A method and system for assessing the situation of a football match

Abstract

This invention relates to the field of match situation analysis technology, and discloses a method and system for assessing the situation of a football match. The invention acquires tracking data and event data from a football match, performs feature extraction processing to obtain first feature data and second feature data, performs time alignment to obtain target fusion data, analyzes the defensive style, offensive phase, and threat level of the teams in the match, obtains information on defensive style, offensive phase, and threat level, determines football match analysis prompts, and inputs them into a large language model trained on big data to obtain the match situation assessment result. This invention achieves multi-dimensional analysis of the match situation by fusing tracking data and event data, avoiding the limitation of existing technologies that can only analyze single-dimensional data. It analyzes the team's defensive style, offensive phase, and threat level holistically, and finally generates prompts which are input into the large language model to obtain an accurate football match situation analysis based on the overall match.

Description

Technical Field

[0001] This invention relates to the field of match situation analysis technology, specifically to a method and system for assessing the situation of a football match. Background Technology

[0002] In football matches, situational assessment is a crucial step in understanding the game's progress, conducting post-match analysis, and evaluating player and team performance. Situational assessment methods mainly fall into two categories: The first is player modeling based on theoretical dynamics models, which uses mathematical models to describe player behavior. This approach lacks the ability to comprehensively analyze the overall game situation. The second category assesses the specific actions of individual players, such as passing quality, shooting success rate, and interception effectiveness. However, this method only provides analysis of individual performance and fails to consider the team's overall tactical layout, resulting in an incomplete understanding and analysis of the game situation.

[0003] Therefore, current technologies for assessing the situation in football matches neglect the importance of overall tactical planning and teamwork, and cannot analyze the situation from a holistic perspective. Summary of the Invention

[0004] In view of this, the present invention provides a method and system for assessing the situation of a football match, in order to address how to improve the comprehensiveness of football match situation analysis from a global perspective.

[0005] In a first aspect, the present invention provides a method for assessing the situation of a football match, the method comprising:

[0006] Acquire tracking and event data at any point in a football match;

[0007] The tracking data and event data are subjected to feature extraction processing to obtain first feature data after processing the tracking data and second feature data after processing the event data, respectively.

[0008] The first feature data and the second feature data are time-aligned to obtain the time-aligned target fusion data.

[0009] The target fusion data is analyzed to determine the defensive style, offensive phase, and threat level of the teams in the game, resulting in defensive style information, offensive phase information, and threat level information. The target fusion data represents the data set obtained by aligning the first feature data and the second feature data to the same timeline.

[0010] Football match analysis prompts are determined based on the target fusion data, defensive style information, offensive phase information, and threat level information.

[0011] The football match analysis prompts are input into a large language model trained on big data from real football matches to obtain the football match situation assessment results.

[0012] This invention acquires tracking data and event data at any moment during a football match. Feature extraction processing is performed on the tracking data and event data to obtain first feature data after tracking data processing and second feature data after event data processing, ensuring sufficient valuable feature data for subsequent situation analysis. The first and second feature data are time-aligned to obtain time-aligned target fusion data, ensuring consistency of data from different sources on the timeline and allowing for a global perspective in subsequent situation analysis. The target fusion data is analyzed to assess the defensive style, offensive phase, and threat level of the teams, yielding defensive style information, offensive phase information, and threat level information. Football match analysis prompts are determined based on these information, combining the above three analyses with all the fusion data. This ensures the generated prompts have better descriptive capabilities, making the subsequent analysis results of the large language model more closely reflect the actual match process. The football match analysis prompts are input into a large language model trained on big data from real football matches to obtain a football match situation assessment result. By utilizing the trained large language model, a comprehensive football match situation analysis result is generated.

[0013] In one optional implementation, the first feature data includes at least the ball's position information, team information, the identification information of the defending team, the player's identity information, the player's position information, the player's speed information, the coordinates of the court, the match number information, the remaining time of the match, the identification information of the player currently in possession of the ball, and the information of the team currently in possession of the ball.

[0014] The second feature data includes at least the team information where the event occurred, the identity information of the players in the team where the event occurred, the event type information, the event result information, the starting position information of the event, and the ending position information of the event.

[0015] In this embodiment of the invention, the first feature data includes static information about various players, teams, stadiums, and balls, while the second feature data includes dynamic information about events. This provides a sufficient amount of valuable feature data for subsequent situational analysis, thereby improving the accuracy of the analysis and helping the large language model to better understand the game situation.

[0016] In one optional implementation, the target fusion data is analyzed to determine the defensive style of the participating teams, yielding defensive style information, specifically including:

[0017] Based on player position information, team information, stadium coordinates, ball position information, current ball possession player identification information, current ball possession team information, event type information, start position information of the event, and end position information of the event, cluster analysis is performed on the attacking and defending teams to obtain the position information of each attacking line and each defensive line on the field.

[0018] Based on the positional information of each offensive line and each defensive line, determine the permutation and combination information between the offensive and defensive lines;

[0019] The defensive style information of the teams in the match is determined based on the permutation and combination information.

[0020] This invention utilizes cluster analysis based on multi-dimensional data such as player position information, team information, and stadium coordinates to obtain the position information of each offensive and defensive line. This allows for a comprehensive understanding of the team's tactical arrangements during the game. Based on the arrangement and combination information between the offensive and defensive lines, the tactical changes of the team can be analyzed and judged in real time, making the analysis more relevant to the actual game situation and generating specific defensive style information. This provides a data foundation related to defensive style for determining subsequent game prompts.

[0021] In one optional implementation, the target fusion data is analyzed during the offensive phase of a match to obtain offensive phase information, specifically including:

[0022] The first relative position information between the player and the ball is determined based on the player's position information, the coordinates of the field, and the position information of the ball;

[0023] The type of offensive line controlling the ball is determined based on the position information of each offensive line, the starting position information of the event, the ending position information of the event, and the first relative position information.

[0024] Determine offensive phase information based on the type of offensive line that controls the ball.

[0025] This invention, through determining the first relative position information between the player and the ball based on the player's position information, the field coordinate information, and the ball's position information, can obtain a more accurate analysis of the offensive situation. Based on the position information of each offensive line, the start and end position information of the event, and the first relative position information, the type of offensive line controlling the ball is determined, making the classification of offensive methods clearer and more evidence-based. The offensive phase information is determined based on the type of offensive line controlling the ball, providing a data foundation related to offensive phase information for the generation of subsequent match prompts.

[0026] In one optional implementation, the target fusion data is analyzed to determine the threat level of the participating teams, yielding threat level information, specifically including:

[0027] Threat analysis of the court area is based on the court's coordinate information;

[0028] The second relative position information between the ball and the court is determined based on the event type information, the ball's position information, and the court area.

[0029] The threat level of the teams in the match is determined based on the second relative position information.

[0030] This invention, through its embodiments, divides the threat analysis field area based on the field's coordinate information. By combining event type information, ball position information, and field area information, it determines the second relative position information between the ball and the field, thereby determining the threat level information in a dynamic football match and providing a data foundation related to threat level information for the subsequent generation of match prompts.

[0031] In one optional implementation, determining football match analysis prompts based on the target fusion data, defensive style information, offensive phase information, and threat level information specifically includes:

[0032] The data sets of first and second feature data aligned to the same timeline, along with the corresponding defensive style information, offensive phase information, and threat level information, are input into a preset prompt word template to obtain football match analysis prompt words; wherein, the preset prompt word template represents a question template obtained by semantic feature analysis based on football match analysis data.

[0033] This invention integrates first and second feature data aligned to the same timeline, and combines them with corresponding defensive style information, offensive phase information, and threat level information to generate prompts with better descriptive capabilities. The subsequent analysis results of the large language model are more consistent with the actual game process, ensuring that the analysis results are not only comprehensive but also have higher accuracy.

[0034] Secondly, the present invention provides a football match situation assessment system, the system comprising:

[0035] The data acquisition module is used to acquire tracking data and event data at any moment during a football match;

[0036] The feature extraction module is used to perform feature extraction processing on the tracking data and event data to obtain first feature data after processing the tracking data and second feature data after processing the event data, respectively.

[0037] The time alignment module is used to align the first feature data and the second feature data in time to obtain the time-aligned target fusion data.

[0038] The analysis module is used to analyze the target fusion data to determine the defensive style, offensive phase, and threat level of the teams in the game, and to obtain defensive style information, offensive phase information, and threat level information; wherein, the target fusion data represents the data set obtained by aligning the first feature data and the second feature data to the same timeline;

[0039] The prompt word generation module is used to determine football match analysis prompt words based on the target fusion data, defensive style information, offensive phase information, and threat level information.

[0040] The evaluation results module is used to input the football match analysis prompts into a large language model trained on big data from real football matches to obtain the football match situation evaluation results.

[0041] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the football match situation assessment method described in the first aspect or any corresponding embodiment thereof.

[0042] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the football match situation assessment method described in the first aspect or any corresponding embodiment thereof.

[0043] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the football match situation assessment method described in the first aspect or any corresponding embodiment thereof. Attached Figure Description

[0044] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0045] Figure 1 This is a flowchart illustrating a football match situation assessment method according to an embodiment of the present invention;

[0046] Figure 2 This is a flowchart illustrating another method for assessing the situation of a football match according to an embodiment of the present invention;

[0047] Figure 3 This is a schematic diagram of the defensive style arrangement and combination of the football match situation assessment method according to an embodiment of the present invention;

[0048] Figure 4 This is a schematic diagram of the field area division in the football match situation assessment method according to an embodiment of the present invention;

[0049] Figure 5 This is a flowchart illustrating another method for assessing the situation of a football match according to an embodiment of the present invention.

[0050] Figure 6 This is a structural block diagram of a football match situation assessment system according to an embodiment of the present invention;

[0051] Figure 7 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] Football is a typical team-based competitive sport, where situational assessment plays a crucial role in the course of the game, post-match analysis, and evaluation of players and teams. Furthermore, in team competitions, simply relying on scoring or winning the game is insufficient for evaluating player performance and conducting tactical analysis.

[0054] Traditional methods for assessing the situation in football matches fall into two categories. One involves using theoretical dynamic models to model players. While this method has some theoretical basis, it only provides a basic, general model that cannot be applied to different scenarios or offer a comprehensive situational analysis. The other method focuses on evaluating the actions of players, especially those with the ball, such as assessing the quality of a pass, the success of a shot, or the success of an interception. This type of method can only evaluate the specific actions of individual players, neglecting the importance of overall tactical planning and teamwork in football, and failing to analyze the game situation from a holistic perspective.

[0055] This invention provides a method for assessing the situation of a football match, applicable to scenarios such as real-time football match analysis, post-match analysis, player evaluation, team evaluation, and tactical research. By integrating tracking data and event data, it achieves multi-dimensional analysis of the match situation. Through time-aligned processing, it enables dynamic assessment at different stages of the match, avoiding the limitation of existing technologies that can only analyze single-dimensional data. It provides a comprehensive analysis of the team's defensive style, offensive phases, and threat level, and the generated prompts, combined with a large language model, yield a more accurate football match situation analysis based on the overall game.

[0056] According to an embodiment of the present invention, a method for assessing the situation of a football match is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0057] This embodiment provides a method for assessing the situation of a football match, which can be used with the aforementioned computer. Figure 1 This is a flowchart of a football match situation assessment method according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps:

[0058] Step S101: Obtain tracking data and event data at any moment during a football match.

[0059] In this embodiment, the tracking data may include all tracking data of players and the ball within the playing field area. The playing field area refers to the entire area on a football field where players can run freely. Players can be virtual characters in a ball game simulation program, real players in a ball game, intelligent robots during a ball game, or intelligent agents in the field of artificial intelligence; the type of player is not specifically limited here. Tracking data is collected through wearable devices (such as sensors on sports vests or shoes worn by players) or extracted from game videos using computer vision technology. The GPS or other positioning technologies built into the wearable devices ensure that the precise position, speed, acceleration, and other parameters of each player are captured in real time.

[0060] It is understandable that when players are in a virtual environment, tracking data of players and the ball can be tracked and extracted through a program within the virtual environment. When players are in a real ball game environment, players and the ball can be pre-configured or equipped with positioning and motion capture devices, which capture tracking data of each player and the ball in real time. The tracking data at any moment in the game is the raw data extracted from any moment in any game. The data tracking time interval can be preset, for example, data can be extracted once every 1 second or 1 second. The data extraction time interval can also be defined by presetting the number of tracking data. For example, in a 90-minute game, the number of tracking data can be set to 5000, in which case data can be extracted once every 1 second. This embodiment of the invention does not impose specific limitations on this.

[0061] In this embodiment, event data refers to descriptions of key events in the match, including passes, shots, and fouls. Event data can be acquired through manual annotation (e.g., real-time recording by commentators or data analysts) or through automated match analysis software. Automated match analysis software can automatically identify and label events in the match based on player actions and ball movement patterns in the match video.

[0062] For example, in the 15th minute of the match, player A passes the ball from midfield to player B. Tracking data at this moment records the real-time positions and speeds of players A and B. Event data, based on automated match analysis software, annotates the start and end times of this passing action, the identities of players A and B, and the starting and ending positions of the pass. Only by obtaining sufficiently accurate and comprehensive data can we provide data support for subsequent feature extraction, time alignment, and situational analysis, forming the foundation for completing the entire football match situational analysis.

[0063] Step S102: Perform feature extraction processing on the tracking data and event data to obtain the first feature data after processing the tracking data and the second feature data after processing the event data, respectively.

[0064] Specifically, for tracking data, interpolation is needed to complete the data, and for event data, different marking methods need to be uniformly processed to obtain a consistent description.

[0065] For example, the tracking data records player A's specific location and speed at 15 minutes and 12 seconds into the game. The processed first feature data will retain player A's identity, location, and speed information at that moment. The event data records player A's passing action at that moment. The processed second feature data will retain the start and end times of the pass, player A's identity, and the result of the pass (whether it was successfully passed to the target player).

[0066] This embodiment considers that the raw features of the tracking data contain many data features, which are difficult for a large language model to understand due to their complex temporal sequence. Some features, such as the position and speed of the ball and players, are highly relevant to game situation assessment and prediction, while others, such as player roles, are relatively less relevant. Furthermore, considering that some features highly relevant to game situation assessment and prediction cannot be directly obtained from the raw features of the tracking data, it is necessary to perform feature engineering on the raw features of the tracking data. Feature engineering can specifically include advanced feature extraction and feature filtering. For example, features like player roles, which have relatively low relevance to game situation assessment and prediction, can be filtered out, while features with relatively high relevance can be retained.

[0067] Step S103: Time-align the first feature data and the second feature data to obtain time-aligned target fusion data.

[0068] Specifically, a time alignment algorithm is needed to align tracking and event data from the same match at different granularities and with different annotation methods, thus obtaining tracking and event data at the same moment. Time alignment ensures that the analysis model can combine tracking and event data at the same time; only when the timelines are consistent can the situational analysis be comprehensive and accurate.

[0069] For example, player A makes a pass at 15 minutes and 12 seconds into the game, and the event data marks the time of the pass; while the tracking data records player position information 10 times per second. The result of time alignment will ensure that at 15 minutes and 12 seconds, there is both the pass event data of player A and the tracking data of his position and speed.

[0070] Step S104: Analyze the defensive style, offensive phase, and threat level of the teams in the match using the target fusion data to obtain defensive style information, offensive phase information, and threat level information; wherein, the target fusion data represents the data set obtained by aligning the first feature data and the second feature data to the same timeline.

[0071] Specifically, defensive style needs to be determined through cluster analysis of player positions. Using clustering algorithms such as K-means, the position of each player is clustered to form the offensive and defensive lines of the offensive and defensive sides. The relative positions of the offensive and defensive lines determine the team's defensive style.

[0072] The offensive phase is analyzed based on the relative positions of the ball and the attacking players. The offensive process is divided into three main stages: organizing the attack, passing the attack, and threatening the attack. If the ball is relatively far away and the players are organizing the attack, it is in the organizing phase; when the ball is passed to an area closer to the opponent's goal, it enters the threatening attack phase.

[0073] Threat level analysis is based on the ball's position on the field, dividing the field into multiple zones (e.g., the 16 zones defined by Athletic), each with a different threat level.

[0074] Step S105: Determine football match analysis prompts based on the target fusion data, defensive style information, offensive phase information, and threat level information.

[0075] It should be noted that cue word design describes the tracking data, event data, and results of the match situation analysis to guide the large language model in further evaluating the match situation. Cue words transmit key information about the match to the large language model, ensuring that the model can perform accurate analysis based on this information. The cue word generation process can be automated, for example, by using a pre-set template and fitting specific information into the template to generate cue words.

[0076] For example, the following prompts could be designed to provide an overview of the competition:

[0077] "Based on the data from the current moment of the match and the data from previous moments, as well as the aforementioned data (target fusion data, defensive style information, offensive phase information, and threat level information), please tell me the current time of the match, the red card and yellow card substitutions, and from an overall perspective, tell me the current stage of the match and the reasons for analyzing that stage."

[0078] Based on the overall situation of the football match, the following prompts are designed:

[0079] "Based on the data provided (target fusion data, defensive style information, offensive phase information, and threat level information), I would like you to help me analyze the overall game situation from the following three aspects."

[0080] First, from the perspective of team formation, please help me analyze the current formations of the home team and the away team, and the reasons for this analysis, including the advantages and disadvantages of the current formations.

[0081] Second, given the known data including the positions and speeds of all players, please help me analyze the current overall situation of the home team from the perspective of player position distribution and speed, and explain the reasons for this conclusion.

[0082] Third, given the ball's position and speed in the data, please analyze the current situation of the game from the perspective of the ball's speed and position, and explain the reasons for this conclusion.

[0083] Fourth, from the perspective of the relative positions of the ball and players, please help me analyze the current situation and strategies of the home and away teams, and the reasons for this conclusion.

[0084] Finally, based on the above analysis, a summary of the overall situation of the competition is presented.

[0085] Step S106: Input the football match analysis prompts into a large language model trained on big data from real football matches to obtain the football match situation assessment results.

[0086] It should be noted that in this embodiment of the invention, the large language model (e.g., GPT-3.5-turbo) is trained using a large amount of historical football match data, and thus possesses the ability to analyze the match situation. When a prompt word is input into the large language model, the model generates a comprehensive analysis of the current match situation based on the context of the prompt word information. The football match situation assessment results include at least a summary of the match overview, macro-level situation analysis (formation, player positions, ball position, and the relative positions of the ball and players), and local situation analysis (weak areas and strong areas).

[0087] For example, the large language model outputs the following results for summarizing the competition overview:

[0088] "Based on the provided tracking data, we can perform the following analysis: Match time: Current frame number is 621. If we use the total number of frames in the first half, 28081, as a reference, we can estimate as follows: Assuming the first half is approximately 45 minutes long, then there are approximately 28081 / 45 to 624 frames per minute. Therefore, frame 621 roughly corresponds to the end of the first minute of the first half."

[0089] Match Phase: Since the match is in the first half, we can assume that the game is still in its early stages. This period is typically characterized by both teams trying to control the tempo, establish a rhythm on the field, and find weaknesses in the opponent's defense.

[0090] Positional data on the field indicates that the game is in the early stages of the first half, with a potential attacking activity on the right side of the pitch, and the players are just beginning to implement their tactical plans.

[0091] Player Position Analysis: The home team's goalkeeper and several defenders are positioned in their own half, suggesting the home team may be looking for opportunities to launch an attack from a defensive posture. The goalkeeper's position, especially close to the goal, implies the home team is currently in a defensive posture. 2. Ball Position Analysis: The ball is positioned high near the right edge, potentially indicating an ongoing attack down the flank or a long pass across midfield. 3. Relative Positions of the Ball and Players: The forwards are positioned closer to the ball in the opponent's half, possibly preparing to receive crosses or launch attacks using the high space of the opponent's defense.

[0092] Analysis from a formation perspective: Considering the players' positions, the home team may have fielded a formation that emphasizes midfield control, such as 4-4-2 or 4-2-3-1. The players are spread out in a wide attacking and midfield area to control the ball and launch attacks, while the center-backs remain compact to block the opponent's attacks.

[0093] This invention acquires tracking data and event data at any moment during a football match. Feature extraction processing is performed on the tracking data and event data to obtain first feature data after tracking data processing and second feature data after event data processing, ensuring sufficient valuable feature data for subsequent situation analysis. The first and second feature data are time-aligned to obtain time-aligned target fusion data, ensuring consistency of data from different sources on the timeline and allowing for a global perspective in subsequent situation analysis. The target fusion data is analyzed to assess the defensive style, offensive phase, and threat level of the teams, yielding defensive style information, offensive phase information, and threat level information. Football match analysis prompts are determined based on these information, combining the above three analyses with all the fusion data. This ensures the generated prompts have better descriptive capabilities, making the subsequent analysis results of the large language model more closely reflect the actual match process. The football match analysis prompts are input into a large language model trained on big data from real football matches to obtain a football match situation assessment result. By utilizing the trained large language model, a comprehensive football match situation analysis result is generated.

[0094] This embodiment provides a method for assessing the situation of a football match, which can be used with the aforementioned computer. Figure 2 This is a flowchart of a football match situation assessment method according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps:

[0095] Step S201: Obtain tracking data and event data for any moment in the football match. For details, please refer to [link to relevant documentation]. Figure 1 Step S101 of the illustrated embodiment will not be described again here.

[0096] Step S202 involves performing feature extraction processing on the tracking data and event data to obtain first feature data after processing the tracking data and second feature data after processing the event data, respectively. For details, please refer to [link to relevant documentation]. Figure 1 Step S102 of the illustrated embodiment will not be described again here.

[0097] Step S203: Time-align the first feature data and the second feature data to obtain time-aligned target fusion data.

[0098] In one optional implementation, the first feature data includes at least the ball's position information, team information, the identification information of the defending team, the player's identity information, the player's position information, the player's speed information, the coordinates of the court, the match number information, the remaining time of the match, the identification information of the player currently in possession of the ball, and the information of the team currently in possession of the ball.

[0099] The second feature data includes at least the team information where the event occurred, the identity information of the players in the team where the event occurred, the event type information, the event result information, the starting position information of the event, and the ending position information of the event.

[0100] In this embodiment of the invention, the first feature data includes static information about various players, teams, stadiums, and balls, while the second feature data includes dynamic information about events. This provides a sufficient amount of valuable feature data for subsequent situational analysis, thereby improving the accuracy of the analysis and helping the large language model to better understand the game situation.

[0101] Step S204: Analyze the defensive style, offensive phase, and threat level of the teams in the match using the target fusion data to obtain defensive style information, offensive phase information, and threat level information; wherein, the target fusion data represents the data set obtained by aligning the first feature data and the second feature data to the same timeline.

[0102] Specifically, step S204 above includes:

[0103] Step S2041: Analyze the defensive style of the teams in the target fusion data to obtain defensive style information.

[0104] Specifically, step S2041 above includes:

[0105] A1 performs cluster analysis on attacking and defending teams based on player position information, team information, stadium coordinates, ball position information, current ball-controlling player identification information, current ball-controlling team information, event type information, start position information of the event, and end position information of the event, to obtain the position information of each attacking line and each defensive line on the field.

[0106] Specifically, the real-time position information of each player is obtained, including their X and Y coordinates on the field, the ball's position, the player in possession of the ball, and the team in possession of the ball. The starting and ending positions from the event data are also included. A clustering algorithm is used to group the players from both the attacking and defending teams based on their position information, resulting in the spatial distribution of the players. The attacking team's players are divided into three attacking lines: forwards, midfielders, and defenders. The defending team's players are divided into three defensive lines: defenders, midfielders, and forwards. Finally, the attacking and defensive lines are divided, and the average position of each attacking and defensive line is calculated.

[0107] A2. Based on the positional information of each offensive line and each defensive line, determine the permutation and combination information between the offensive and defensive lines.

[0108] Specifically, based on the relative positions of the attacking and defensive lines, there are various combinations of the three attacking lines and the three defensive lines. For example, if all three attacking lines are in front of the defensive lines, it indicates that the attacking side has a strong offensive posture. If all three defensive lines are in front of the attacking lines, it indicates that the defending side has adopted a solid defensive strategy. If the attacking and defensive lines partially intersect, it indicates that the two sides may be in the midfield battle stage.

[0109] A3. Determine the defensive style information of the teams in the match based on the permutation and combination information.

[0110] Specifically, there is no fuzzy classification. Based on the arrangement of the six formation lines (3 for the attacking side and 3 for the defending side), and through combination relationships... There are 20 possible scenarios. Specifically, both the attacking and defending teams are arranged with a defensive-midfield-forward formation. Therefore, you only need to select 3 lines from the 6 lines as the attacking line to automatically form the attacking line, and the remaining 3 lines will automatically form the defensive line. When the attacking and defensive lines are exactly in the same position, a fixed rule can be set, such as arranging them closer to your own goal.

[0111] For example, refer to Figure 3The entire attacking and defending teams are clustered into three attacking lines (a, b, c) and defensive lines (d, e, f). The defensive style is divided into 20 cases based on the distribution of the six lines. 0 represents the attacking team and 1 represents the defending team. Each combination corresponds to a label (e.g., 001011, where 0 is located at positions 1, 2, and 4, so the total number of 0 positions is 1+2+4=7, and the total number of 1 positions is 2+5+6=14). The direction from right to left represents the attack towards the defensive goal. The combination 000111 corresponds to low-pressing defense, and the combination 111000 corresponds to high-pressing defense.

[0112] This invention utilizes cluster analysis based on multi-dimensional data such as player position information, team information, and stadium coordinates to obtain the position information of each offensive and defensive line. This allows for a comprehensive understanding of the team's tactical arrangements during the game. Based on the arrangement and combination information between the offensive and defensive lines, the tactical changes of the team can be analyzed and judged in real time, making the analysis more relevant to the actual game situation and generating specific defensive style information. This provides a data foundation related to defensive style for determining subsequent game prompts.

[0113] Step S2042: Analyze the target fusion data for the offensive phase of the match teams to obtain offensive phase information.

[0114] Specifically, step S2042 above includes:

[0115] B1 determines the first relative position information between the player and the ball based on the player's position information, the coordinates of the field, and the position information of the ball.

[0116] Specifically, the actual distance between a player and the ball can be calculated using geometric formulas and the distance formula between two points.

[0117] B2, determine the type of attack line controlling the ball based on the position information of each attack line, the starting position information of the event, the ending position information of the event, and the first relative position information.

[0118] Specifically, in step A1, the position information of each attack line has been determined and represented by the average position of all players on the line. The starting and ending positions of the pass are obtained from the event data. For example, at the current moment, the ball is controlled by player A and the event is a pass. Based on the starting and ending positions of the pass, the position information of each attack line, and the first relative position, it is determined which attack line the ball is on during the pass.

[0119] B3, determine the offensive phase information based on the type of offensive line controlling the ball.

[0120] For details, please refer to Figure 3 Based on the fact that the ball is controlled by the first (a) attack line, the second (b) attack line, and the third (c) attack line, the attack is divided into three phases: organizing the attack, passing the attack, and threatening the attack.

[0121] This invention, through determining the first relative position information between the player and the ball based on the player's position information, the field coordinate information, and the ball's position information, can obtain a more accurate analysis of the offensive situation. Based on the position information of each offensive line, the start and end position information of the event, and the first relative position information, the type of offensive line controlling the ball is determined, making the classification of offensive methods clearer and more evidence-based. The offensive phase information is determined based on the type of offensive line controlling the ball, providing a data foundation related to offensive phase information for the generation of subsequent match prompts.

[0122] Step S2043: Analyze the threat level of the teams in the target fusion data to obtain threat level information.

[0123] Specifically, step S2043 above includes:

[0124] C1, based on the coordinate information of the court, divides the court area for threat analysis.

[0125] Specifically, such as Figure 4 As shown in (a), the court can be divided into 15 zones, with five zones horizontally and three zones vertically; as... Figure 4 As shown in (b), the field can also be divided into several smaller zones both longitudinally and laterally, following Athletic's proposed 16-zone division method. For example, the field can be longitudinally divided into three attacking zones (your own half, midfield, and the opponent's half), and each zone can be further subdivided into multiple smaller zones; such as... Figure 4 As shown in (c), it can also follow Manchester City coach Pep Guardiola's 20-zone system. Each zone is assigned a unique identifier. For example, zones could be numbered 1 through 16 to represent different zones.

[0126] C2 determines the second relative position information between the ball and the court based on the event type information, the ball's position information, and the court area.

[0127] For example, if the ball is currently in the opponent's half (number 10) and the event is a shot, the relative distance between the ball and the goal area (such as area number 16) will affect the threat assessment. If the ball is close to the penalty area, the threat is higher; if the ball is far from the penalty area, the threat is lower.

[0128] C3, determine the threat level of the teams in the match based on the second relative position information.

[0129] This invention, through its embodiments, divides the threat analysis field area based on the field's coordinate information. By combining event type information, ball position information, and field area information, it determines the second relative position information between the ball and the field, thereby determining the threat level information in a dynamic football match and providing a data foundation related to threat level information for the subsequent generation of match prompts.

[0130] Specifically, the threat level is divided into several levels, namely low threat, medium threat, and high threat.

[0131] For example, player C initiates a pass from the center line, the ball is at (50, 30), the event type is pass, and the ball's destination is inside the opponent's penalty area (number 10); at this time, the distance between the ball and the opponent's goal is calculated to be higher than the preset value, and the pass enters the threat zone (number 10).

[0132] Step S205: Determine football match analysis prompts based on the target fusion data, defensive style information, attacking phase information, and threat level information. For details, please refer to [link to relevant documentation]. Figure 1 Step S105 of the illustrated embodiment will not be described again here.

[0133] Step S206: Input the football match analysis prompts into a large language model trained on big data from real football matches to obtain the football match situation assessment results. For details, please refer to [link to relevant documentation]. Figure 1 Step S106 of the illustrated embodiment will not be described again here.

[0134] This embodiment provides a method for assessing the situation of a football match, the process of which includes the following steps:

[0135] Step S501: Obtain tracking data and event data for any moment in the football match. For details, please refer to [link to relevant documentation]. Figure 1 Step S101 of the illustrated embodiment will not be described again here.

[0136] Step S502 involves performing feature extraction processing on the tracking data and event data to obtain first feature data after processing the tracking data and second feature data after processing the event data, respectively. For details, please refer to [link to relevant documentation]. Figure 1 Step S102 of the illustrated embodiment will not be described again here.

[0137] Step S503: Time-align the first feature data and the second feature data to obtain time-aligned target fusion data. For details, please refer to [link to details]. Figure 1 Step S103 of the illustrated embodiment will not be described again here.

[0138] Step S504: Analyze the target fusion data to determine the defensive style, offensive phase, and threat level of the participating teams, obtaining defensive style information, offensive phase information, and threat level information. The target fusion data represents the data set obtained by aligning the first feature data and the second feature data to the same timeline. For details, please refer to [link to details]. Figure 1 Step S104 of the illustrated embodiment will not be described again here.

[0139] Step S505: Determine football match analysis prompts based on the target fusion data, defensive style information, offensive phase information, and threat level information.

[0140] Specifically, step S505 includes: inputting the data set of the first feature data and the second feature data aligned to the same timeline, along with the corresponding defensive style information, offensive phase information, and threat level information, into a preset prompt word template to obtain football match analysis prompt words; wherein, the preset prompt word template represents a question template obtained by semantic feature analysis based on football match analysis data.

[0141] Understandably, given the diverse dimensions and types of the raw data, directly using this data for analysis would increase the processing burden on large language models and potentially lead to biased analysis results. Therefore, by using prompt word templates, the raw competition data is transformed into structured and semantically clear descriptions, improving descriptive capabilities and ensuring the accuracy of the model's analysis results.

[0142] It should be noted that cue word templates are formatted descriptive frameworks that embed key information from the game (defensive style, offensive phase, threat level, target fusion data) into specific locations within the template. Each template can be used to generate different types of questions or descriptions.

[0143] For example, the content of a prompt word template is as follows:

[0144] "The match data is as follows: At time A, team B1 is at position b, team C1 is at position c, the current attacking team is in area X, the defending team adopts defensive strategy Y, the threat level is Z, did the action of the attacking player s put more pressure on the defending team than Z?" The positions of each letter above are placeholders and can be replaced by actual data.

[0145] Step S506: Input the football match analysis prompts into a large language model trained on big data from real football matches to obtain the football match situation assessment results. For details, please refer to [link to relevant documentation]. Figure 1 Step S106 of the illustrated embodiment will not be described again here.

[0146] This invention integrates first and second feature data aligned to the same timeline, and combines them with corresponding defensive style information, offensive phase information, and threat level information to generate prompts with better descriptive capabilities. The subsequent analysis results of the large language model are more consistent with the actual game process, ensuring that the analysis results are not only comprehensive but also have higher accuracy.

[0147] The present invention provides a method for assessing the situation of a football match, the overall process of which can be referred to as follows: Figure 5 The system acquires the raw features of tracking data and event data at any given moment during the match. It then extracts and processes these raw features to obtain the first feature data after processing. Similarly, it extracts and processes the raw features of the event data to obtain the second feature data after processing. A small-scale match situation analysis model is designed from three perspectives: defensive style, offensive phase, and threat level. Prompt words are designed for the feature data of the tracking and event data, as well as the small-scale situation analysis model. Finally, the feature data of the tracking and event data, the small-scale situation analysis model, and the prompt words are input into a large-scale match situation assessment language model to obtain the football match situation assessment results. These results include a match overview summary, macro-level situation analysis (formation, player positions, ball position, and relative positions of the ball and players), and local situation analysis (weak areas and strong areas).

[0148] Compared to existing technologies that only assess players in contact with the ball, this method efficiently utilizes tracking and event data from all players and the ball within the playing area, enabling a comprehensive and interpretable situational assessment from a global perspective. Unlike existing small-scale situational assessment models that can only evaluate one type of data or situation, this method achieves temporal alignment of tracking and event data, and comprehensively assesses the multi-layered, multi-faceted game situation. Furthermore, the game situational assessment results can assist players and coaches in tactical analysis and post-match debriefing.

[0149] This embodiment also provides a football match situation assessment system, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the system described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0150] This embodiment provides a football match situation assessment system, such as... Figure 6 As shown, it includes:

[0151] The data acquisition module 601 is used to acquire tracking data and event data at any moment during a football match;

[0152] Feature extraction module 602 is used to perform feature extraction processing on the tracking data and event data to obtain first feature data after processing the tracking data and second feature data after processing the event data, respectively.

[0153] The time alignment module 603 is used to align the first feature data and the second feature data in time to obtain time-aligned target fusion data.

[0154] The analysis module 604 is used to analyze the target fusion data to determine the defensive style, offensive phase, and threat level of the teams in the game, and to obtain defensive style information, offensive phase information, and threat level information; wherein, the target fusion data represents the data set obtained by aligning the first feature data and the second feature data to the same timeline;

[0155] The prompt word generation module 605 is used to determine football match analysis prompt words based on the target fusion data, defensive style information, offensive phase information, and threat level information.

[0156] The evaluation result module 606 is used to input the football match analysis prompts into a large language model trained on big data from real football matches to obtain the football match situation evaluation result.

[0157] In one optional implementation, the first feature data includes at least the ball's position information, team information, the identification information of the defending team, the player's identity information, the player's position information, the player's speed information, the coordinates of the court, the match number information, the remaining time of the match, the identification information of the player currently in possession of the ball, and the information of the team currently in possession of the ball.

[0158] The second feature data includes at least the team information where the event occurred, the identity information of the players in the team where the event occurred, the event type information, the event result information, the starting position information of the event, and the ending position information of the event.

[0159] In one alternative implementation, the analysis module 604 specifically includes:

[0160] The clustering analysis unit is used to perform clustering analysis on attacking and defending teams based on player position information, team information, field coordinate information, ball position information, current ball-controlling player identification information, current ball-controlling team information, event type information, start position information of the event, and end position information of the event, to obtain the position information of each attacking line and each defensive line on the field.

[0161] The permutation and combination unit is used to determine the permutation and combination information between the offensive and defensive lines based on the position information of each offensive line and each defensive line.

[0162] The defensive style unit is used to determine the defensive style information of the teams in the game based on the permutation and combination information.

[0163] In one alternative implementation, the analysis module 604 specifically includes:

[0164] The relative position unit is used to determine the first relative position information between the player and the ball based on the player's position information, the coordinate information of the court, and the position information of the ball;

[0165] An attack line type unit is used to determine the attack line type of the ball control based on the position information of each attack line, the start position information of the event, the end position information of the event, and the first relative position information.

[0166] The offensive phase unit is used to determine offensive phase information based on the type of offensive line in possession of the ball.

[0167] In one alternative implementation, the analysis module 604 specifically includes:

[0168] The threat area analysis unit is used to divide the threat analysis field area based on the field's coordinate information;

[0169] The second relative position unit is used to determine the second relative position information between the ball and the court based on the event type information, the ball's position information, and the court area.

[0170] Threat level unit, used to determine the threat level information of the playing team based on the second relative position information.

[0171] In one optional implementation, the prompt word generation module 605 specifically includes:

[0172] The analysis prompt word generation unit is used to input the data set of first feature data and second feature data aligned to the same timeline, as well as the corresponding defensive style information, offensive phase information and threat level information, into a preset prompt word template to obtain football match analysis prompt words; wherein, the preset prompt word template represents a question template obtained by semantic feature analysis based on football match analysis data.

[0173] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0174] In this embodiment, the football match situation assessment system is presented in the form of functional units. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0175] This invention also provides a computer device having the above-described features. Figure 6 The football match situation assessment system shown.

[0176] Please see Figure 7 , Figure 7 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 7 As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 7 Take a processor 10 as an example.

[0177] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0178] The memory 20 stores instructions executable by at least one processor 10 to cause at least one processor 10 to perform the method shown in the above embodiments.

[0179] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0180] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0181] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0182] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0183] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0184] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and all such modifications and variations fall within the defined scope.

Claims

1. A method for assessing the situation of a football match, characterized in that, The method includes: Acquire tracking and event data at any point in a football match; The tracking data and event data are subjected to feature extraction processing to obtain first feature data after tracking data processing and second feature data after event data processing. The first feature data includes at least the ball's position information, team information, the identification information of the defending team, the player's identity information, the player's position information, the player's speed information, the coordinates of the field, the match number information, the remaining time of the match, the identification information of the player currently in possession of the ball, and the information of the team currently in possession of the ball. The second feature data includes at least the team information of the event, the identity information of the players in the team where the event occurred, the event type information, the event result information, the starting position information of the event, and the ending position information of the event. The first feature data and the second feature data are time-aligned to obtain the time-aligned target fusion data. The target fusion data is analyzed to determine the defensive style, offensive phase, and threat level of the teams in the game, resulting in defensive style information, offensive phase information, and threat level information. The target fusion data represents the data set obtained by aligning the first feature data and the second feature data to the same timeline. Football match analysis prompts are determined based on the target fusion data, defensive style information, offensive phase information, and threat level information. The football match analysis prompts are input into a large language model trained on big data from real football matches to obtain the football match situation assessment results. Specifically, the defensive style analysis of the target fusion data is performed on the teams' defensive styles to obtain defensive style information, including: Based on player position information, team information, stadium coordinates, ball position information, current ball possession player identification information, current ball possession team information, event type information, start position information of the event, and end position information of the event, cluster analysis is performed on the attacking and defending teams to obtain the position information of each attacking line and each defensive line on the field. Based on the positional information of each offensive line and each defensive line, determine the permutation and combination information between the offensive and defensive lines; The defensive style information of the teams in the match is determined based on the permutation and combination information.

2. The method according to claim 1, characterized in that, The target fusion data is analyzed during the offensive phase of the match to obtain offensive phase information, specifically including: The first relative position information between the player and the ball is determined based on the player's position information, the coordinates of the field, and the position information of the ball; The type of offensive line controlling the ball is determined based on the position information of each offensive line, the starting position information of the event, the ending position information of the event, and the first relative position information. Determine offensive phase information based on the type of offensive line that controls the ball.

3. The method according to claim 1, characterized in that, The target fusion data is analyzed to determine the threat level of the teams in the match, resulting in threat level information, specifically including: Threat analysis of the court area is based on the court's coordinate information; The second relative position information between the ball and the court is determined based on the event type information, the ball's position information, and the court area. The threat level of the teams in the match is determined based on the second relative position information.

4. The method according to any one of claims 1-3, characterized in that, The process of determining football match analysis prompts based on the target fusion data, defensive style information, offensive phase information, and threat level information specifically includes: The data sets of first and second feature data aligned to the same timeline, along with the corresponding defensive style information, offensive phase information, and threat level information, are input into a preset prompt word template to obtain football match analysis prompt words; wherein, the preset prompt word template represents a question template obtained by semantic feature analysis based on football match analysis data.

5. A football match situation assessment system, characterized in that, The system includes: The data acquisition module is used to acquire tracking data and event data at any moment during a football match; The feature extraction module is used to perform feature extraction processing on the tracking data and event data to obtain first feature data after processing the tracking data and second feature data after processing the event data. The first feature data includes at least the ball's position information, team information, the identification information of the defending team, the player's identity information, the player's position information, the player's speed information, the coordinates of the field, the match number information, the remaining time of the match, the identification information of the player currently in possession of the ball, and the information of the team currently in possession of the ball. The second feature data includes at least the team information of the event, the identity information of the players in the team where the event occurred, the event type information, the event result information, the starting position information of the event, and the ending position information of the event. The time alignment module is used to align the first feature data and the second feature data in time to obtain the time-aligned target fusion data. The analysis module is used to analyze the target fusion data to determine the defensive style, offensive phase, and threat level of the teams in the game, and to obtain defensive style information, offensive phase information, and threat level information; wherein, the target fusion data represents the data set obtained by aligning the first feature data and the second feature data to the same timeline; The prompt word generation module is used to determine football match analysis prompt words based on the target fusion data, defensive style information, offensive phase information, and threat level information. The evaluation result module is used to input the football match analysis prompts into a large language model trained on big data from real football matches to obtain the football match situation evaluation result. The analysis module is specifically used for: Based on player position information, team information, stadium coordinates, ball position information, current ball possession player identification information, current ball possession team information, event type information, start position information of the event, and end position information of the event, cluster analysis is performed on the attacking and defending teams to obtain the position information of each attacking line and each defensive line on the field. Based on the positional information of each offensive line and each defensive line, determine the permutation and combination information between the offensive and defensive lines; The defensive style information of the teams in the match is determined based on the permutation and combination information.

6. A computer device, characterized in that, include: A memory and a processor are communicatively connected, the memory stores computer instructions, and the processor executes the computer instructions to perform the football match situation assessment method according to any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the football match situation assessment method according to any one of claims 1-4.

8. A computer program product, characterized in that, Includes computer instructions for causing a computer to execute the football match situation assessment method according to any one of claims 1-4.

Images