence period of the key game event can be determined based on the exact moment of the event's occurrence; for example, 10 seconds before and after the moment of the key game event can be considered the occurrence period. The character status data of the key virtual characters reflects their state during the occurrence period of the key game event. This data can include, for example, the position data and orientation angle data of the key virtual character at at least one moment during the occurrence period. The position data can specifically be represented as the three-dimensional coordinates of the key virtual character in the virtual game scene, and the orientation angle data can specifically be represented as the Euler angle data of the key virtual character's camera, including yaw, pitch, and roll.

[0075] Optionally, in addition to including the role status data of each of the above-mentioned at least two key virtual characters, the event-related data may also include other data, such as the role status data of other virtual characters in the team to which each key virtual character belongs during the occurrence period of the game's key event, or the game environment data at the location where the game's key event occurs, etc. This application embodiment does not limit the specific content included in the event-related data.

[0076] For example, taking a kill event as a key game event, the game's telemetry application programming interface (API) can be called to capture kill events that occur during the game. Relevant game data within 10 seconds before and after the kill event can be obtained as event-related data corresponding to the kill event. For example, the position and orientation angle data of the key virtual character who kills, and the position and orientation angle data of the key virtual character who is killed can be obtained to form the event-related data corresponding to the kill event.

[0077] S402: Determine the field of view of the target virtual character based on the character status data of the target virtual character in the event association data.

[0078] The target virtual character can be any key virtual character. For example, a key virtual character with a preset identity can be selected from at least two key virtual characters as the target virtual character. For instance, when a key game event is a kill event, the key virtual character who is the one being killed can be the target virtual character. Alternatively, each key virtual character can be used as the target virtual character separately. For instance, when a key game event is a kill event, the key virtual character who kills and the key virtual character who is the one being killed can be used as target virtual characters respectively. This application does not limit the target virtual character in any way.

[0079] The field of view of a target virtual character is used to indicate the range that the target virtual character can see. It is usually in the form of a three-dimensional field of view cone, such as a three-dimensional cone with the location of the target virtual character as the origin, the yaw angle direction as the central axis, and a fixed field of view angle.

[0080] For example, see Figure 5 , Figure 5This diagram illustrates the field of view of a target virtual character provided in an embodiment of this application. (a) indicates a top view of the field of view, (b) indicates a side view of the field of view, and (c) indicates a three-dimensional perspective view of the field of view. Without considering obstructions or the use of a virtual high-powered lens by the target virtual character, the complete field of view of the target virtual character can cover an observation range of 200° to 220° in the horizontal direction. In the vertical direction, with the yaw angle of the target virtual character as the central axis, the field of view of the target virtual character can cover an observation range of 60° above and 75° below. Furthermore, since the target virtual character has a corresponding viewing distance, the field of view of the target virtual character is typically determined by combining this viewing distance with the aforementioned observation range angles. It should be understood that if the distance between the observed target and the target virtual character is greater than the visible distance, or if the angle of the observed target relative to the target virtual character exceeds the aforementioned observation angle range, then the target virtual character can be considered unable to observe the observed target; conversely, if the distance between the observed target and the target virtual character is less than or equal to the visible distance, or if the angle of the observed target relative to the target virtual character does not exceed the aforementioned observation angle range, then the target virtual character can be considered able to observe the observed target.

[0081] For example, when determining the field of view of a target virtual character, the aforementioned three-dimensional field of view cone can be constructed based on the position data and orientation angle data in the character state data of the target virtual character. It should be understood that if the event-related data only includes the character state data of the target virtual character at one moment, then only the field of view of the target virtual character at that moment needs to be constructed; if the event-related data includes the character state data of the target virtual character at multiple moments, then the field of view of the target virtual character at each of those moments needs to be constructed.

[0082] S403: Based on the target virtual character's field of vision and the character status data of other key virtual characters, determine the target virtual character's observation status of other key virtual characters during key game events.

[0083] The observation status of the target virtual character towards other key virtual characters indicates whether the target virtual character has detected the key virtual character during the occurrence of a key game event. It should be understood that if the key virtual character falls within the target virtual character's field of vision, it can be assumed that the target virtual character has detected the key virtual character during the occurrence of the key game event; conversely, if the key virtual character is outside the target virtual character's field of vision, it can be assumed that the target virtual character has not detected the key virtual character during the occurrence of the key game event.

[0084] For example, when specifically determining the observation state of a target virtual character towards other key virtual characters, the position data of the key virtual character can be extracted from the role status data of the other key virtual characters. Then, based on the position data of the target virtual character and the position data of the key virtual character, the relative positional relationship between the target virtual character and the key virtual character can be determined. Then, based on the field of view of the target virtual character and the relative positional relationship, it can be determined whether the key virtual character falls within the field of view of the target virtual character, thereby obtaining the observation state of the target virtual character towards the key virtual character.

[0085] It should be understood that determining the above observation states requires ensuring time alignment between data; that is, it requires ensuring that the field of view of the target virtual character, the position data of the target virtual character, and the position data of the key virtual characters used in the calculation all correspond to the same moment. Furthermore, through the above method, the observation state of the target virtual character towards other key virtual characters at least at one moment within the occurrence period of a key game event can be determined.

[0086] S404: Based on the observation status, determine the visual replay information corresponding to key game events.

[0087] Vision recap information is used to indicate the visual observation status of the target virtual character during the period of a critical game event. For example, it can indicate whether the target virtual character discovered other key virtual characters when the critical game event occurred.

[0088] For example, when only key virtual characters with preset identities are used as target virtual characters, the observation status of the target virtual character towards other key virtual characters can be directly used as the aforementioned vision recap information; for example, when the target virtual character is a key virtual character who is killed in a kill event, the observation status of the target virtual character towards other key virtual characters who are killers can be directly used as the vision recap information corresponding to the kill event.

[0089] Alternatively, when each key virtual character is taken as a target virtual character, the observation status of each key virtual character towards other key virtual characters can be comprehensively considered to summarize the visual recap information of the key events in the game. For example, when the key virtual character who is the killer and the key virtual character who is the killer in a kill event are taken as target virtual characters, the visual recap information of the kill event can be summarized based on the observation status of the key virtual character who is the killer towards the key virtual character who is the killer, and the observation status of the key virtual character who is the killer towards the key virtual character who is the killer. This visual recap information can indicate whether the key virtual character who is the killer and the key virtual character who is the killer have a two-way discovery, a one-way discovery, or no discovery of each other.

[0090] The game replay method provided in this application innovatively proposes a fine-grained replay scheme based on key game events. This scheme helps players objectively replay the virtual character's field of vision during key game events. Specifically, this method focuses the replay on key game events during the game process. Based on the character status data of the key virtual character (the virtual character participating in the key game event) during the occurrence of the key game event, replay information is determined, thereby providing players with event-level fine-grained replay information. This allows players to efficiently replay the game using replay information related to key game events, without wasting time and energy focusing on other unimportant game events during the game. Furthermore, this method can objectively construct the field of vision of any key virtual character and accurately determine the observation status of the key virtual character towards other key virtual characters during key game events, thereby obtaining field of vision recap information. This field of vision recap information can reliably reflect the field of vision observation of the key virtual character during key game events. Compared with related technologies where players judge the field of vision observation based on their personal subjective experience, this application can provide players with objective and accurate field of vision recap information, thus ensuring that the recap results have high accuracy and reliability.

[0091] Mechanism for determining the field of view observation situation

[0092] • Determination mechanism of field of view range

[0093] In one possible implementation, the character status data of the key virtual character in the aforementioned event-related data may specifically include the key virtual character's position data and orientation angle data. The position data indicates the key virtual character's position in the virtual game scene, and can be represented, for example, as the key virtual character's three-dimensional coordinate data. The orientation angle data indicates the key virtual character's orientation in the virtual game scene, and can be represented, for example, as Euler angle data, specifically including yaw angle, pitch angle, and roll angle. The yaw angle indicates the key virtual character's orientation rotation about the Y-axis in the three-dimensional coordinate system, the pitch angle indicates the key virtual character's orientation rotation about the X-axis in the three-dimensional coordinate system, and the roll angle indicates the key virtual character's orientation rotation about the Z-axis in the three-dimensional coordinate system.

[0094] When the character status data of the key virtual character includes the position data and orientation angle data of the key virtual character, the field of view of the target virtual character can be determined by the following S11, that is, the above S402 includes the following S11 (not shown in the figure):

[0095] S11: Determine the orientation vector of the target virtual character based on the orientation angle data of the target virtual character; determine the field of view range of the target virtual character based on the position data, orientation vector, and field of view distance threshold of the target virtual character.

[0096] The orientation vector of the target virtual character is used to indicate the relationship between the orientation of the target virtual character and the origin of the three-dimensional coordinate system of the virtual game scene in the form of a vector. Specifically, it can include a forward vector determined by the yaw angle and pitch angle, a rightward vector determined by the yaw angle, and a preset upward vector.

[0097] For example, taking the three-dimensional coordinate system of the virtual game scene as a left-handed coordinate system (with the positive X-axis pointing to the right, the positive Y-axis pointing forward, and the positive Z-axis pointing upward) built on the fifth generation Unreal Engine (UE5), the position data of the target virtual character i is obtained. And the Euler angle data (yaw) of the target virtual character i. i pitch i and roll i Then, the yaw can be processed using the following formulas (1) and (2). i pitch i Convert them to their corresponding radians respectively. and :

[0098] (1)

[0099] (2)

[0100] Then, the forward vector f of the target virtual character i is determined by the following equations (3) to (5). i Right-directed vector r i and upward vector u i :

[0101] (3)

[0102] (4)

[0103] (5)

[0104] Calculate the forward vector f i At that time, UE5 agreed to first yaw around the Z-axis. i Then pitch around the Y-axis i Calculate the rightward vector r. i When, ignore pitch i and roll i Only considering yaw iCalculate the upward vector u. i Considering that virtual characters in games can typically only turn their heads left and right, the upward vector u can be directly preset. i Pointing directly upwards.

[0105] After calculating the orientation vector of the target virtual character using the above method, the target virtual character's three-dimensional coordinate data, the orientation vector, and the field of view distance threshold can be combined to determine the target virtual character's field of view range. For example, the position indicated by the target virtual character's three-dimensional coordinate data can be used as the vertex of a three-dimensional field of view cone, the direction indicated by the orientation vector can be used as the centerline direction of the three-dimensional field of view cone, and the field of view distance threshold can be used as the side length of the three-dimensional field of view cone. Combined with the field of view range angle of the target virtual character in the game (pre-set), a three-dimensional field of view cone representing the field of view range of the target virtual character can be constructed.

[0106] It should be noted that the aforementioned field of view distance threshold is a parameter indicating the observation distance of the target virtual character. It can be the observation distance set for the target virtual character during game development, or it can dynamically change based on the virtual game environment or virtual game items used by the target virtual character during the occurrence of a critical game event. For example, if there are obstructions around the target virtual character during the occurrence of a critical game event, or if the target virtual character is in a harsh virtual weather environment, the observation distance of the target virtual character can be reduced to obtain the field of view distance threshold. As another example, if the target virtual character uses a virtual high-powered scope during the occurrence of a critical game event, the observation distance of the target virtual character can be increased based on the performance of the virtual high-powered scope to obtain the field of view distance threshold. This application does not specifically limit the field of view distance threshold in this embodiment.

[0107] In this way, by combining the position data and orientation angle data of the target virtual character, the field of vision of the target virtual character can be accurately and reliably determined. Furthermore, by combining the position data, orientation angle data, and field of vision distance range, the field of vision range of the target virtual character can be accurately described, providing reliable data support for subsequently determining the observation status of the target virtual character towards other key virtual characters.

[0108] • Determination mechanism of observation state

[0109] In one possible implementation, the observation state of the target virtual character towards the key virtual character can be determined based on the relative positional relationship between the key virtual character and the target virtual character. That is, S403 above may include the following S21 to S23 (not shown in the figure):

[0110] S21: Determine the relative position information between the target virtual character and the key virtual character based on the position data of the target virtual character and the key virtual character.

[0111] The relative position information between the target virtual character and the key virtual character is used to indicate the relative positional relationship between the target virtual character and the key virtual character in the virtual game scene. Specifically, it may include relative direction and relative distance. The relative direction is used to indicate the direction of the line connecting the positions of the target virtual character and the key virtual character in the virtual game scene, and the relative distance is used to indicate the length of the line connecting the positions of the target virtual character and the key virtual character.

[0112] For example, the relative position information between the target virtual character i and the key virtual character j can be calculated using the following equations (6) to (8):

[0113] (6)

[0114] (7)

[0115] (8)

[0116] in, Let be the relative vector between the target virtual character i and the key virtual character j. Let i be the relative distance between the target virtual character i and the key virtual character j. The normalized relative direction between the target virtual character i and the key virtual character j.

[0117] S22: Determine the relative projection data of the key virtual character to the target virtual character based on the relative direction and the orientation vector of the target virtual character.

[0118] The relative projection data of the key virtual character to the target virtual character is used to indicate the observation vector data obtained by projecting the key virtual character onto the target virtual character's observation coordinate system. Specifically, based on the target virtual character's orientation vector and position data, an observation coordinate system for the target virtual character can be constructed. This observation coordinate system has its origin at the position indicated by the target virtual character's position data, and uses the rightward, forward, and upward vectors of the target virtual character's orientation vector as the X-axis, Y-axis, and Z-axis, respectively. According to the relative direction between the target virtual character and the key virtual character, the key virtual character can be projected onto the target virtual character's observation coordinate system to obtain the corresponding relative projection data. Based on this relative projection data, it is possible to more accurately analyze whether the key virtual character falls within the target virtual character's field of view.

[0119] For example, the relative projection data of the key virtual character j to the target virtual character i can be calculated using the following equations (9) to (11):

[0120] (9)

[0121] (10)

[0122] (11)

[0123] in, Let be the projection component of the key virtual character j onto the X-axis of the observation coordinate system of the target virtual character i. Let be the projection component of the key virtual character j onto the Y-axis of the observation coordinate system of the target virtual character i. The projection component of the key virtual character j onto the Z-axis of the observation coordinate system of the target virtual character i is defined by the projection components along the X, Y, and Z axes. These projection components together constitute the relative projection data of the key virtual character j onto the target virtual character i.

[0124] S23: Based on relative projection data and observation angle threshold, as well as relative distance and field of view distance threshold, determine the observation status of the target virtual character towards the key virtual character.

[0125] The observation angle threshold is used to indicate the observation angle range of the target virtual character. Specifically, it can include the observation angle range of the target virtual character in the horizontal direction and the observation angle range in the vertical direction. The observation angle range can be preset for the target virtual character, or it can change dynamically according to the virtual game environment in which the target virtual character is located or the virtual game equipment it is using during the occurrence of key game events.

[0126] Based on relative projection data and observation angle thresholds, it can be determined whether the direction of the key virtual character relative to the target virtual character is within the observation angle range of the target virtual character. Based on relative distance and field-of-view distance thresholds, it can be determined whether the distance between the key virtual character and the target virtual character exceeds the field-of-view observation distance of the target virtual character. Combining the results of the above two aspects, the observation status of the target virtual character towards the key virtual character can be determined, that is, whether the target virtual character can detect the key virtual character.

[0127] As an example, the angle between the key virtual character j and the target virtual character i in the horizontal direction and the angle in the vertical direction can be calculated first using the following equations (12) and (13):

[0128] (12)

[0129] (13)

[0130] in, Let be the horizontal angle between the relative directions of the key virtual character j and the target virtual character i. The angle between the relative directions of the key virtual character j and the target virtual character i in the vertical direction.

[0131] Furthermore, the following formula (14) can be used to determine whether the key virtual character j is within the field of vision of the target virtual character i, that is, to determine the observation state of the target virtual character i towards the key virtual character j:

[0132] (14)

[0133] in, Indicates the observation state of target virtual character i towards key virtual character j; if , ,as well as If all three conditions are met, then This means that if the key virtual character j is within the field of vision of the target virtual character i, the target virtual character i can detect the key virtual character j; conversely, if... , ,as well as If at least one of the three conditions is not met, then This means that the key virtual character j is not within the field of vision of the target virtual character i, and the target virtual character i has failed to discover the key virtual character j.

[0134] in, The above field of view distance threshold is usually set to 100 meters. and The values ​​are determined based on the horizontal observation angle threshold (horizontal_fov) and the vertical observation angle threshold (vertical_fov), respectively. , ; Typically between 70° and 100° Typically, it is between 45° and 80°; it should be understood that if the buffer is to reduce jitter, the above observation angle threshold can be appropriately increased by 2° to 5°.

[0135] As another example, the observation state of the target virtual character i with respect to the key virtual character j can also be determined through an equivalent non-reversible triangular rectangular pyramid determination mechanism.

[0136] Considering ,and Therefore, the observation state of the target virtual character i for the key virtual character j can be determined by the three conditions indicated by the following equations (15) to (17):

[0137] (15)

[0138] (16)

[0139] (17)

[0140] If all three conditions are met, it can be determined that the key virtual character j is within the field of vision of the target virtual character i, and the target virtual character i can discover the key virtual character j; otherwise, if at least one condition is not met, it can be determined that the key virtual character j is not within the field of vision of the target virtual character i, and the target virtual character i has failed to discover the key virtual character j.

[0141] It should be noted that, if Too small, such as (e.g. 10) -5 If cm), it can be directly determined that the key virtual character j is not within the field of vision of the target virtual character i, so as to avoid the situation of division by 0 during the calculation process and the occurrence of calculation errors.

[0142] In this way, by using the above method, we can determine whether the key virtual character falls within the target virtual character's field of vision from two dimensions: the observation angle and the observation distance of the target virtual character towards the key virtual character. Based on this, we can determine the target virtual character's observation status towards the key virtual character, thereby ensuring that the determined observation status is objective and accurate.

[0143] • Determination mechanism of field of view review information

[0144] In one possible implementation, at least two key virtual characters in a game's key event may include a first target virtual character and a second target virtual character. That is, a game key event may involve two key virtual characters. When determining the field of view recap information, these two key virtual characters can be respectively designated as target virtual characters (the first target virtual character and the second target virtual character), thereby determining the observation state of each key virtual character towards the other key virtual characters in the game key event. For example, a game key event may be a kill event, the first target virtual character may be the key virtual character who is the killer in the kill event, and the second target virtual character may be the key virtual character who is the killer in the kill event.

[0145] Accordingly, the visual replay information corresponding to the aforementioned key game events can be determined through the following S31 to S34, that is, the aforementioned S404 may include the following S31 to S34 (not shown in the figure):

[0146] S31: If the first target virtual character's first observation state of the second target virtual character indicates that the first target virtual character has discovered the second target virtual character, and the second target virtual character's second observation state of the first target virtual character indicates that the second target virtual character has discovered the first target virtual character, then the vision recap information indicates that the first target virtual character and the second target virtual character have discovered each other.

[0147] For example, the first observation state of the first target virtual character i towards the second target virtual character j, and the second observation state of the second target virtual character j towards the first target virtual character i, can be determined by the following equations (18) and (19), respectively:

[0148] (18)

[0149] (19)

[0150] in, This is the first observation state. This is the second observation state. and This can be determined through the observation state determination mechanism described above.

[0151] Based on the first and second observation states described above, the mutual discovery state can be further determined using the following formula (20):

[0152] (20)

[0153] in, In a mutually discovered state. If and If both are true, it means that the first target virtual character i can detect the second target virtual character j, and the second target virtual character j can also detect the first target virtual character i. In this case, This indicates that the first target virtual character i and the second target virtual character j can discover each other. Conversely, if and If at least one value is false, it means that the first target virtual character i did not find the second target virtual character j, and / or the second target virtual character j did not find the first target virtual character i. In this case, This indicates that the first target virtual character i and the second target virtual character j have not discovered each other.

[0154] It should be understood that, if Then, it can be determined that the vision recap information indicates that the first target virtual character i and the second target virtual character j have discovered each other. Taking a key game event as a kill event, with the first target virtual character and the second target virtual character being the killer and the killed in the kill event, respectively, as an example, the vision recap information corresponding to the kill event can indicate that the key virtual character as the killer and the key virtual character as the killed have discovered each other.

[0155] S32: If the first observation state indicates that the first target virtual character has discovered the second target virtual character, and the second observation state indicates that the second target virtual character has not discovered the first target virtual character, then the vision recap information indicates that the first target virtual character has unilaterally discovered the second target virtual character.

[0156] For example, in the above In this case, further judgment can be made. Is it true? This indicates that the first target virtual character i has spotted the second target virtual character j, but the second target virtual character j has not spotted the first target virtual character i. In this case, it can be determined that the vision replay information indicates that the first target virtual character i unilaterally spotted the second target virtual character j. Taking the game's key event as a kill event, and the first target virtual character and the second target virtual character as the killer and the killed in the kill event, respectively, as an example, the vision replay information corresponding to the kill event can indicate that the key virtual character as the killer unilaterally spotted the key virtual character as the killed.

[0157] S33: If the first observation state indicates that the first target virtual character has not discovered the second target virtual character, and the second observation state indicates that the second target virtual character has discovered the first target virtual character, then it is determined that the vision recap information indicates that the second target virtual character has unilaterally discovered the first target virtual character.

[0158] For example, in the above In this case, further judgment can be made. Is it true? If the second target virtual character j has spotted the first target virtual character i, but the first target virtual character i has not spotted the second target virtual character j, then the vision replay information indicates that the second target virtual character j unilaterally spotted the first target virtual character i. Taking a key game event as a kill event, with the first and second target virtual characters being the killer and the killed respectively, the vision replay information corresponding to this kill event can indicate that the killed key virtual character unilaterally spotted the killer key virtual character.

[0159] S34: If the first observation state indicates that the first target virtual character has not discovered the second target virtual character, and the second observation state indicates that the second target virtual character has not discovered the first target virtual character, then it is determined that the vision recap information indicates that the first target virtual character and the second target virtual character have not discovered each other.

[0160] For example, in the above In the case of further judgment and confirmation and If both are false, it means that the first target virtual character i did not discover the second target virtual character j, and the second target virtual character j also did not discover the first target virtual character i. In this case, it can be determined that the vision replay information indicates that the first target virtual character i and the second target virtual character j did not discover each other. Taking the game's key event as a kill event, and the first target virtual character and the second target virtual character as the killer and the killed in the kill event, respectively, as an example, the vision replay information corresponding to the kill event can indicate that the key virtual character as the killer and the key virtual character as the killed did not discover each other.

[0161] In this way, by comprehensively considering the observation status of the two key virtual characters towards each other in the key events of the game, the visual recap information corresponding to the key events of the game can be summarized. This visual recap information can then be used to indicate the discovery relationship between the two key virtual characters, thereby providing players with comprehensive and accurate visual recap information related to the key events of the game.

[0162] • Face-to-face determination mechanism

[0163] In one possible implementation, if the vision recap information indicates that the first target virtual character and the second target virtual character have discovered each other, it can be further determined whether they are face-to-face when they discover each other. That is, the method provided in this application embodiment can also include the following S41 (not shown in the figure):

[0164] S41: When the vision recap information indicates that the first target virtual character and the second target virtual character have discovered each other, determine the face-to-face determination result between the first target virtual character and the second target virtual character based on the pitch and yaw angles in the orientation angle data of the first target virtual character and the pitch and yaw angles in the orientation angle data of the second target virtual character.

[0165] The face-to-face determination result is used to indicate whether the first target virtual character and the second target virtual character have discovered each other face-to-face. If the first target virtual character is facing the second target virtual character and the second target virtual character is also facing the first target virtual character, then it can be considered that the first target virtual character and the second target virtual character have discovered each other face-to-face. The face-to-face determination result can be determined based on the pitch angle and yaw angle of the first target virtual character and the second target virtual character respectively.

[0166] For example, the yaw angle difference and pitch angle difference between the first target virtual character i and the second target virtual character j can be calculated using the following equations (21) and (22):

[0167] (twenty one)

[0168] (twenty two)

[0169] in, The difference in yaw angle between the first target virtual character i and the second target virtual character j. and These are the yaw angle and pitch angle from the orientation angle data of the first target virtual character i, respectively. and These are the yaw angle and pitch angle in the orientation angle data of the second target virtual character j, respectively.

[0170] Furthermore, the face-to-face determination result between the first target virtual character i and the second target virtual character j can be determined by the following formula (23):

[0171] (twenty three)

[0172] in, The result of the face-to-face determination between the first target virtual character i and the second target virtual character j; if both conditions are met. ,as well as These two conditions can determine That is, it indicates that the first target virtual character i and the second target virtual character j discover each other face to face; conversely, if at least one of the above conditions is not met, then it can be determined that... This means that the first target virtual character i and the second target virtual character j do not discover each other face to face. The preset face-to-face judgment angle threshold is usually set between 20° and 45°, for example, it can be equal to 45°.

[0173] S42: Add the face-to-face judgment results to the visual replay information.

[0174] The aforementioned face-to-face judgment results can be added to the visual recap information corresponding to key game events, so as to more accurately reflect the discovery situation between the first target virtual character and the second target virtual character.

[0175] In this way, after determining that the first target virtual character and the second target virtual character have discovered each other, the face-to-face judgment result between the two can be determined in more detail and accurately. This face-to-face judgment result is added to the vision replay information, which can make the vision replay information more accurately reflect the vision discovery situation between the first target virtual character and the second target virtual character when key game events occur, thereby helping players to conduct more granular game replays.

[0176] It should be understood that if a key event in the game involves more key virtual characters, such as three or more key virtual characters, the discovery relationships between each key virtual character can also be determined through the above method, thereby summarizing and obtaining comprehensive vision review information.

[0177] Team evaluation mechanism

[0178] In one possible implementation, this application embodiment may also provide team debriefing information corresponding to key game events, so as to help players review the team collaboration situation when key game events occur.

[0179] To determine the team's debriefing information, the event-related data corresponding to the key game events must also include the character status data of each virtual character in the key team during the time period of the key game event. Specifically, a key team refers to at least one team belonging to each of the at least two key virtual characters in the key game event. Taking a kill event as an example, the key team could be the team to which the key virtual character who was killed belonged, or the team to which the key virtual character who killed belonged, or it could include both teams. The character status data of the virtual characters in the key team could, for example, include the virtual character's 3D position data.

[0180] Accordingly, the team replay information corresponding to the key events in the game can be determined through the following steps S51 to S52. That is, the method provided in this application embodiment may also include the following steps S51 to S52 (not shown in the figure):

[0181] S51: Determine the team evaluation result corresponding to the key team based on the role status data of each virtual role in the key team.

[0182] Team evaluation results are used to indicate the performance of key teams on preset evaluation dimensions during the occurrence of key events in the game. These preset evaluation dimensions are dimensions referenced when evaluating the collaboration of key teams, and may include, but are not limited to, team formation, team cohesion, team coverage, team tactical effectiveness, team support performance, and overall team performance.

[0183] When determining the team evaluation result for a key team, the evaluation parameters for reference in determining the team evaluation result can be calculated first based on the role status data of each virtual role in the key team. Then, an evaluation algorithm corresponding to a preset evaluation dimension can be used to determine the team evaluation result based on the aforementioned evaluation parameters. Alternatively, a pre-trained neural network model for determining the team evaluation result can be used to intelligently determine the team evaluation result based on the role status data of each virtual role in the key team. This application does not limit the calculation method of the team evaluation result in any way.

[0184] S52: Based on the team evaluation results, determine the team debriefing information corresponding to key game events.

[0185] Team debriefing information is used to indicate the performance of key teams during critical game events. Specifically, the team evaluation results of the aforementioned key teams can be directly used as the team debriefing information, or the team evaluation results of the aforementioned key teams can be further analyzed and organized to obtain the team debriefing information.

[0186] In this way, by combining the role status data of each virtual character in the key team during the occurrence of key game events, the team's performance in those events can be evaluated, thus providing players with corresponding team debriefing information. This method innovatively proposes team-level debriefing information, which helps players objectively and accurately analyze their team's performance, thereby enabling them to formulate better team tactics in the future.

[0187] In one possible implementation, the team evaluation result can be determined from at least one preset evaluation dimension, including team formation, team cohesion, team coverage, team tactical effectiveness, or overall team performance. Specifically, S51 can include the following S511 (not shown in the figure):

[0188] S511: Based on the role status data of each virtual role in the key team, determine at least one of the following as the team evaluation result: formation result, cohesion assessment result, coverage assessment result, tactical effectiveness assessment result, and comprehensive assessment result.

[0189] The formation result corresponding to the key team indicates the formation structure formed by the arrangement of various virtual roles within the key team. Common formation structures include, but are not limited to, centralized, linear, triangular, diamond, L-shaped, dispersed, and single-person formations. The formation result corresponding to the key team can be any of these common formation structures. For example, when determining the formation result, the arrangement position of each virtual role in the key team can be determined based on their respective role status data. Then, based on the relationship between the arrangement positions of the virtual roles and the formation conditions corresponding to the various common formation structures mentioned above, the formation structure conforming to the arrangement positions of the virtual roles in the key team is determined, and this formation structure is taken as the formation result corresponding to the key team.

[0190] The cohesion assessment result for a key team indicates the degree of clustering among the virtual roles within that team. If the virtual roles are close to each other, the assessment result indicates high cohesion; conversely, if they are far apart, the assessment result indicates low cohesion. For example, when determining the cohesion assessment result, the distance between the virtual roles can be calculated based on their location data, and then the assessment result can be determined based on these distances. Optionally, the movement direction of the virtual roles can also be incorporated as reference information when determining the cohesion assessment result, to avoid classifying virtual roles that are close in location but have significantly different movement directions as having high cohesion.

[0191] The coverage assessment result for the key team indicates the extent to which the formation of each virtual character within the key team covers the virtual game scene. If the formation of the virtual characters within the key team covers a large area of ​​the virtual game scene, or if two virtual characters within the key team are far apart, the coverage assessment result indicates that the key team has a large coverage area. Conversely, if the formation of the virtual characters within the key team covers a small area of ​​the virtual game scene, or if the distances between the virtual characters within the key team are all small, the coverage assessment result indicates that the key team has a small coverage area. For example, when determining the coverage assessment result, the distance between every two virtual characters in the key team can be determined, and the largest distance can be selected. Then, the coverage assessment result is determined based on this largest distance.

[0192] The tactical effectiveness assessment result for the key team indicates the game benefits brought by the formation of the virtual characters within the key team. This tactical effectiveness assessment result is usually strongly correlated with the formation results mentioned above; that is, the tactical effectiveness assessment result can indicate the effectiveness of the formation formed by the key team in the current game scenario. For example, a neural network model for evaluating tactical effectiveness can be pre-trained, and then, using this model, combined with the formation results of the key team and game scenario information at the time of the key game event, the tactical effectiveness assessment result can be determined.

[0193] The overall evaluation result for a key team is determined based on at least one of the following: cohesion evaluation result, coverage evaluation result, tactical effectiveness evaluation result, or key team role survival rate. The key team role survival rate refers to the proportion of surviving virtual characters within the key team out of all virtual characters in that team. For example, since the cohesion evaluation result, coverage evaluation result, and tactical effectiveness evaluation result can typically be expressed as corresponding scores, the overall evaluation result can be determined by weighting and summing these results according to preset weights. The resulting sum indicates the overall performance of the key team in key game events.

[0194] It should be noted that when determining the team evaluation results based on the role status data of each virtual character in the key team during the occurrence of key game events, the role status data of each virtual character can be preprocessed first. Taking the role status data of each virtual character as position data as an example, a corresponding position sequence can be constructed for each virtual character. Then, the position sequence can be processed by de-jittering and interpolation to remove obvious drift points. In addition, processing such as unifying the coordinate system and scaling, and eliminating translation, rotation and scale differences can also be performed.

[0195] In this way, the team evaluation results can be determined from the dimensions of team formation, team cohesion, team coverage, team tactical effectiveness, and team overall performance. This ensures that the team evaluation results can cover the information needed by players when reviewing team collaboration, thus guaranteeing the practicality of the team review information.

[0196] • Determination mechanism of formation result

[0197] In one possible implementation, a pre-defined formation evaluation mechanism or a pre-trained formation recognition model can be used to determine the formation results corresponding to the aforementioned key teams. That is, S511 may include the following S511-11 or S511-12 (not shown in the figure):

[0198] S511-11: Based on the number of surviving virtual characters in the key team and the position data of each surviving virtual character, determine the formation reference information; and based on the formation reference information and the reference information conditions corresponding to various preset formation structures, determine the target reference information conditions satisfied by the formation reference information, and take the preset formation structure corresponding to the target reference information conditions as the formation result.

[0199] The event-related data includes the role status data of each virtual character in the key team, specifically including the survival status and location data (such as 3D coordinate data) of the virtual character. Based on this, the number of surviving virtual characters in the key team can be determined, and the formation reference information can be determined according to the number of surviving virtual characters and the location data of each surviving virtual character. This formation reference information is used as a reference when determining the formation result corresponding to the key team.

[0200] As an example, based on the location data of each surviving virtual character and the actual game terrain, the locations of each surviving virtual character can be connected to construct a nearest neighbor graph. From this graph, graph structure features such as main direction, convex hull, corners, connectivity, and minimum spanning tree can be extracted, and the distances between the surviving virtual characters can be calculated. The aforementioned nearest neighbor graph, the extracted graph structure features, and the distances between the surviving virtual characters can all be used as formation reference information.

[0201] Common formation structures include centralized, linear, triangular, rhomboid, L-shaped, dispersed, and single-person formations. A centralized formation is characterized by small distances between virtual characters, a small coverage area for groups of virtual characters, and strong connectivity in the nearest neighbor graph. Correspondingly, if the average distance between surviving virtual characters in the formation reference information is less than a preset distance threshold, the convex hull area of ​​the nearest neighbor graph is less than a preset area threshold, and the edge weights in the nearest neighbor graph are all less than preset edge weight thresholds, then the formation result for the key team can be determined to be centralized. A linear formation is characterized by virtual characters arranged along a main direction, a nearest neighbor graph with a length greater than its width, and a convex hull shape approximating a line segment. Correspondingly, if the main direction of the nearest neighbor graph in the formation reference information is obvious, and the length along the main direction is much greater than the width in the vertical direction, and the proportion of points within the straight-line model is high, then the formation result for the key team can be determined to be linear. The triangular formation structure is characterized by three or more surviving virtual characters forming a triangular frame with evenly distributed corner points. Correspondingly, if the convex hull in the formation reference information has three vertices forming an approximately equilateral or equiangular triangle, and the alignment error with the triangular template is small, then the formation result of the key team can be determined to be triangular. The rhomboid formation structure is characterized by four approximately equidistant corner points and a symmetrical rhomboid shape. Correspondingly, if the convex hull in the formation reference information has four vertices with paired intersecting diagonals, and the alignment error with the rhomboid template is small, then the formation result of the key team can be determined to be rhomboid. The L-shaped formation structure is characterized by surviving virtual characters arranged along two approximately intersecting line segments. Correspondingly, if the nearest neighbor graph in the formation reference information has two distinct line segment branches, and the angle between the line segment branches is close to a right angle, then the formation result of the key team can be determined to be L-shaped. A distributed formation structure is characterized by large distances between surviving virtual characters, multiple loose subgroups, and wide formation coverage. Correspondingly, if the average distance between surviving virtual characters in the formation reference information is greater than a preset distance threshold, the convex hull area is large, and multiple connected components exist in the nearest neighbor graph, then the formation result of the key team can be determined to be distributed. A single-person formation structure is characterized by at least one virtual character being far removed from the main group, forming a single-point wandering or exploration point. Correspondingly, if the positional distance of surviving virtual characters reveals a small cluster of size 1 that is significantly distant from the main cluster, then the formation result of the key team can be determined to be single-person.

[0202] As another example, the distance between each pair of virtual characters can be calculated based on the location data of each surviving virtual character in the key team. Based on this distance and the number of surviving virtual characters, formation reference information can be determined, thereby determining the formation result of the key team.

[0203] Taking a maximum of 4 virtual characters in a team as an example, we can obtain the 3D coordinate data of each of the four virtual characters in the key team, i.e., P={p1, p2, p3, p4}, where, =[x k ,y k ,z k ] T k equals 1, 2, 3, or 4 as mentioned above. Next, based on the three-dimensional coordinate data of the virtual character, a survival determination is made; if p... k =[0,0,0] T If the k-th virtual character in the key team has been eliminated, then it can be assumed that the k-th virtual character has been eliminated; based on this, the set of surviving characters A={(k, ) The number of surviving virtual characters N = For ease of calculation, the distance between virtual characters can be calculated by considering only their two-dimensional coordinates, i.e., by using the following formula (24):

[0204] (twenty four)

[0205] in, Let i be the distance between the i-th surviving virtual character and the j-th surviving virtual character in the key team. and These are the X-axis and Y-axis coordinates of the i-th virtual character, respectively. and These are the X-axis and Y-axis coordinates of the j-th virtual character, respectively.

[0206] Then all non-zero distances can be expanded into a one-dimensional set D= Therefore, the above parameters N and set D can be determined as array reference information.

[0207] If N=1, the formation result can be directly determined as a single formation, and the confidence score for this formation result can be set to 1.0. If N<2, the formation result can be determined as a scattered formation, and the confidence score for this formation result can be set to 0.3 (preset), as a fallback for the case where N=1. If D= If so, the formation result can be determined to be CONCENTRATED, and the confidence score of the formation result can be set to 0.9 (preset).

[0208] Furthermore, distance statistics can be calculated using the following equations (25) to (28), and these distance statistics can also be added to the formation reference information.

[0209] (25)

[0210] (26)

[0211] (27)

[0212] (28)

[0213] in, The average distance between surviving virtual characters. The standard deviation of the distance between surviving virtual characters. The maximum distance between surviving virtual characters. This represents the minimum distance between surviving virtual characters. Based on the above distance statistics, the formation outcome of key teams can be determined using the following rules:

[0214] like If the value is less than 50, the formation result can be determined as centralized, and the confidence score for this formation result can be set to 0.9 (preset). If N > 300, the formation result can be determined as SCATTERED, and the confidence score for this formation result can be set to 0.4 (preset). If N = 2, the formation result can be determined as LINEAR, and the confidence score for this formation result can be set to 0.7 (preset).

[0215] If N=3, the formation result can be determined as a LINEAR using the line type determination method described below. If the formation result is determined to be a LINEAR, the confidence score for that formation result is set to 0.8 (preset). If the formation result is determined to be a non-LINEAR, the formation result is determined to be a TRIANGULAR, and the confidence score for that formation result is set to 0.85 (preset).

[0216] If N=4, the formation result can first be determined using the diamond-shaped determination method described below. If the formation result is determined to be diamond-shaped, the confidence score for this formation result is set to 0.9 (preset). Otherwise, the formation result can be determined using the line-shaped determination method described below. If the formation result is determined to be line-shaped, the confidence score for this formation result is set to 0.7 (preset). Otherwise, the formation result can be determined using the L-shaped determination method described below. If the formation result is determined to be L-shaped, the confidence score for this formation result is set to 0.8 (preset). Otherwise, the formation result can be directly determined to be scattered, and the confidence score for this formation result is set to 0.6 (preset).

[0217] For cases other than those described above, the formation result can be set to SCATTERED by default, and the confidence score of the formation result can be set to 0.5 by default (preset).

[0218] The following sections introduce the methods for determining line type, diamond type, and L-type:

[0219] Line type determination method: First construct a two-dimensional point set In each pair (i,j), i < j; if >10 -6 The slope is then calculated using the following formula (29). :

[0220] (29)

[0221] Collect all valid slopes (i.e., slopes that are not zero) to form a set S. If S = Then the formation result can be determined to be linear; if S The standard deviation of the slope can then be calculated using the following formula (30). :

[0222] , (30)

[0223] like If the value is less than 0.5, the formation result can be determined as a linear (LINEAR) formation.

[0224] Rhombus determination method: This method is only executed when N=4. First, the two-dimensional center point c of each surviving virtual character is determined by the following formula (31):

[0225] (31)

[0226] Then, determine the set of distances R to the two-dimensional center point c. .

[0227] Next, the standard deviation of the center distance is calculated using the following formula (32). :

[0228] , (32)

[0229] like If the value is less than 30, then the formation result can be determined as diamond (DIAMOND).

[0230] L-shaped determination method: This method is only executed when N≥3. Candidate center point p is selected from the position data of each surviving virtual character, and the relative vector angle between the position q of other virtual characters and the candidate center point p is calculated by the following formula (33):

[0231] (33)

[0232] Then, collect the above set of relative vector angles A={ If there are two angles and It satisfies approximately orthogonality, that is, it satisfies the condition indicated by the following equation (34):

[0233] or (34)

[0234] Then the formation result can be determined to be L-shaped (L_SHAPE); conversely, if all surviving virtual characters are enumerated as candidate center points and none of them satisfy the conditions indicated by equation (34), then the formation result can be determined not to be L-shaped (L_SHAPE).

[0235] S511-12: Using the formation recognition model, determine the formation result based on the position data of each surviving virtual character in the key team.

[0236] The formation recognition model refers to a neural network model that can recognize the formation structure formed by virtual characters in a game. This formation recognition model can be, for example, a large language model (LLM), or other pre-trained models specifically designed for recognizing formation structures, such as convolutional neural networks (CNN) models, feedforward neural networks (FNN) models, etc. This application does not limit the model structure of the formation recognition model in any way.

[0237] For example, when the formation recognition model is an LLM (Limited Linear Model), a prompt can be constructed based on the position data of each virtual role in the key team to instruct the execution of the formation recognition task. After this prompt is input into the LLM, the LLM can automatically perform the formation recognition task based on the position data of each virtual role in the key team, thereby outputting the corresponding formation result. Alternatively, when the formation recognition model is another model specifically designed for formation structures, the position data of each virtual role in the key team can be directly input into the model to obtain the formation recognition result output by the model.

[0238] Thus, through the above methods, a pre-designed formation recognition mechanism, combined with the position data of each surviving virtual character, can quickly and accurately determine the formation of key teams, saving relevant computing resources. Alternatively, a formation recognition model can be used to automatically and intelligently determine the formation of key teams, ensuring the accuracy of the formation recognition results.

[0239] • Determination mechanism of cohesion evaluation result

[0240] In one possible implementation, a pre-defined cohesion assessment mechanism or a pre-trained cohesion assessment model can be used to determine the cohesion assessment results for the aforementioned key teams. That is, S511 may include the following S511-21 or S511-22 (not shown in the figure):

[0241] S511-21: Determine cohesion reference information based on the number of surviving virtual characters in the key team or the location data of each surviving virtual character; and use a preset cohesion algorithm to determine the cohesion score as the cohesion assessment result based on the cohesion reference information.

[0242] The event-related data includes the role status data of each virtual role in the key team, specifically including the survival status and location data (such as three-dimensional coordinate data) of the virtual role. Based on this, the number of surviving virtual roles in the key team can be determined, and cohesion reference information can be determined based on the number of surviving virtual roles and the location data of each surviving virtual role. This cohesion reference information is used as a reference when determining the cohesion assessment result of the key team.

[0243] As an example, based on the location data of each surviving virtual character and the actual game terrain, the locations of each surviving virtual character can be connected to construct a nearest neighbor graph. From this graph, graph structure features such as main direction, convex hull, connectivity, and minimum spanning tree can be extracted, and the distance between each surviving virtual character can be calculated. The aforementioned nearest neighbor graph, the extracted graph structure features, and the distance between each surviving virtual character can all serve as cohesion reference information.

[0244] Furthermore, the distances between surviving virtual characters during the critical game events can be aligned with historical distributions to obtain an initial cohesion score. This historical distribution records the mapping relationship between distances between virtual characters and cohesion scores, derived from statistical analysis of historical game records. Then, by incorporating the average side length of the minimum spanning tree and the consistency of the velocity directions of the surviving virtual characters, the initial cohesion score can be adjusted to obtain the final cohesion score. Additionally, short-term anomalies can be smoothed using a windowed mean or exponential method to ensure curve stability. Thus, by using the above methods, the cohesion of key teams is jointly measured from three dimensions: distance, connectivity, and frequency of convergence, enhancing the ability to characterize the essence of team collaboration.

[0245] As another example, the distance between each pair of virtual characters can be calculated based on the location data of each surviving virtual character in the key team. Based on this distance and the number of surviving virtual characters, cohesion reference information can be determined, thereby determining the cohesion assessment result of the key team.

[0246] Taking a maximum of 4 virtual characters in a team as an example, if the number of surviving virtual characters in the key team is N≤1, then the cohesion score can be determined to be 0 by default. If the number of surviving virtual characters in the key team is N≥2, then the cohesion score can be determined by the following formula (35):

[0247] (35)

[0248] in, It is the average distance between the surviving virtual characters, which can be calculated using equation (25).

[0249] S511-22: Using a cohesion assessment model, determine the cohesion assessment results based on the role status data of each virtual role in the key team.

[0250] A cohesion assessment model is a neural network model that can evaluate the cohesion of a team composed of virtual characters in a game. Such a cohesion assessment model can be, for example, an LLM, or other pre-trained models specifically designed to evaluate team cohesion, such as CNN models, FNN models, etc. This application does not impose any limitations on the model structure of the cohesion assessment model.

[0251] For example, when the cohesion assessment model is an LLM (Limited Line Management) model, prompts can be constructed based on the location data of each virtual role in the key team to instruct the execution of the cohesion assessment task. After inputting these prompts into the LLM model, it can automatically perform the cohesion assessment task based on the location data of each virtual role in the key team, thereby outputting the corresponding cohesion assessment result. Alternatively, when the cohesion assessment model is another model specifically designed for assessing team cohesion, the location data of each virtual role in the key team can be directly input into the model to obtain the cohesion assessment result output by that model.

[0252] Thus, through the above methods, a pre-designed cohesion assessment mechanism, combined with the distances between surviving virtual characters, can quickly and accurately determine the cohesion assessment results for key teams, saving relevant computing resources and improving the efficiency of cohesion assessment. Alternatively, a cohesion assessment model can be used to automatically and intelligently determine the cohesion assessment results for key teams, ensuring the accuracy of the cohesion assessment results.

[0253] • Determination mechanism of coverage evaluation result

[0254] In one possible implementation, a pre-defined coverage assessment mechanism or a pre-trained coverage assessment model can be used to determine the coverage assessment results for the aforementioned key teams. That is, S511 may include the following S511-31 or S511-32 (not shown in the figure):

[0255] S511-31: Determine coverage reference information based on the number of surviving virtual characters in the key team or the location data of each surviving virtual character; and use a preset coverage algorithm to determine the coverage score as the coverage evaluation result based on the coverage reference information.

[0256] The event-related data includes the role status data of each virtual role in the key team, specifically including the survival status and location data (such as 3D coordinate data) of the virtual role. Based on this, the number of surviving virtual roles in the key team can be determined, and based on the number of surviving virtual roles and the location data of each surviving virtual role, coverage reference information can be determined. This coverage reference information is used as a reference when determining the coverage assessment results corresponding to the key team.

[0257] As an example, based on the location data of each surviving virtual character and the actual game terrain, the locations of each surviving virtual character can be connected to construct a nearest neighbor graph. Graph structure features such as convex hulls can be extracted from this graph, and the distances between the surviving virtual characters can be calculated. The aforementioned nearest neighbor graph, the extracted graph structure features, and the distances between the surviving virtual characters can all serve as reference information for coverage.

[0258] Furthermore, the extent of a key team's coverage of the virtual game scene can be measured using the extension of the nearest neighbor graph, taking into account diameter, convex hull area, view union, and keypoint control. Specifically, the distance between the two furthest virtual characters in the key team can be calculated, the union of the key team's formation enclosed area and the visible area (the area reachable by the virtual characters' vision) can be calculated, and proximity dwell time and reachability can be determined for key points and resource points. Then, a specific algorithm is used to calculate the corresponding coverage score by combining the above parameters. Optionally, the ideal coverage range can be differentiated for different stages such as search, map control, and adversarial scenarios. In this way, geometric coverage and semantic coverage (view domain and keypoints) can be integrated to provide interpretable coverage assessment.

[0259] As another example, the distance between the two farthest virtual characters in a critical team can be calculated based on the location data of each surviving virtual character in the critical team. Based on this distance and the number of surviving virtual characters, coverage reference information can be determined, thereby determining the coverage assessment result of the critical team.

[0260] Taking a maximum of 4 virtual characters in a team as an example, if the number of surviving virtual characters in the key team is N≤1, the coverage score can be determined by default to be 0.2. If the number of surviving virtual characters in the key team is N≥2, the coverage score can be determined by the following formula (36):

[0261] (36)

[0262] in, It is the maximum distance between the surviving virtual characters, which can be calculated using equation (27).

[0263] S511-32: Using a coverage assessment model, determine the coverage assessment results based on the role status data of each virtual role in the key team.

[0264] Coverage assessment model refers to a neural network model that can assess the coverage of a team composed of virtual characters in a game. This coverage assessment model can be, for example, an LLM, or other pre-trained models specifically designed to assess team coverage, such as CNN models, FNN models, etc. This application embodiment does not impose any limitations on the model structure of the coverage assessment model.

[0265] For example, when the coverage assessment model is an LLM (Limited Line Management) system, prompts can be built based on the location data of each virtual role within the key team to instruct the execution of a coverage assessment task. After this prompt is input into the LLM, it can automatically perform the coverage assessment task based on the location data of each virtual role within the key team, thereby outputting the corresponding coverage assessment result. Alternatively, when the coverage assessment model is another model specifically designed for assessing team coverage, the location data of each virtual role within the key team can be directly input into the model to obtain its output coverage assessment result.

[0266] Thus, through the above methods, a pre-designed coverage assessment mechanism, combined with the distances between surviving virtual characters, can quickly and accurately determine the coverage assessment results for key teams, saving relevant computing resources and improving coverage assessment efficiency. Alternatively, a coverage assessment model can be used to automatically and intelligently determine the coverage assessment results for key teams, ensuring the accuracy of the coverage assessment results.

[0267] • Determination mechanism of tactical effectiveness evaluation result

[0268] In one possible implementation, a pre-defined tactical effectiveness evaluation mechanism or a pre-trained tactical effectiveness evaluation model can be used to determine the tactical effectiveness evaluation results for the aforementioned key teams. That is, S511 may include the following S511-41 or S511-42 (not shown in the figure):

[0269] S511-41: Determine the tactical effectiveness assessment result based on the tactical effectiveness parameters corresponding to the formation results and the identification confidence level corresponding to the formation results.

[0270] The tactical effectiveness parameter is determined based on the win-loss record of the formations corresponding to key teams in historical game matches. Specifically, the number of wins for each common formation structure in historical game matches can be counted. Then, based on the number of wins and the total number of games played by that formation structure, the win rate corresponding to that formation structure can be calculated as the tactical effectiveness parameter for that formation structure. In this step, the tactical effectiveness parameter for the formation structure indicated by the formation results of the key teams can be applied.

[0271] For example, taking a maximum of 4 virtual characters in a team as an example, if the number of surviving virtual characters N ≤ 1 in the key team, the tactical effectiveness score can be determined by default to 0.3. If the number of surviving virtual characters N ≥ 2 in the key team, the tactical effectiveness score can be determined by combining the tactical effectiveness parameters corresponding to each formation result; specifically, the tactical effectiveness parameters corresponding to various common formation structures... For example, see the following formula (37):

[0272] (37)

[0273] Furthermore, the tactical effectiveness assessment results can be calculated using the following formula (38):

[0274] tactical_effectiveness= ·formation_score (38)

[0275] Among them, tactical_effectiveness is the tactical effectiveness assessment result, and formation_score is the confidence level of the formation result identification.

[0276] S511-42: Using a tactical effectiveness assessment model, determine the tactical effectiveness assessment result based on at least one of the following: formation result, cohesion assessment result, or coverage assessment result.

[0277] A tactical effectiveness evaluation model refers to a neural network model that can evaluate the tactical effectiveness of a team composed of virtual characters in a game. This tactical effectiveness evaluation model can be, for example, an LLM, or other pre-trained models specifically designed to evaluate the tactical effectiveness of a team, such as CNN models, FNN models, etc. The embodiments of this application do not impose any limitations on the model structure of the tactical effectiveness evaluation model.

[0278] As an example, when the tactical effectiveness assessment model is LLM, prompts can be constructed based on the formation results, cohesion assessment results, and coverage assessment results of the key team to instruct the execution of tactical effectiveness assessment tasks. After inputting these prompts into the LLM, the LLM can automatically perform coverage assessment tasks based on the formation results, cohesion assessment results, and coverage assessment results of the key team, thereby outputting the corresponding tactical effectiveness assessment results. Optionally, when constructing the prompts, the number and resources of enemy virtual characters, as well as the game terrain where the key team is located, can also be considered.

[0279] As another example, when the tactical effectiveness assessment model is another model specifically designed to evaluate team tactical effectiveness, this model can be trained first. Specifically, training samples can be recorded within short time windows, including formation structure, team cohesion, team coverage, enemy and friendly numbers and resource differences, game map terrain, and version. Labels are assigned to these training samples based on the rounds or team battle results after the window; in the absence of combat, advance and resource gains can be used as proxy labels. Then, the initial model is used to determine the predicted tactical effectiveness assessment result based on the training samples. The initial model is then trained based on the difference between the labels and the predicted tactical effectiveness assessment result. After model training is complete, the model can be deployed in real-world applications, enabling it to determine the tactical effectiveness assessment result for a key team based on the input key team's formation results, cohesion assessment results, coverage assessment results, the differences in numbers and resources between the key team and the enemy team, and the game terrain. It should be understood that in online applications, group calibration can be performed under different game versions and opponent strengths to ensure cross-environment stability. Thus, by linking formation recognition with contextual features, a tactical effectiveness evaluation framework that is practical, explainable, and transferable can be formed.

[0280] Thus, through the above methods, a pre-designed tactical effectiveness assessment mechanism, combined with the formation results of key teams, can quickly determine the tactical effectiveness assessment results for key teams, saving relevant computing resources and improving the efficiency of tactical effectiveness assessment. Alternatively, a tactical effectiveness assessment model can be used to automatically and intelligently determine the tactical effectiveness assessment results for key teams, combining the formation results, cohesion assessment results, coverage assessment results, the differences in personnel and resources between key teams and enemy teams, and the game terrain, ensuring the accuracy of the tactical effectiveness assessment results.

[0281] Furthermore, with a maximum of 4 virtual roles in a team, if the number of surviving virtual roles in the key team is N≤1, the overall evaluation score (corresponding to the overall evaluation result) can be defaulted to 0.3. If the number of surviving virtual roles in the key team is N≥2, it can be determined by the following formula (39) based on the cohesion score, coverage score, tactical effectiveness score, and survival rate of the key team:

[0282] (39)

[0283] Where survival_rate=N / 4, N is the number of virtual characters surviving in the key team.

[0284] Therefore, by efficiently and quickly considering the formation results, cohesion assessment results, coverage assessment results, tactical effectiveness assessment results, and survival rate of the key team, a comprehensive assessment result for the key team can be determined.

[0285] • Provide tactical advice and information

[0286] In one possible implementation, this application embodiment can also provide corresponding tactical recommendations based on the team evaluation results corresponding to the key teams. That is, the method provided by this application embodiment may further include the following S61 (not shown in the figures):

[0287] S61: Based on the team assessment results, determine the target tactical recommendations for key teams.

[0288] The target tactical suggestion information is a tactical suggestion information given to the key team based on the team evaluation results. For example, it can indicate the formation structure that the key team should adopt in key events of the game, or it can indicate the reasonable positional arrangement between virtual characters in the key team. This application embodiment does not limit the target tactical suggestion information in any way.

[0289] As an example, a pre-defined tactical suggestion mechanism or a pre-trained tactical suggestion generation model can be used to determine the target tactical suggestion information corresponding to the aforementioned key teams. That is, S61 above can include the following S61-11 or S61-12 (not shown in the figure):

[0290] S61-11: Based on the team's evaluation results and the game environment information during the period of the occurrence of key game events, determine the reference battle situation information; and find the target battle situation information that matches the reference battle situation information in the suggested mapping relationship, and determine the tactical suggestion information corresponding to the target battle situation information in the suggested mapping relationship as the target tactical suggestion information.

[0291] Game environment information refers to the descriptive information of the area where the key team is located on the virtual game map during the occurrence of a key game event. Specifically, the area where the key team is located can be determined based on the location of each virtual character in the virtual game map. Then, the terrain information, building information, and relevant information of other virtual characters within the area (such as location, equipped virtual equipment, etc.) can be obtained as the game environment information. This application embodiment does not specifically limit the game environment information.

[0292] Reference battle situation information is used to indicate the battle situation situation of a key team during the occurrence of a key game event. As an example, the team evaluation results corresponding to the key team and the aforementioned game environment information can be directly used as the reference battle situation information. As another example, the team evaluation results corresponding to the key team and the aforementioned game environment information can be summarized and organized to obtain reference battle situation information describing the battle situation situation of the key team. For example, based on the team evaluation results corresponding to the key team and the aforementioned game environment information, prompt information can be constructed to instruct the execution of the battle situation situation summarization task. This prompt information is then input into the LLM (Local Management Module), allowing the LLM to analyze the current battle situation situation of the key team based on the prompt information and output the reference battle situation information. This application does not limit the method of determining the reference battle situation information in any way.

[0293] The suggested mapping relationship is used to record the tactical suggestions corresponding to various preset battle situation information. These preset battle situation information can be derived from summarizing battle situation forms that have appeared in historical games, specifically describing frequently occurring battle situation forms in the game. The tactical suggestions corresponding to each preset battle situation information are tactical adjustment suggestions given for the battle situation form indicated by that preset battle situation information. Specifically, these can be suggestions for adjusting the positioning of virtual characters, or suggestions for the movement paths of virtual characters; for example, when the preset battle situation information indicates that the virtual characters in the team are scattered and about to engage the enemy team, the tactical suggestions can suggest that the virtual characters in the team be grouped into a triangle or rhombus shape, and provide suggested spacing and mutual protection relationships. Another example is when the preset battle situation information indicates that the team's formation... When the formation is linear and the team is in an open area, tactical advice can instruct the team to change to a zigzag formation or spread out their flanks to create crossfire and avoid damage from penetration. For example, when the preset battle situation information indicates that the virtual characters within the team are too concentrated and the enemy team possesses area-of-effect capabilities, the tactical advice can instruct the virtual characters within the team to appropriately disperse to a safe mutual protection distance. Furthermore, when the preset battle situation information indicates that the team is in the search and map control phase, the tactical advice can instruct the team to expand their coverage area while maintaining communication links and support radius, and so on. This application does not limit the preset battle situation information or its corresponding tactical advice in any way.

[0294] After determining the reference battle situation information based on the team evaluation results and game environment information corresponding to the key teams, the preset battle situation information that matches the reference battle situation information can be found from the above-mentioned suggestion mapping relationship and used as the target battle situation information. For example, a semantic analysis model can be used to determine the reference semantic features corresponding to the reference battle situation information and the preset semantic features corresponding to each preset battle situation information in the suggestion mapping relationship. Then, the feature similarity between the reference semantic features and each preset semantic feature is calculated, and the preset semantic feature with the highest feature similarity with the reference semantic features is determined. The preset battle situation information corresponding to the preset semantic feature is used as the target battle situation information. Correspondingly, the tactical suggestion information corresponding to the target battle situation information recorded in the suggestion mapping relationship can be used as the target tactical suggestion information. This target tactical suggestion information is the suggestion information given for the performance of the key teams during the occurrence of key events in the game.

[0295] S61-12: Generates target tactical suggestion information based on team evaluation results and game environment information during the period of key game events, using a tactical suggestion generation model.

[0296] A tactical suggestion generation model refers to a neural network model that can provide tactical suggestions to a team based on the current state of the game. This tactical suggestion generation model can be, for example, an LLM, or other pre-trained models specifically designed for generating tactical suggestions, such as CNN models, FNN models, etc. This application embodiment does not impose any limitations on the model structure of the tactical suggestion generation model.

[0297] For example, when the tactical suggestion generation model is an LLM (Limited Management Model), prompts can be constructed based on the team evaluation results of the key teams and the game environment information during the occurrence of key events. These prompts instruct the LLM to automatically execute the tactical suggestion generation task based on the team evaluation results of the key teams and the game environment information during the occurrence of key events, thereby outputting the corresponding target tactical suggestion information. Alternatively, when the tactical suggestion generation model is another model specifically designed for generating tactical suggestions, the team evaluation results of the key teams and the game environment information during the occurrence of key events can be directly input into the model to obtain the target tactical suggestion information output by the model.

[0298] Therefore, through the above methods, when players review their games, they can receive further tactical suggestions related to the performance of key teams. These suggestions can help players better learn teamwork mechanisms in similar key game events, thus aiding their rapid improvement. Furthermore, when determining these tactical suggestions, a pre-designed tactical suggestion generation mechanism can be used, combining the team evaluation results of the key teams with game environment information, to quickly and accurately identify targeted tactical suggestions, improving the efficiency of tactical suggestion generation. Alternatively, a tactical suggestion generation model can be used to automatically and intelligently generate tactical suggestions, ensuring the accuracy of the provided tactical advice.

[0299] See Figure 6 , Figure 6 This is a flowchart illustrating a front-end game replay method provided in an embodiment of this application. The method is executed by a computer device, such as a terminal device or a server. Figure 6 As shown, the game replay method includes the following steps:

[0300] S601: Display a virtual game map and mark the target locations of at least two key virtual characters involved in key game events on the virtual game map.

[0301] A virtual game map refers to the game map corresponding to a player's game match. It is used to visually display the geographical distribution of the virtual game scene where the player-controlled virtual character is located in a map format. This virtual game map can be any of the following: a two-dimensional map, a 2.5-dimensional map, or a three-dimensional map. It can include elements such as virtual roads, virtual buildings, virtual forests, virtual deserts, and virtual oceans in the virtual game scene. This application is private and does not impose any limitations on the virtual game map.

[0302] A key game event refers to a predefined type of event that occurs between at least two key virtual characters, who are the virtual characters participating in the key game event. For example, a key game event can be determined based on the game type. In SLG, FPS / TPS, or MOBA games, a kill event can be considered a key game event, and the virtual character who kills and the virtual character who is killed in that kill event are the aforementioned key virtual characters. Of course, in practical applications, other types of game events, such as attack events or healing events, can also be considered key game events. This application does not limit the type of key game event in any way.

[0303] In this embodiment, the target locations of at least two key virtual characters involved in key game events can be marked on the virtual game map. The target location of a key virtual character refers to its position at the time the key game event occurs. Specifically, marking the target location of a key virtual character on the virtual game map can be achieved by displaying a preset type of marker (such as a circular marker, a square marker, or a positioning marker) at the target location. It should be noted that in practical applications, the target locations of all key game events occurring in a game match can be marked on the virtual game map, or only the target locations of some key game events (such as key game events related to a specific player) can be marked. This embodiment does not impose any limitations on the key game events and key virtual characters represented on the virtual game map.

[0304] For example, after a game ends, the players who participated in the game can trigger the viewing of the replay information of the game in the game application. In response to the player's triggering of the viewing of the replay information, the game application can display a virtual game map and mark the target locations of key virtual characters involved in the key game events that occurred in the game on the virtual game map.

[0305] Taking kill events as a key game event as an example, see [link / reference]. Figure 7 andFigure 8 , Figure 7 and Figure 8 These are schematic diagrams illustrating two virtual game maps provided in embodiments of this application. Figure 7 In the virtual game map shown, circular markers with numbers and specific colors are used to indicate the target locations of key virtual characters involved in each kill event that occurs in the game. The numbers of the circular markers are used to indicate the order in which the kill events of the key virtual characters occur in the game, and the colors of the circular markers are used to indicate the role of the key virtual character in the kill event. For example, black indicates that the key virtual character is the killer, and white indicates that the key virtual character is the killer. Figure 8 The virtual game map shown is Figure 7 The virtual game map shown is a partial area; players can zoom in and move around on the virtual game map to examine in detail the kill events that occurred in specific areas within the map. For example, Figure 8 The virtual game map includes specific areas where the 7th, 8th, and 9th kills in a match are concentrated. Optionally, the virtual game map may also include the ranges of the various safe zones that appear during the match.

[0306] S602: In response to the first operation triggered by the target position of the target virtual character in the virtual game map, display the event recap information corresponding to the key game event.

[0307] The target virtual character is any key virtual character represented in the virtual game map. Taking a kill event as an example, the target virtual character can be any killer or killed player in any kill event that occurs in the game.

[0308] The first operation refers to the operation used to trigger the viewing of event recap information corresponding to the key events in the game to which the target virtual character belongs. The first operation can be any of the following: a click operation, a double click operation, or a long press operation. The first operation can also be, for example, triggering the display of a recap menu related to the key events in the game to which the target virtual character belongs, and selecting to view the event recap information in the recap menu. This application embodiment does not limit the first operation in any way.

[0309] The event recap information corresponding to key game events helps players review the occurrence and development of those events. This recap information may include at least field-of-view recap information, which indicates the target virtual character's observation status towards other key virtual characters during the event, such as whether the target virtual character detected other key virtual characters. Optionally, the field-of-view recap information may also indicate the observation status of other key virtual characters towards the target virtual character during the event, such as whether other key virtual characters detected the target virtual character. If the field-of-view recap information indicates that the target virtual character and other key virtual characters mutually detected each other, it may further indicate whether the detection occurred face-to-face. The specific mechanism for determining the field-of-view recap information can be found in the above description. Figure 4 The process indicated by S401 to S404 in the illustrated embodiment.

[0310] Optionally, the event review information corresponding to the aforementioned key game events may also include team review information corresponding to the key game events. This team review information is determined based on the team evaluation results of the key teams. The key teams are at least one team from the teams to which the at least two key virtual characters involved in the key game events belong. The team evaluation results are used to indicate the performance of the key teams on a preset evaluation dimension during the period in which the key game events occurred.

[0311] That is, the event recap information corresponding to the key game events can also include team recap information. Specifically, this team recap information can include at least one of the team evaluation results of the key team to which the target virtual character belongs, or the team evaluation results of the key teams to which other key virtual characters belong. The team evaluation results here can include, for example, at least one of the following: the key team's formation results, cohesion evaluation results, coverage evaluation results, tactical effectiveness evaluation results, and comprehensive evaluation results during the period in which the key game event occurred. The formation results can be any of the common formation structures, such as concentrated, linear, triangular, diamond, L-shaped, dispersed, and single-player formations. The cohesion evaluation results can be a cohesion score indicating the degree of clustering of the virtual characters in the key team. The coverage evaluation results can be a coverage score indicating the extent to which the formation of the virtual characters in the key team covers the virtual game scene. The tactical effectiveness evaluation results can be a tactical effectiveness score indicating the game benefits brought by the formation of the virtual characters in the key team. The comprehensive evaluation result can be a weighted sum of the above cohesion score, coverage score, tactical effectiveness score, and the key team's character survival rate. The specific mechanism for determining the team's evaluation results can be found in the process indicated by S51 to S52 in the above embodiments.

[0312] Optionally, the event recap information corresponding to the aforementioned key game events may also include target tactical suggestion information for the key teams. This target tactical suggestion information is used to indicate the team tactical suggestions given to the key teams in the key game events, and it is determined based on the team evaluation results corresponding to the aforementioned key teams.

[0313] That is, the event recap information corresponding to key game events can also include target tactical suggestions for key teams. These suggestions can, for example, indicate the formation structure the key team should adopt during the key game event, or the appropriate positional arrangement of virtual characters within the key team. The mechanism for determining these target tactical suggestions can be found in the S61 instruction process described in the above embodiment.

[0314] For example, players can target Figure 7 or Figure 8 Clicking on any key virtual character's target location marked on the virtual game map triggers a first action, targeting that virtual character's location. In response to this first action, the game application can display... Figure 9 The diagram shown illustrates the replay information interface, which includes event replay information corresponding to the kill event of the target virtual character.

[0315] Specifically, such as Figure 9 As shown, the event recap information can include basic event information 901 corresponding to the kill event (such as the poison circle number, the time of occurrence, and the distance between the location and the poison circle boundary), and basic information 902 of the key virtual characters who were the killer and the kill (such as nickname, location at the time of the kill, changes in health during the kill, virtual equipment, virtual scope, virtual helmet, and virtual armor used in the kill, posture at the time of the kill, and whether the kill occurred inside a virtual building). It can also include the details of the kill event itself. The information 903 on the visual recap between the key virtual character who is the killer and the key virtual character who is the killer (such as whether the other party was spotted, whether it was a face-to-face spot, the difference in yaw angle and pitch angle) may also include the team evaluation results 904 of the key teams to which the key virtual character who is the killer and the key virtual character who is the killer belong in the kill event (such as the overall score, cohesion score, coverage score, tactical effectiveness score), and may also include the tactical advice information 905 given to the key team to which the target virtual character (taking the key virtual character who is the killer in the kill event as an example) belongs.

[0316] In the game replay method provided in this application embodiment, a virtual game map marked with the target locations of key virtual characters involved in key game events can be displayed to the player first, so that the player can understand the location of the key virtual characters when the key game events occur from a macro perspective. Then, in response to the first operation triggered by the player based on the target location of the target virtual character (any key virtual character) in the virtual game map, the event replay information corresponding to the key game event to which the target virtual character belongs is displayed accordingly. The event replay information includes the field of view replay information corresponding to the key game event, which can indicate the observation status of the target virtual character towards other key virtual characters in the key game event. Thus, the player is provided with field of view replay information that can objectively reflect the field of view observation of the target virtual character, providing accurate and reliable replay information from the dimension of field of view observation, thereby helping the player to better conduct fine-grained and targeted replays of key game events.

[0317] Furthermore, the aforementioned event recap information can also include team recap information corresponding to key game events. This team recap information can reflect the performance of key teams in specific evaluation dimensions during key game events. Therefore, by providing players with team-level recap information, players can objectively and accurately analyze the performance of key teams in key game events, thus helping them to formulate better team tactics in the future.

[0318] In addition, the event recap information may also include target tactical suggestions for key teams. These suggestions indicate the appropriate tactics that the key team should take when a key game event occurs. By providing these suggestions, players can better learn the team collaboration mechanisms in similar key game events, thereby helping them improve quickly.

[0319] Mark locations on the virtual game map.

[0320] In one possible implementation, when marking the target location of a key virtual character on the virtual game map, the system can further indicate the key game event to which the key virtual character belongs, and its role in that key game event. Furthermore, the system can also mark the target locations of teammates within the same team on the virtual game map. That is, the "marking the target locations of at least two key virtual characters involved in a key game event on the virtual game map" in S601 above can include the following S71 (not shown in the figure):

[0321] S71: Mark the target locations of key virtual characters involved in multiple key game events on the virtual game map, and at the target location of each key virtual character, mark the identity of the key virtual character in the key game event, as well as the occurrence sequence number of the key game event in the whole game.

[0322] In a game match, multiple key events usually occur. Accordingly, when reviewing the game match, the target locations of the key virtual characters involved in each key event can be marked on the virtual game map. This allows players to have a macroscopic understanding of the locations of each key event when reviewing the game.

[0323] In addition, when marking the target location of a key virtual character on the virtual game map, the identity of the key virtual character in key game events and the sequence number of the key game events to which the key virtual character belongs in the entire game can also be marked. For example, the identity of the key virtual character in key game events can be distinguished by configuring different colors for the target location markers, and the sequence of occurrence of the key game events to which the key virtual character belongs in the entire game can be indicated by configuring numbers for the target location markers.

[0324] For example, taking a kill event as a key game event, see [example]. Figure 7 and Figure 8 The virtual game map shown uses circular markers to indicate the target locations of key virtual characters in each kill event. Each circular marker can be configured with a corresponding color and number. The color of the circular marker indicates the role of the key virtual character in the kill event; for example, black indicates that the key virtual character is the killer, and white indicates that the key virtual character is the killer. The number of the circular marker indicates the order in which the kill event to which the key virtual character belongs occurs in the game.

[0325] Optionally, the method provided in this application embodiment may further include the following S72 (not shown in the figures):

[0326] S72: Mark the target location of the associated virtual character for each key game event on the virtual game map, and at the target location of each associated virtual character, mark the team to which the associated virtual character belongs, as well as the sequence number of the key game event in the entire game.

[0327] Associated virtual characters refer to teammates of key virtual characters, that is, virtual characters who establish a teammate relationship with the key virtual character before the start of the game, and are controlled by other players.

[0328] In the virtual game map, in addition to marking the target locations of key virtual characters in critical game events, it's also possible to mark the target locations of their teammates during those events. This allows players to understand the support available to teammates during critical game events. When marking the target locations of related virtual characters, different markers can be used than those corresponding to the key virtual character's target location. For example, if a circular marker is used for the key virtual character's target location, other shapes (such as squares) can be used to mark the target locations of related virtual characters, or even smaller circular markers can be used.

[0329] When marking the target location of an associated virtual character, you can also mark the team to which the associated virtual character belongs, as well as the sequence number of the key game events associated with the associated virtual character in the entire game. For example, for key virtual characters and associated virtual characters belonging to the same team, you can indicate that they belong to the same team by configuring the same color for the identifiers of their respective target locations, and you can indicate the sequence of the key game events to which they belong in the entire game by configuring the same number for the identifiers of their respective target locations.

[0330] For example, taking a kill event as a key game event, see [example]. Figure 8 The virtual game map shown uses large circular markers to indicate the target locations of key virtual characters in each kill event, and smaller circular markers to indicate the target locations of related virtual characters. Each smaller circular marker can be configured with a corresponding color and number. The color of the smaller circular marker indicates the team to which the related virtual character belongs in the kill event; for example, black indicates that the related virtual character and the key virtual character who is the killer belong to the same team, and white indicates that the related virtual character and the key virtual character who is the killer belong to the same team. The number of the smaller circular marker indicates the order in which the kill events associated with the related virtual character occur in the game.

[0331] In this way, the target locations of key virtual characters and their teammates in each key game event can be marked on the virtual game map simultaneously. Furthermore, the identities of the key virtual characters and their teammates in the key game event, as well as the order in which the key game event occurs, can be marked. This allows players to more intuitively understand the order and location of key game events throughout the game through the virtual game map. It also makes it easier for players to know the relevant support situation of teammates in the key game events. In other words, the virtual game map concisely presents a wealth of replay information related to key game events.

[0332] • Overlay a death heat layer onto the virtual game map

[0333] In one possible implementation, a virtual game map can also be used to represent the distribution of death events in a game match. That is, the method provided in this application embodiment may also include the following S81 (not shown in the figure):

[0334] S81: Overlay a death heat layer onto the virtual game map.

[0335] The death heatmap layer is used to indicate the frequency of virtual deaths in different areas of a virtual game map. Specifically, the death heatmap layer uses different color depths to represent the occurrence of virtual deaths in each area of ​​the virtual game map. For example, it can obtain the location coordinates of each virtual character killed in a game match, and then, based on these coordinates, calculate the frequency of virtual deaths in each area of ​​the virtual game map, thus generating the death heatmap layer. Areas with a high number of virtual deaths can be represented by darker color blocks, while areas with fewer virtual deaths can be represented by lighter color blocks.

[0336] For example, Figure 10 This is a schematic diagram of a death thermal layer provided in an embodiment of this application. For example... Figure 10 As shown, a death heat map is overlaid on the virtual game map. This death heat map uses different shades of gray to indicate the heat of virtual death events in each area. A higher gray depth in an area indicates that more virtual death events have occurred in that area, while a lower gray depth indicates that fewer virtual death events have occurred in that area.

[0337] Optionally, high-risk areas for immediate death upon landing can be further marked in the aforementioned death heat map. These high-risk areas are determined based on the coordinates of the death locations of virtual characters whose survival time in a game is less than a preset duration (e.g., 180 seconds). These high-risk areas can indicate areas where players face a high risk of death upon landing.

[0338] In this way, by combining the virtual game map and the death heat map, players can be provided with a new dimension of replay information. This information can help players intuitively and accurately identify areas in the virtual game map where virtual deaths are likely to occur, thus enabling them to formulate reasonable game strategies for subsequent gameplay.

[0339] • Replay key events in the game

[0340] In one possible implementation, embodiments of this application can support players in viewing recorded videos corresponding to key game events, and the video will show the field of view of key virtual characters. That is, the method provided by embodiments of this application may further include the following S91 to S92 (not shown in the figures):

[0341] S91: In response to a second operation triggered based on the target location of the target virtual character in the virtual game map, or in response to a video viewing operation triggered based on event recap information, play the event recording video of the key game event to which the target virtual character belongs.

[0342] The second operation refers to the operation of triggering the viewing of the event recording video corresponding to the key game event to which the target virtual character belongs. The second operation can be any one of the following: a click operation, a double click operation, or a long press operation. The second operation can also be, for example, triggering the display of a replay menu related to the key game event to which the target virtual character belongs, and selecting to view the event recording video in the replay menu. This application embodiment does not limit the second operation in any way.

[0343] The video viewing operation refers to the operation of requesting to view the recorded video of a key game event, triggered by a replay information interface that carries event recap information. For example, the replay information interface carries event recap information and deploys a video viewing control. Players can trigger the video viewing operation by clicking, double-clicking, or long-pressing the video viewing control. Alternatively, players can trigger the video viewing operation by triggering preset gesture operations on the replay information interface, such as swiping up, swiping down, swiping left, or swiping right. This embodiment of the application does not limit the video viewing operation in any way.

[0344] Game key event recording video refers to video used to reflect the situation in the virtual game scene during the period when the game key event occurs. Specifically, it can be a video obtained by recording the virtual game scene within 10 seconds before and after the moment when the game key event occurs.

[0345] For example, players can long-press on a marker indicating the target location of a virtual character on a virtual game map to trigger the second operation mentioned above; alternatively, a video viewing control can be deployed in the replay information interface that carries event recap information corresponding to key game events, allowing players to trigger a video viewing operation by clicking on the control. In response to the second operation or the video viewing operation, the game application can play a recorded video of the key game event involving the target virtual character, including content from various locations within the virtual game scene within 10 seconds before and after the key game event.

[0346] S92: In the event recording video, mark the field of vision range of at least two key virtual characters in the key game events.

[0347] In the recorded video of the key events in the game, the field of vision of at least two key virtual characters in the key events is also marked. The field of vision of the key virtual characters can be determined by the above S402, and can be specifically represented as the three-dimensional cone of the field of vision of the key virtual characters.

[0348] For example, see Figure 11 , Figure 11 This is a schematic diagram illustrating an event recording video provided in an embodiment of this application. For example... Figure 11 As shown, the field of vision of the key virtual character 1101 is represented by the three-dimensional field of vision cone 1102, and the field of vision of the key virtual character 1103 is represented by the three-dimensional field of vision cone 1104. Since the key virtual character 1103 is included in the three-dimensional field of vision cone 1102 and the key virtual character 1101 is included in the three-dimensional field of vision cone 1104, it can be determined that the key virtual character 1101 and the key virtual character 1103 discover each other in the key events of the game.

[0349] In this way, the above methods can provide players with recorded videos of key game events, allowing them to quickly review those events. Furthermore, the recorded videos can be further annotated with the field of vision of at least two key virtual characters during the key events. This allows players to intuitively understand how each key virtual character observes the other key virtual characters, enabling an objective and accurate review of the key virtual characters' field of vision during those events.

[0350] • View teammates' support routes

[0351] In one possible implementation, embodiments of this application can support players in viewing the movement trajectories of teammate characters during key game events. That is, the method provided in embodiments of this application may further include the following S101 (not shown in the figures):

[0352] S101: In response to a third operation triggered based on the target location of the target virtual character in the virtual game map, or in response to a support trajectory viewing operation triggered based on event replay information, mark the movement trajectory of the target virtual character's teammate virtual character during the period of occurrence of the key game event in the virtual game map.

[0353] The third operation refers to the operation used to trigger the viewing of the movement trajectory of the virtual teammates of the target virtual character in the key game event to which the target virtual character belongs. The third operation can be any of the following: a click operation, a double click operation, or a long press operation. The third operation can also be, for example, triggering the display of a replay menu related to the key game event to which the target virtual character belongs, and selecting to view the teammate support trajectory in the replay menu. This application embodiment does not limit the third operation in any way.

[0354] The "support trajectory viewing" operation refers to the action triggered by a replay information interface that carries event replay information, requesting to view the support trajectory of a teammate's virtual character during key game events. For example, the replay information interface carries event replay information and deploys a support trajectory viewing control. Players can trigger the support trajectory viewing operation by clicking, double-clicking, or long-pressing the control. Alternatively, players can trigger the support trajectory viewing operation by triggering preset gesture operations within the replay information interface, such as swiping up, swiping down, swiping left, or swiping right. This embodiment of the application does not impose any limitations on the support trajectory viewing operation.

[0355] A teammate virtual character refers to a virtual character that has a teammate relationship with the target virtual character, and is usually controlled by other players. The movement trajectory of a teammate virtual character during the occurrence of a key game event refers to the movement route of the teammate virtual character during the occurrence of the key game event, which may be, for example, the movement route within 10 seconds before and after the occurrence of the key game event.

[0356] For example, a player can double-click on a marker indicating the target virtual character's location on a virtual game map to trigger the aforementioned third operation; alternatively, a trajectory viewing control can be deployed in the replay information interface that carries event replay information corresponding to key game events, allowing the player to trigger a trajectory viewing operation by clicking the control. In response to the aforementioned third operation or the trajectory viewing operation, the game application can display a virtual game map and mark the movement trajectories of each of the target virtual character's teammates during the time period of the key game event; for example, see... Figure 12 , Figure 12 A schematic diagram of the support trajectory provided in the embodiments of this application, such as Figure 12 As shown, the movement trajectories 1201, 1202 and 1203 of the target virtual character's three teammate virtual characters are marked on the virtual game map. It can be seen from these three movement trajectories that when a key event occurs in the game, the target virtual character's three teammate virtual characters all move towards the target virtual character's location 1204 in order to provide support to the target virtual character.

[0357] In this way, players can be provided with the movement trajectories of their teammates' virtual characters when key game events occur. This allows players to intuitively review the support provided by their teammates' virtual characters during those key events, providing them with objective and accurate review information on teammate support.

[0358] Based on the game replay method provided in the preceding embodiments, this application also provides a corresponding game replay device. The following, in conjunction with... Figure 13 and Figure 14 To explain, Figure 13 This application provides a schematic diagram of the structure of a game replay device 1300, which includes:

[0359] The data acquisition module 1301 is used to acquire event-related data corresponding to key game events. The key game events are preset type events that occur between at least two key virtual characters. The event-related data includes at least the character status data of the at least two key virtual characters during the occurrence period of the key game events.

[0360] The field of view determination module 1302 is used to determine the field of view range of the target virtual character based on the character status data of the target virtual character in the event association data, wherein the target virtual character is any one of the key virtual characters;

[0361] The observation state determination module 1303 is used to determine the observation state of the target virtual character towards other key virtual characters in the key events of the game, based on the field of vision of the target virtual character and the character state data of other key virtual characters.

[0362] The replay information determination module 1304 is used to determine the visual replay information corresponding to the key game events based on the observation state.

[0363] Optionally, the event association data further includes the character status data of each virtual character in the key team during the occurrence period of the key game event, wherein the key team is at least one of the teams to which the at least two key virtual characters belong; the device further includes:

[0364] The team evaluation module is used to determine the team evaluation result corresponding to the key team based on the role status data of each virtual character in the key team. The team evaluation result is used to indicate the performance of the key team in a preset evaluation dimension during the occurrence period of the key game event.

[0365] The debriefing information determination module 1304 is also used to determine the team debriefing information corresponding to the key game events based on the team evaluation results.

[0366] Optionally, the character status data of the key virtual character includes the position data and orientation angle data of the key virtual character; the field of view determination module 1302 is specifically used for:

[0367] Based on the orientation angle data of the target virtual character, the orientation vector of the target virtual character is determined; based on the position data of the target virtual character, the orientation vector, and the field of view distance threshold, the field of view range of the target virtual character is determined.

[0368] Optionally, the observation state determination module 1303 is specifically used for:

[0369] Based on the location data of the target virtual character and the location data of the key virtual character, the relative position information between the target virtual character and the key virtual character is determined, and the relative position information includes relative direction and relative distance;

[0370] Based on the relative direction and the orientation vector of the target virtual character, determine the relative projection data of the key virtual character to the target virtual character;

[0371] Based on the relative projection data and the observation angle threshold, as well as the relative distance and the field of view distance threshold, the observation state of the target virtual character towards the key virtual character is determined.

[0372] Optionally, the at least two key virtual characters include a first target virtual character and a second target virtual character; the debriefing information determination module 1304 is specifically used for:

[0373] If the first target virtual character's first observation state of the second target virtual character indicates that the first target virtual character has discovered the second target virtual character, and the second target virtual character's second observation state of the first target virtual character indicates that the second target virtual character has discovered the first target virtual character, then it is determined that the vision recap information indicates that the first target virtual character and the second target virtual character have discovered each other.

[0374] If the first observation state indicates that the first target virtual character has discovered the second target virtual character, and the second observation state indicates that the second target virtual character has not discovered the first target virtual character, then it is determined that the field of view recap information indicates that the first target virtual character has unilaterally discovered the second target virtual character;

[0375] If the first observation state indicates that the first target virtual character has not discovered the second target virtual character, and the second observation state indicates that the second target virtual character has discovered the first target virtual character, then it is determined that the field of view recap information indicates that the second target virtual character has unilaterally discovered the first target virtual character;

[0376] If the first observation state indicates that the first target virtual character has not discovered the second target virtual character, and the second observation state indicates that the second target virtual character has not discovered the first target virtual character, then it is determined that the vision recap information indicates that the first target virtual character and the second target virtual character have not discovered each other.

[0377] Optionally, the device further includes:

[0378] The face-to-face determination module is used to determine the face-to-face determination result between the first target virtual character and the second target virtual character based on the pitch angle and yaw angle in the orientation angle data of the first target virtual character and the pitch angle and yaw angle in the orientation angle data of the second target virtual character when the vision review information indicates that the first target virtual character and the second target virtual character have discovered each other.

[0379] The review information determination module 1304 is also used to add the face-to-face judgment result to the visual review information.

[0380] Optionally, the team evaluation module is specifically used for:

[0381] Based on the role status data of each virtual role in the key team, determine at least one of the following: formation result, cohesion assessment result, coverage assessment result, tactical effectiveness assessment result, and comprehensive assessment result, as the team assessment result;

[0382] The formation result indicates the formation structure formed by the arrangement of the virtual characters in the key team; the cohesion assessment result indicates the degree of aggregation of the virtual characters in the key team; the coverage assessment result indicates the coverage of the virtual game scene by the formation of the virtual characters in the key team; the tactical effectiveness assessment result indicates the game benefits brought by the formation of the virtual characters in the key team; and the comprehensive assessment result is determined based on at least one of the cohesion assessment result, the coverage assessment result, the tactical effectiveness assessment result, or the character survival rate of the key team.

[0383] Optionally, the virtual character's status data includes the virtual character's survival status and location data, and the team evaluation module includes a formation result determination submodule, which is used for:

[0384] Based on the number of surviving virtual characters in the key team and the location data of each surviving virtual character, formation reference information is determined; and based on the formation reference information and the reference information conditions corresponding to various preset formation structures, the target reference information conditions satisfied by the formation reference information are determined, and the preset formation structure corresponding to the target reference information conditions is taken as the formation result.

[0385] or,

[0386] The formation result is determined by using a formation recognition model based on the position data of each surviving virtual character in the key team.

[0387] Optionally, the virtual character's status data includes the virtual character's survival status and location data, and the team evaluation module includes a cohesion determination submodule, which is used for:

[0388] Based on the number of surviving virtual characters in the key team, or the location data of each surviving virtual character, cohesion reference information is determined; and a preset cohesion algorithm is used to determine a cohesion score as the cohesion evaluation result based on the cohesion reference information.

[0389] or,

[0390] The cohesion assessment result is determined based on the role status data of each virtual role in the key team using the cohesion assessment model.

[0391] Optionally, the virtual character's status data includes the virtual character's survival status and location data, and the team evaluation module includes a coverage determination submodule, which is used for:

[0392] Based on the number of surviving virtual characters in the key team, or the location data of each surviving virtual character, coverage reference information is determined; and a preset coverage algorithm is used to determine a coverage score as the coverage evaluation result based on the coverage reference information.

[0393] or,

[0394] The coverage assessment result is determined based on the role status data of each virtual role in the key team using the coverage assessment model.

[0395] Optionally, the team evaluation module includes a tactical effectiveness determination submodule, which is used for:

[0396] The tactical effectiveness evaluation result is determined based on the tactical effectiveness parameters corresponding to the formation result and the identification confidence level corresponding to the formation result. The tactical effectiveness parameters are determined based on the win-loss record of the formation result in historical game matches.

[0397] or,

[0398] The tactical effectiveness assessment result is determined using a tactical effectiveness assessment model based on at least one of the formation result, the cohesion assessment result, or the coverage assessment result.

[0399] Optionally, the device further includes:

[0400] The tactical recommendation determination module is used to determine the target tactical recommendation information corresponding to the key team based on the team evaluation results.

[0401] Optionally, the tactical suggestion determination module is specifically used for:

[0402] Based on the team's evaluation results and the game environment information during the time period of the key game events, reference battle information is determined; and target battle information that matches the reference battle information is searched in the suggested mapping relationship, and the tactical suggestion information corresponding to the target battle information in the suggested mapping relationship is determined as the target tactical suggestion information. The suggested mapping relationship records the tactical suggestion information corresponding to various preset battle information.

[0403] or,

[0404] The tactical suggestion generation model generates target tactical suggestion information based on the team's evaluation results and the game environment information during the period in which the key game events occur.

[0405] Figure 14 This application provides a schematic diagram of the structure of a game replay device 1400, which includes:

[0406] The map display module 1401 is used to display a virtual game map and mark the target positions of at least two key virtual characters involved in key game events on the virtual game map. The target positions are the positions of the key virtual characters when the key game events occur.

[0407] The replay information display module 1402 is used to respond to a first operation triggered based on the target position of the target virtual character in the virtual game map, and to display the event replay information corresponding to the key game event. The target virtual character is any one of the key virtual characters. The event replay information includes at least vision replay information, which is used to indicate the observation status of the target virtual character towards other key virtual characters in the key game event.

[0408] Optionally, the event recap information also includes team recap information corresponding to the key game events. The team recap information is determined based on the team evaluation results of the key teams. The key teams are at least one team among the teams to which the at least two key virtual characters belong. The team evaluation results are used to indicate the performance of the key teams on a preset evaluation dimension during the period in which the key game events occurred.

[0409] Optionally, the event debriefing information may also include target tactical suggestion information corresponding to the key team. The target tactical suggestion information is used to indicate the team tactical suggestions given to the key team in the key game event, and it is determined based on the team evaluation results.

[0410] Optionally, the map display module 1401 is specifically used for:

[0411] Mark the target locations of the key virtual characters involved in each of the multiple key game events on the virtual game map, and at the target location of each key virtual character, mark the identity of the key virtual character in the key game event and the occurrence sequence number of the key game event in the whole game;

[0412] The map display module 1401 is also used to mark the target location of the associated virtual character for each key game event on the virtual game map, wherein the associated virtual character is a teammate of the key virtual character; and at the target location of each associated virtual character, mark the team to which the associated virtual character belongs, and the occurrence sequence number of the key game event in the whole game.

[0413] Optionally, the map display module 1401 is further configured to:

[0414] A death heat map is overlaid on the virtual game map to indicate the popularity of virtual death events in various areas of the virtual game map.

[0415] Optionally, the device further includes:

[0416] The video playback module is used to play the event recording video of the key game event to which the target virtual character belongs in response to a second operation triggered based on the target location of the target virtual character in the virtual game map, or in response to a video viewing operation triggered based on the event recap information;

[0417] The field of view annotation module is used to annotate the field of view range of each of the at least two key virtual characters in the key game events in the recorded video of the event.

[0418] Optionally, the device further includes:

[0419] The trajectory annotation module is used to, in response to a third operation triggered based on the target location of the target virtual character in the virtual game map, or in response to a support trajectory viewing operation triggered based on the event replay information, to annotate the movement trajectory of the target virtual character's teammate virtual characters in the virtual game map during the occurrence period of the key game event.

[0420] This application also provides a computer device for executing the above-described game replay method. Specifically, the computer device may be a terminal device or a server. The terminal device and server provided in this application will be described below from the perspective of hardware implementation.

[0421] See Figure 15 , Figure 15 This is a schematic diagram of the structure of the terminal device provided in the embodiments of this application. For example... Figure 15 As shown, for ease of explanation, only the parts related to the embodiments of this application are shown. For specific technical details not disclosed, please refer to the method section of the embodiments of this application. The terminal can be any terminal device including mobile phones, tablets, personal digital assistants (PDAs), point-of-sale (POS) terminals, in-vehicle computers, etc. Taking a computer as an example:

[0422] Figure 15 This is a block diagram illustrating a portion of the structure of a computer associated with the terminal provided in an embodiment of this application. (Reference) Figure 15The computer includes: a radio frequency (RF) circuit 1510, a memory 1520, an input unit 1530 (including a touch panel 1531 and other input devices 1532), a display unit 1540 (including a display panel 1541), a sensor 1550, an audio circuit 1560 (connected to a speaker 1561 and a microphone 1562), a wireless fidelity (WiFi) module 1570, a processor 1580, and a power supply 1590, etc. Those skilled in the art will understand that... Figure 15 The computer architecture shown does not constitute a limitation on the computer and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0423] The memory 1520 can be used to store software programs and modules. The processor 1580 executes various computer functions and data processing by running the software programs and modules stored in the memory 1520. The memory 1520 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the computer (such as audio data, telephone directory, etc.). In addition, the memory 1520 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0424] Processor 1580 is the control center of the computer, connecting various parts of the computer through various interfaces and lines. It performs various computer functions and processes data by running or executing software programs and / or modules stored in memory 1520, and by calling data stored in memory 1520. Optionally, processor 1580 may include one or more processing units; preferably, processor 1580 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may also not be integrated into processor 1580.

[0425] In this embodiment, the processor 1580 included in the terminal is used to execute the steps in the game replay method described in the foregoing embodiments.

[0426] See Figure 16 , Figure 16This is a schematic diagram of the structure of a server 1600 provided in an embodiment of this application. The server 1600 can vary significantly due to different configurations or performance, and may include one or more central processing units (CPUs) 1622 (e.g., one or more processors) and memory 1632, and one or more storage media 1630 (e.g., one or more mass storage devices) for storing application programs 1642 or data 1644. The memory 1632 and storage media 1630 can be temporary or persistent storage. The program stored in the storage media 1630 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the server. Furthermore, the CPU 1622 may be configured to communicate with the storage media 1630 and execute the series of instruction operations in the storage media 1630 on the server 1600.

[0427] Server 1600 may also include one or more power supplies 1626, one or more wired or wireless network interfaces 1650, one or more input / output interfaces 1658, and / or one or more operating systems, such as Windows Server. TM Mac OS X TM Unix TM Linux TM FreeBSD TM etc.

[0428] The steps performed by the server in the above embodiments can be based on this Figure 16 The server structure shown is as follows. The CPU 1622 is used to execute the steps in the game replay methods described in the foregoing embodiments.

[0429] This application also provides a computer-readable storage medium for storing a computer program that performs the steps in the game replay method described in the foregoing embodiments.

[0430] This application also provides a computer program product or computer program that includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps in the game replay methods described in the foregoing embodiments.

[0431] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0432] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0433] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0434] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0435] If the integrated unit is implemented as 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 technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing computer programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0436] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0437] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A game review method, characterized by, The method comprises: obtaining event association data corresponding to a game key event, the game key event being a preset type event occurring between at least two key virtual characters, the event association data at least including character state data of the at least two key virtual characters respectively within an occurrence time period of the game key event; determining a field of view range of a target virtual character according to the character state data of the target virtual character in the event association data, the target virtual character being any one of the key virtual characters; determining an observation state of the target virtual character to other key virtual characters in the game key event according to the field of view range of the target virtual character and the character state data of the other key virtual characters; determining field of view review information corresponding to the game key event based on the observation state.

2. The method of claim 1, wherein, The event association data further includes character state data of each virtual character in a key team respectively within the occurrence time period of the game key event, the key team being at least one team among teams to which the at least two key virtual characters belong respectively; the method further comprises: determining a team evaluation result corresponding to the key team according to the character state data of each virtual character in the key team, the team evaluation result being used to indicate performance of the key team in a preset evaluation dimension within the occurrence time period of the game key event; determining team review information corresponding to the game key event based on the team evaluation result.

3. The method according to claim 1 or 2, characterized in that, The character state data of the key virtual character includes position data and orientation angle data of the key virtual character; The determining of the field of view range of the target virtual character according to the character state data of the target virtual character in the event association data comprises: determining an orientation vector of the target virtual character according to the orientation angle data of the target virtual character; and determining the field of view range of the target virtual character based on the position data of the target virtual character, the orientation vector, and a field of view distance threshold.

4. The method of claim 3, wherein, The determining of the observation state of the target virtual character to other key virtual characters in the game key event according to the field of view range of the target virtual character and the character state data of the other key virtual characters comprises: determining relative position information between the target virtual character and the key virtual characters according to the position data of the target virtual character and the position data of the key virtual characters, the relative position information including a relative direction and a relative distance; determining relative projection data of the key virtual characters to the target virtual character according to the relative direction and the orientation vector of the target virtual character; determining the observation state of the target virtual character to the key virtual characters based on the relative projection data and an observation angle threshold, and the relative distance and the field of view distance threshold.

5. The method according to any one of claims 1 to 4, characterized in that, The at least two key virtual characters include a first target virtual character and a second target virtual character; and the determining of the field of view review information corresponding to the game key event based on the observation state comprises: If the first observation state of the first target virtual character indicates that the first target virtual character discovers the second target virtual character, and the second observation state of the second target virtual character indicates that the second target virtual character discovers the first target virtual character, it is determined that the field-of-view review information indicates that the first target virtual character and the second target virtual character discover each other. If the first observation state indicates that the first target virtual character discovers the second target virtual character, and the second observation state indicates that the second target virtual character does not discover the first target virtual character, it is determined that the field-of-view review information indicates that the first target virtual character discovers the second target virtual character unilaterally. If the first observation state indicates that the first target virtual character does not discover the second target virtual character, and the second observation state indicates that the second target virtual character discovers the first target virtual character, it is determined that the field-of-view review information indicates that the second target virtual character discovers the first target virtual character unilaterally. If the first observation state indicates that the first target virtual character does not discover the second target virtual character, and the second observation state indicates that the second target virtual character does not discover the first target virtual character, it is determined that the field-of-view review information indicates that the first target virtual character and the second target virtual character do not discover each other.

6. The method of claim 5, wherein, The method further comprises: In a case where the field-of-view review information indicates that the first target virtual character and the second target virtual character discover each other, determining a face-to-face determination result between the first target virtual character and the second target virtual character according to a pitch angle and a yaw angle in the orientation angle data of the first target virtual character and a pitch angle and a yaw angle in the orientation angle data of the second target virtual character. Adding the face-to-face determination result to the field-of-view review information.

7. The method according to any one of claims 2 to 6, characterized in that, The determining of the team evaluation result corresponding to the key team according to the role state data of each virtual character in the key team comprises: Determining at least one of a formation result, a cohesion evaluation result, a coverage evaluation result, a tactical effectiveness evaluation result, and a comprehensive evaluation result corresponding to the key team as the team evaluation result according to the role state data of each virtual character in the key team; wherein the formation result is used to indicate a formation structure formed by the arrangement of each virtual character in the key team, the cohesion evaluation result is used to indicate the degree of gathering of each virtual character in the key team, the coverage evaluation result is used to indicate the coverage degree of the formation composed of each virtual character in the key team to the virtual game scene, the tactical effectiveness evaluation result is used to indicate the game income brought by the formation composed of each virtual character in the key team, and the comprehensive evaluation result is determined according to at least one of the cohesion evaluation result, the coverage evaluation result, the tactical effectiveness evaluation result, or the role survival rate of the key team.

8. The method of claim 7, wherein, The role state data of the virtual role includes a survival state and position data of the virtual role, and the formation result is determined by the following manner: A formation reference information is determined according to the number of the virtual roles surviving in the key team and the position data of each of the virtual roles surviving; and a target reference information condition is determined according to the formation reference information and reference information conditions corresponding to each of a plurality of preset formation structures, the preset formation structure corresponding to the target reference information condition is taken as the formation result. Alternatively, The formation result is determined by a formation identification model according to the position data of each of the virtual roles surviving in the key team. The role state data of the virtual role includes a survival state and position data of the virtual role, and the cohesion evaluation result is determined by the following manner:

9. The method according to claim 7 or 8, characterized in that, A cohesion reference information is determined according to the number of the virtual roles surviving in the key team or the position data of each of the virtual roles surviving; and a cohesion score is determined as the cohesion evaluation result according to the cohesion reference information by using a preset cohesion algorithm. Alternatively, The cohesion evaluation result is determined by a cohesion evaluation model according to the role state data of each of the virtual roles in the key team. The role state data of the virtual role includes a survival state and position data of the virtual role, and the coverage evaluation result is determined by the following manner:

10. The method according to any one of claims 7 to 9, characterized in that, A coverage reference information is determined according to the number of the virtual roles surviving in the key team or the position data of each of the virtual roles surviving; and a coverage score is determined as the coverage evaluation result according to the coverage reference information by using a preset coverage algorithm. Alternatively, The coverage evaluation result is determined by a coverage evaluation model according to the role state data of each of the virtual roles in the key team. The tactical effectiveness evaluation result is determined by the following manner:

11. The method according to any one of claims 7 to 10, characterized in that, The tactical effectiveness evaluation result is determined according to a tactical effectiveness parameter corresponding to the formation result and an identification confidence corresponding to the formation result, the tactical effectiveness parameter being determined according to a win or lose situation of the formation result in a historical game session. Alternatively, The tactical effectiveness evaluation result is determined by a tactical effectiveness evaluation model according to at least one of the formation result, the cohesion evaluation result or the coverage evaluation result. The method further includes:

12. The method according to any one of claims 2 to 11, characterized in that, A target tactical suggestion information corresponding to the key team is determined according to the team evaluation result. The target tactical suggestion information corresponding to the key team is determined according to the team evaluation result, including:

13. The method of claim 12, wherein, ​ According to the team evaluation result and game environment information in the occurrence period of the game key event, reference battle information is determined; and target battle information matched with the reference battle information is searched in a suggestion mapping relationship, and tactical suggestion information corresponding to the target battle information in the suggestion mapping relationship is determined as the target tactical suggestion information, and the suggestion mapping relationship records tactical suggestion information corresponding to each of a plurality of preset battle information; Or, According to the team evaluation result and game environment information in the occurrence period of the game key event, the target tactical suggestion information is generated through a tactical suggestion generation model.

14. A game replay method, characterized by, The method comprises: displaying a virtual game map and marking target positions of at least two key virtual characters involved in a game key event in the virtual game map, the target positions being positions of the key virtual characters in the game key event; in response to a first operation triggered based on a target position of a target virtual character in the virtual game map, displaying event review information corresponding to the game key event, the target virtual character being any of the key virtual characters, and the event review information at least including field of view review information for indicating an observation state of the target virtual character on other key virtual characters in the game key event.

15. The method of claim 14, wherein, The event review information further includes team review information corresponding to the game key event, the team review information being determined according to a team evaluation result of a key team, the key team being at least one of teams to which the at least two key virtual characters respectively belong, and the team evaluation result being used to indicate a performance of the key team in a preset evaluation dimension in an occurrence period of the game key event.

16. The method of claim 15, wherein, The event review information further includes target tactical suggestion information corresponding to the key team, the target tactical suggestion information being used to indicate a team tactical suggestion given to the key team in the game key event, and the target tactical suggestion information being determined according to the team evaluation result.

17. The method according to any one of claims 14 to 16, characterized in that, The marking of the target positions of the at least two key virtual characters involved in the game key event in the virtual game map comprises: marking target positions of the key virtual characters involved in a plurality of game key events in the virtual game map, and marking identities of the key virtual characters in the game key events and occurrence serial numbers of the game key events in the entire game at the target positions of the key virtual characters. The method further comprises: marking target positions of associated virtual characters of each game key event in the virtual game map, the associated virtual characters being teammate characters of the key virtual characters, and marking teams to which the associated virtual characters belong and occurrence serial numbers of the game key events in the entire game at the target positions of the associated virtual characters.

18. The method according to any one of claims 14 to 17, characterized in that, The method further comprises: superimposing a death heat map layer on the virtual game map, the death heat map layer being used to indicate a heat of virtual death events in each region in the virtual game map.

19. The method according to any one of claims 14 to 18, characterized in that, The method further comprises: in response to a second operation triggered based on a target position of the target virtual character in the virtual game map, or in response to a video viewing operation triggered based on the event review information, playing an event recording video of the game key event to which the target virtual character belongs; in the event recording video, marking a field of view range of each of the at least two key virtual characters in the game key event.

20. The method according to any one of claims 14 to 19, characterized in that, The method further comprises: in response to a third operation triggered based on the target position of the target virtual character in the virtual game map, or in response to a support trajectory viewing operation triggered based on the event review information, marking a moving trajectory of a teammate virtual character of the target virtual character in the virtual game map within a time period of occurrence of the game key event.

21. A game review apparatus, characterized by The device comprises: a data acquisition module configured to acquire event-related data corresponding to a game key event, the game key event being a preset type event occurring between at least two key virtual characters, and the event-related data at least including character state data of each of the at least two key virtual characters within a time period of occurrence of the game key event; a field of view determination module configured to determine a field of view range of a target virtual character according to character state data of the target virtual character in the event-related data, the target virtual character being any one of the key virtual characters; an observation state determination module configured to determine an observation state of the target virtual character to other key virtual characters in the game key event according to the field of view range of the target virtual character and character state data of the other key virtual characters; a review information determination module configured to determine view review information corresponding to the game key event based on the observation state.

22. A game replay apparatus, characterized by The device comprises: a map display module configured to display a virtual game map and mark target positions of at least two key virtual characters involved in a game key event in the virtual game map, the target positions being positions of the key virtual characters at a time of occurrence of the game key event; a review information display module configured to display event review information corresponding to the game key event in response to a first operation triggered based on a target position of a target virtual character in the virtual game map, the target virtual character being any one of the key virtual characters, and the event review information at least including view review information for indicating an observation state of the target virtual character to other key virtual characters in the game key event.

23. A computer device, comprising: The device comprises a processor and a memory; The memory is configured to store a computer program; The processor is configured to execute the game review method according to any one of claims 1 to 20 according to the computer program.

24. A computer-readable storage medium, characterized in that, The computer readable storage medium is configured to store a computer program, and the computer program is executed by an electronic device to implement the game review method according to any one of claims 1 to 20.

25. A computer program product comprising a computer program, characterised in that, The computer program is executed by a processor to implement the game review method according to any one of claims 1 to 20.

Images