Method and system for recommending cards in card game

Abstract

The invention discloses a card recommendation method and system in a card game, and relates to the technical field of games. Comprising the following steps: collecting current game information of players in real time, namely collecting card group information, tracking extracted cards, monitoring cards in a remaining card pool, analyzing card groups of opponents and recording game states; based on historical game data of the player and the current game information, determining a game style of the player and a targeted strategy in a current game; predicting a card most suitable for the current situation according to the current game information and the targeted strategy; and after the player selects the recommended card from the card pool, analyzing the combination effect of the newly added card and the current card group, and measuring the effectiveness of the recommendation. According to the method, the recommended cards can practically meet the requirements of the current situation, and the method has the remarkable beneficial effects of improving the decision-making ability of players, enhancing the game fun and improving the winning rate.

Description

Technical Field

[0001] The present invention relates to the field of game technology, and in particular to a method and system for recommending cards in a card game. Background Art

[0002] With the rapid development of information technology, online card games have become a form of interactive entertainment widely participated by players around the world. Such games usually involve complex strategies, real-time decision-making and game analysis, thus attracting a large number of players. Traditional card games mainly rely on players' intuition and experience to make decisions. However, with the increase in game content and the number of cards, the amount of information faced by players in the decision-making process has increased exponentially, which has led to players often feeling confused when choosing cards and finding it difficult to quickly identify the optimal strategy and the best card.

[0003] Although the industry has gradually proposed a series of intelligent recommendation systems, the existing card recommendation technology still has many shortcomings. For example, many recommendation systems often ignore the real-time game status and dynamic changes of players and cannot provide accurate real-time feedback. Secondly, many recommendation systems lack in-depth analysis of players' game styles and fail to accurately identify players' strategic tendencies in different situations, which leads to doubts about the effectiveness of the recommendation results. Summary of the invention

[0004] In view of the problems of the existing card recommendation technology, such as insufficient real-time feedback and insufficient in-depth analysis of player styles, the present invention is proposed.

[0005] Therefore, the problem to be solved by the present invention is how to accurately recommend the most suitable card for the current situation in a dynamic environment.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides a method for recommending cards in a card game, comprising: collecting current game information of players in real time during the game; determining the player's game style and targeted strategy in the current game based on the player's historical game data and the current game information; predicting the most suitable card for the current situation based on the current game information and the targeted strategy; and when the player selects a recommended card from a card pool, analyzing the combination effect of the newly added card and the current deck, and measuring the effectiveness of the recommendation.

[0008] As a preferred scheme of the card recommendation method in the card game described in the present invention, the real-time collection of the player's current game information includes collecting deck information, tracking drawn cards, monitoring remaining card pool cards, analyzing the opponent's deck and recording game status.

[0009] As a preferred solution of the card recommendation method in the card game of the present invention, the process of determining the game style of the player includes: extracting historical game records from the player database, and using a classification algorithm to preliminarily classify the overall style of the player according to the historical game records. Assuming that the player's style is divided into offensive, defensive and controlling, the classification formula is:

[0010]

[0011] Among them, G p is the player’s style classification result, G agg , G def , G ctrl Respectively represent offensive, defensive, and controlling styles; ω j is the weight of the jth game in the historical data, B j is the proportion of cards of a certain style used by the player in the jth game, and n is the number of historical game rounds of the player.

[0012] As a preferred embodiment of the card recommendation method in the card game of the present invention, the process of determining the targeted strategy in the current game includes: comparing the current deck structure with the player's historical deck, and preliminarily determining the current game strategy based on the consistency of the player's commonly used strategies; determining the strategy composition of the current deck by analyzing the proportion of various types of cards in the current deck; judging whether the player's strategy for this game is early attack or late burst based on the cost distribution of the current deck, and comprehensively obtaining the targeted strategy in the current game.

[0013] As a preferred solution of the card recommendation method in the card game of the present invention, the prediction of the most suitable card for the current situation includes the following operations: constructing a decision tree prediction model for card recommendation, the prediction model comprehensively considers the synergy between the card and the existing deck, the preliminarily marked player style and the strategic response of the opponent's deck; the decision tree decomposes the card recommendation process into multiple decision nodes, each node makes a judgment based on the current input data, including: whether the cost distribution in the current deck is low, whether the opponent's strategy tends to be defensive, and whether the existing hand needs to supplement resources or defense; the prediction model judges each variable in turn, and matches the player's needs with the most suitable card recommendation through a series of "yes / no" judgments, and the final node of the decision tree is the recommended card; each recommended card is scored, wherein the score SC of each recommended card is i The calculation formula is:

[0014] SC i =α×F i +β×P st +γ×P op +δ×P ne

[0015] Among them, α, β, γ and δ are the weight coefficients of each parameter respectively; F i is the card functionality score, P st For the game stage; P op To evaluate the opponent's deck type, a specific score is given; P ne To set the functional requirement score for the current cards according to the situation; establish priorities based on the scores, and select the 2 to 3 cards with the highest scores as recommended options.

[0016] As a preferred solution of the card recommendation method in the card game of the present invention, the analysis of the combination effect of the newly added card and the current deck includes analyzing the overall synergy between the new card and the existing deck and performing synergy analysis based on the new card and the key cards in the existing deck; the analysis of the overall synergy between the new card and the existing deck includes the following steps: evaluating the synergy between the new card and the current deck:

[0017]

[0018] Among them, C sy Score synergy; ω h is the weight of the hth card in the deck, E h is the synergy score between the new card and the hth card in the deck, N is the number of cards in the current deck; calculate the energy consumption balance and evaluate the changes after the introduction of the new card:

[0019]

[0020] Among them, D var is the standard deviation of the energy consumption of the deck, D vag is the average energy consumption of the deck, indicating the average energy consumption of the deck per round, D h is the energy consumption value of the hth card; combined with the impact on the overall energy consumption fluctuation of the deck:

[0021]

[0022] Where, ΔD im This is the effect of adding new cards on the stability of the deck's energy consumption.

[0023] As a preferred solution of the card recommendation method in the card game of the present invention, the collaborative analysis based on the new card and the key cards in the existing deck includes: evaluating the skill triggering probability of the key card after the new card is introduced:

[0024]

[0025] Among them, P triis the probability of triggering the chain skill of the new card and the key card, K is the number of key cards, PS k is the probability of triggering a single skill of the kth key card, M is the number of additional conditions that need to be met in the skill chain, m is the mth additional condition, Q m is the probability of triggering the mth additional condition.

[0026] In a second aspect, an embodiment of the present invention provides a card recommendation system in a card game, which includes a real-time data collection module for collecting players' current game information in real time, including collecting deck information, tracking drawn cards, monitoring remaining card pool cards, analyzing opponents' decks, and recording game status; a game style analysis module for determining the player's game style and targeted strategy in the current game based on the player's historical game data and the current game information; a card prediction module for predicting the most suitable card for the current situation based on the current game information and the targeted strategy; a combination effect analysis module for analyzing the combination effect of newly added cards and the current deck, and measuring the effectiveness of the recommendation; a recommendation feedback module for tracking the frequency of players adopting recommendations and the win rate of the game, measuring the effectiveness of the recommendation, and providing feedback to optimize the recommendation algorithm.

[0027] In a third aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, the steps of the method for recommending cards in a card game as described in the first aspect of the present invention are implemented.

[0028] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, the steps of the method for recommending cards in a card game as described in the first aspect of the present invention are implemented.

[0029] The beneficial effects of the present invention are as follows: by combining historical data with real-time information, the present invention can deeply explore the personalized style and strategic preferences of players, and can not only identify the typical tactics of players, but also adjust the recommended strategies based on the actual situation of the current game. This personalized targeted strategy enables players to be more targeted when facing different opponents, thereby increasing their chances of winning; in addition, combining the current situation and historical data to predict the most suitable cards can help players make decisions quickly in a rapidly changing game. The prediction model not only takes into account the characteristics of the player's deck, but also pays attention to the opponent's strategy, so that the recommended cards can effectively meet the needs of the current situation; it can be seen that the present invention has significant beneficial effects in improving players' decision-making ability, enhancing game fun and improving the winning rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0031] Figure 1 A flowchart of the proposed method for playing cards in a card game.

[0032] Figure 2 This is a structural diagram of the card recommendation system in the card game.

[0033] Figure 3 FIG. 1 is a schematic diagram of the structure of a terminal in the first embodiment of the present invention.

[0034] Figure 4 It is a schematic diagram of the structure of a computer device in the first embodiment of the present invention. DETAILED DESCRIPTION

[0035] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0036] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0037] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0038] Example 1

[0039] Reference Figure 1 to Figure 4 , which is the first embodiment of the present invention, and which provides a method for recommending cards in a card game, such as Figure 1 As shown, the operation steps include:

[0040] S1: During the game, the player's current game information is collected in real time, including collecting deck information, tracking drawn cards, monitoring remaining card pool cards, analyzing the opponent's deck, and recording game status.

[0041] Specifically, collecting deck information includes obtaining a complete list of the player's current deck, counting the number and proportion of each type of cards in the deck, determining the cost distribution of the deck, and identifying key cards and possible combinations in the deck. Key cards include cards that can have a significant impact on the game situation, such as large-scale damage cards, control cards, or rare powerful cards, or cards that can produce powerful synergistic effects when combined with other cards.

[0042] Tracks the cards that have been drawn, including recording the results of each card drawn by the player and updating the player's hand information.

[0043] Monitoring the remaining card pool includes updating the remaining cards in the card pool in real time, calculating the probability of each type of card in the remaining card pool, and marking special or rare cards.

[0044] Specifically, the probability P of each type of card i in the remaining card pool i The calculation process is:

[0045]

[0046] Among them, C i,re is the number of cards of a certain type i in the remaining card pool, N re is the number of cards in the remaining card pool. For example, there are 30 cards in the initial card pool, including 10 offensive cards, 10 defensive cards, and 10 resource replenishment cards. If the player has drawn 6 cards (including 2 offensive cards, 1 defensive card, and 3 resource replenishment cards), the number of cards of each type in the remaining card pool is: 8 offensive cards, 9 defensive cards, and 7 resource replenishment cards; at this time, the total number of cards in the remaining card pool is 24, and the probability of drawing each type of card is: offensive cards are about 0.33, defensive cards are about 0.38, and resource replenishment cards are about 0.29.

[0047] Analyzing the opponent's deck includes recording the cards that the opponent has used and matching this data with known common deck types to infer the opponent's deck type.

[0048] Recording the game status includes tracking the current round number, recording key data such as the health and resource points of both players; and integrating all collected data into a unified data structure.

[0049] By collecting real-time data, we can dynamically understand the player's deck construction and the cards available in the current card pool, ensuring that the recommended cards are relevant to the current game situation and avoiding the recommendation of irrelevant cards.

[0050] S2: Based on the player's historical game data and current game information, determine the player's game style (such as offense, defense, or control) and targeted strategy in the current game.

[0051] Extract historical game records from the player database, preliminarily classify the player's overall style based on the historical deck structure and usage frequency, use the current deck information in S1 to analyze the player's possible strategic tendency in this game, and mark the player's style. The specific operations include:

[0052] Historical game records include the types and configurations of decks used by players in the past, the frequency of use of each card in historical games, and the player's winning rate in different game modes, different opponents, and different strategies.

[0053] According to the historical game records, the classification algorithm is used to preliminarily classify the overall style of the players. Assuming that the player's style can be divided into offensive, defensive and controlling, the classification formula is:

[0054]

[0055] Among them, G p is the player’s style classification result, G agg , G def , G ctrl Respectively represent offensive, defensive, and controlling styles; ω j is the weight of the jth game in the historical data, B j is the proportion of cards of a certain style used by the player in the jth game, and n is the number of historical game rounds of the player.

[0056] Compare the current deck structure with the player's historical deck, and preliminarily determine the current game strategy based on the consistency of the player's commonly used strategies. For example, if the player has used control cards many times in history and the current deck has a similar configuration, it is determined that the control strategy is inclined to be used. By analyzing the proportion of various types of cards in the current deck (offense, defense, or control), the strategy composition of the current deck is determined, and based on the cost distribution of the current deck, it is determined whether the player's strategy for this game is early attack (low-cost deck) or late burst (high-cost deck), that is, a complete targeted strategy in the current game is obtained. For example: if there are many offensive cards in the deck and the cost distribution is low, it can be inferred that the player will adopt an early offensive strategy, trying to quickly weaken the opponent through continuous pressure in the first few rounds of the game; if the deck contains a large number of control or defensive cards, and the cost distribution is high, the player may tend to delay and counter-attack strategies, by weakening the opponent's offensive and relying on high-cost cards in the later stage to win, etc.

[0057] By combining historical game styles and current deck information, the player's style in the current game can be preliminarily marked.

[0058] It should be noted that by identifying the player's gaming style, the system can customize recommended cards for players of different styles, provide more personalized recommendations, and enable players to build decks that better suit their strategic needs.

[0059] S3: Based on the player’s current game information and targeted strategies, use the decision tree prediction model to predict the most suitable card for the current situation.

[0060] S3.1: Based on the real-time data collected in S1 and S2, a decision tree prediction model for card recommendation is constructed. The prediction model uses the player’s current deck information, game status, cards in the remaining card pool and their probabilities, opponent deck information, etc. as input variables. The output of the prediction model is to recommend the card that best suits the current situation.

[0061] The prediction model comprehensively considers the synergy between the card and the existing deck, the player's style that has been initially marked, and the strategic response of the opponent's deck. The specific operations are as follows:

[0062] The decision tree breaks down the card recommendation process into multiple decision nodes, each of which makes decisions based on the current input data, including: Is the cost distribution in the current deck low (early offensive strategy)? Does the opponent's strategy tend to be defensive (more offensive cards are needed)? Does the existing hand need to replenish resources or defense?

[0063] The prediction model will judge each variable in turn (such as the current situation, opponent strategy, card synergy, etc.), and through a series of "yes / no" judgments, it will match the player's needs with the most appropriate card recommendation. The final node of the decision tree is the recommended card.

[0064] S3.2: Select all recommended cards again and use calculation methods to comprehensively evaluate the effectiveness of each card in the current situation. The specific steps are as follows:

[0065] Score each recommended card. The score of each recommended card is SC. i The calculation formula is:

[0066] SC i =α×F i +β×P st +γ×P op +δ×P ne

[0067] Among them, α, β, γ and δ are the weight coefficients of each parameter respectively; F i Score the functionality of the card, such as attack power, control ability, etc. st is the game stage, such as 1 (early), 2 (middle), 3 (late), etc.; P opTo evaluate the opponent's deck type, a specific score is given (such as control type, offensive type, etc.); ne To set the current functional requirement score for the card according to the situation.

[0068] Establish priorities based on the ratings, and select the 2-3 highest-rated cards as recommended options.

[0069] It can be seen that through the predictive model, the system can recommend the most advantageous cards to players based on the current situation, card combinations and player preferences, ensuring that the recommended cards are not only compatible with the existing decks, but also able to cope with the current game situation.

[0070] S4: After the player selects a recommended card from the card pool, analyze the combination effect of the newly added card and the current deck to measure the effectiveness of the recommendation.

[0071] S4.1: Analyzing the combined effect of newly added cards with the current deck includes: analyzing the overall synergy between the new cards and the existing deck and analyzing the synergy between the new cards and the key cards in the existing deck to evaluate whether the new cards can trigger important combination skills or strategies in the deck.

[0072] S4.1.1: Analyze the overall synergy of new cards with existing decks including:

[0073] To evaluate the synergy between a new card and the current deck, a synergy scoring model can be used to calculate the contribution of the new card to the overall deck through weighted average:

[0074]

[0075] Among them, C sy Synergy score, which generally represents the combined effect of the new card and the existing deck. The higher the value, the higher the match between the new card and the deck. h is the weight of the hth card in the deck, E h is the synergy score between the new card and the hth card in the deck, such as damage bonus, skill triggering, etc. N is the number of cards in the current deck; if the new card can directly enhance the skill effect of h cards, it will be given a higher score, such as E h =1.5; if there is no obvious synergy, then E h =1 or lower.

[0076] Furthermore, to evaluate whether the new card can reasonably match the energy consumption of the existing deck, calculate the energy consumption balance and evaluate the changes after the introduction of the new card:

[0077]

[0078] Among them, D varis the standard deviation of the energy consumption of the deck, D vag is the average energy consumption of the deck, indicating the average energy consumption of the deck per round, D h is the energy consumption value of the hth card.

[0079] Combined with the impact on the overall energy consumption fluctuation of the deck:

[0080]

[0081] Where, ΔD im This is the effect of adding new cards on the stability of the deck's energy consumption. If the value is too large, it means that the energy requirement of the new card may be too high, destroying the energy balance of the deck.

[0082] According to ΔD im The numerical value of ΔD is evaluated at multiple levels, such as im ≈0 means slight impact or no impact, 0<ΔD im ≤1 means slightly increasing energy consumption, 1<ΔD im ≤2 for medium impact, etc. Quantify the impact of new cards on energy balance and develop corresponding strategies for different levels of impact. Small increases in energy cost are usually acceptable, especially when the skills or synergies of new cards are very strong; while larger increases in energy cost need to be managed carefully to ensure that the overall energy usage of the deck is not restricted in key rounds.

[0083] S4.1.2: Perform synergy analysis based on the new card and key cards in the existing deck, including:

[0084] Evaluate the probability of skill triggering of key cards after the introduction of new cards, assuming that there are certain conditions (such as number of turns, energy) between cards to trigger chain skills or multiple conditions that need to be met within a specific turn:

[0085]

[0086] Among them, P tri is the probability of triggering the chain skill of the new card and the key card, K is the number of key cards, PS k is the probability of triggering a single skill of the kth key card, M is the number of additional conditions that need to be met in the skill chain (e.g., a round in which a certain energy is reached, another card appears in a certain round), m is the mth additional condition, Q m is the probability of triggering the mth additional condition.

[0087] According to P tri Compared with multi-level thresholds, different levels correspond to different suggestions. Layered analysis can help players better evaluate the synergy value of new cards.

[0088] Based on the addition of new cards, we identify cards in the deck that are redundant or duplicate the functions of new cards, and recommend replacing these cards to improve the overall synergy of the deck. For example, if a new card provides the same or better effect, we may recommend replacing an old card with a similar but weaker effect.

[0089] In addition, based on the properties of the new cards, players may be advised to adjust their game strategies, such as making greater use of the characteristics of the new cards for offense or defense, or changing the core strategy of the deck (such as changing from defense to offense).

[0090] Optionally, if the player does not choose the recommended card, the recommendation parameters are updated and adjusted based on the player's feedback and game results to continuously optimize the prediction model.

[0091] S4.2: Measure the effectiveness of recommendations by tracking how often players adopt them and their game win rates.

[0092] Specifically, a weighted sum is taken based on the frequency with which players adopt recommendations and the win rate of the game. If the result is greater than the preset result threshold, the recommendation is considered valid. Otherwise, the recommendation is considered invalid. The player's choice is recorded, and the recommendation parameters are updated and adjusted based on the player's feedback and game results. The prediction model is continuously optimized to make future card recommendations more accurate and adapt to the personalized needs and changes in game styles of different players.

[0093] Preferably, as the game progresses, cards in the card pool are constantly drawn, and the player's optional cards will decrease. The recommended strategy is updated in real time according to the changes in the card pool, and the recommended card list is dynamically adjusted. Especially in the later stages, the system will give priority to recommending key cards required for the game to help players complete the construction of the final strategy.

[0094] Dynamically adjust the recommendation list to ensure that the cards recommended by the system are always consistent with the current card pool, and can respond flexibly according to the progress of the game to enhance the player's combat experience.

[0095] Furthermore, when recommending cards, the system will provide players with detailed card presentations, including the card's effect, the reason for the recommendation, and the synergy with the current deck. It can also provide simple simulation analysis to help players quickly understand the effect of using recommended cards in specific situations. Through detailed card presentations and recommendation explanations, players can more intuitively understand the value of recommended cards, helping players make more informed card choices and improve game efficiency.

[0096] Furthermore, after the recommendation results are displayed, the system also provides an overall management function for the deck, allowing players to view the current status of the deck (such as cost distribution, attack power, defense power, spell strength, etc.) and make real-time adjustments based on the recommendation results. Players can customize the recommended decks and save multiple deck plans to meet different battle needs.

[0097] Further, such as Figure 2 As shown, this embodiment also provides a card recommendation system in a card game, including:

[0098] A real-time data collection module is used to collect the player's current game information in real time, including collecting deck information, tracking drawn cards, monitoring remaining card pool cards, analyzing the opponent's deck, and recording game status; a game style analysis module is used to determine the player's game style and targeted strategy in the current game based on the player's historical game data and the current game information; a card prediction module is used to predict the most suitable card for the current situation based on the current game information and the targeted strategy; a combination effect analysis module is used to analyze the combination effect of newly added cards and the current deck, and measure the effectiveness of the recommendation; a recommendation feedback module is used to track the frequency of players adopting recommendations and the win rate of the game, measure the effectiveness of the recommendation, and provide feedback to optimize the recommendation algorithm.

[0099] like Figure 4 This embodiment also provides a computer device, which is suitable for the case of a card recommendation method in a card game, including a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to implement the card recommendation method in the card game proposed in the above embodiment.

[0100] like Figure 3 , the computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a button, a trackball or a touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0101] This embodiment further provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for recommending cards in a card game as proposed in the above embodiment.

[0102] In summary, the present invention, by combining historical data with real-time information, can deeply explore the personalized style and strategic preferences of players. It can not only identify the typical tactics of players, but also adjust the recommended strategies based on the actual situation of the current game. This personalized targeted strategy enables players to be more targeted when facing different opponents, thereby increasing their chances of winning. In addition, combining the current situation and historical data to predict the most suitable cards can help players make decisions quickly in a rapidly changing game. The prediction model not only takes into account the characteristics of the player's deck, but also pays attention to the opponent's strategy, so that the recommended cards can effectively meet the needs of the current situation. It can be seen that the present invention has significant beneficial effects in improving players' decision-making ability, enhancing game fun and improving winning rate.

[0103] Example 2

[0104] Referring to Table 1, which is a second embodiment of the present invention, this embodiment provides a method for recommending cards in a card game. In order to verify the beneficial effects of the present invention, a scientific demonstration is carried out through simulation experiments.

[0105] In order to verify the effectiveness of the card recommendation method in the card game proposed by the present invention, the researchers designed a card game called "Magic War" and invited 300 players to participate in a 30-day test. The game has 1,000 different cards, divided into four types: attack, defense, control, and resources. Each player initially randomly receives 30 cards as the starting deck.

[0106] The testing process is as follows:

[0107] First, the researchers developed a card recommendation system based on the method of the present invention. The system collects players' game information in real time, including deck composition, drawn cards, remaining card pool status, opponent deck analysis, game status and other data. The system uses an adaptive algorithm to continuously adjust the probability of cards in the remaining card pool to ensure the accuracy of the recommendation.

[0108] Secondly, the researchers designed a complex player style recognition module that not only analyzes the player's historical game data, but also combines the real-time information of the current game and uses machine learning algorithms to dynamically adjust the player's style label. For example, a player who usually prefers defense may adopt a more aggressive strategy in a specific game, and the system can capture this change in time and adjust the recommended strategy accordingly.

[0109] The decision tree prediction model takes into account the current situation and player style, and continuously optimizes its prediction accuracy through a large number of simulated games. It can provide players with the best card selection suggestions in a complex and changing game environment.

[0110] The comprehensive score of each card is calculated in real time using a dynamic weight adjustment algorithm, taking into account the functionality of the card, the game stage, the opponent's strategy, and the needs of the current situation. This ensures that the recommended cards not only meet the player's long-term strategy, but also specifically address the special needs of the current situation.

[0111] During the 30-day testing period, the system recorded the game data of each player, including indicators such as card recommendation adoption rate, win rate changes, and deck synergy scores. The researchers also invited 100 professional players to form a control group, who used traditional manual card selection methods to play games.

[0112] The test results are shown in the following table:

[0113] Table 1

[0114]

[0115] Through in-depth analysis of the above test results, it can be concluded that in terms of average win rate, the test groups (A, B, C) using the method of the present invention are significantly higher than the control group. The average win rate of the test group is between 56.2% and 58.7%, while that of the control group is between 50.8% and 52.1%. This gap fully proves that the method of the present invention can effectively improve the game performance of players; in particular, the 58.7% win rate of test group A is 6.6 percentage points higher than the best control group (control group 2, 52.1%), which is a significant advantage in competitive games.

[0116] Secondly, the card recommendation adoption rate reflects the practicality and player acceptance of the method of the present invention. The adoption rates of the three test groups were all above 79%, with the highest reaching 82.3%, which shows that players have a high degree of trust and acceptance of the cards recommended by the system. Such a high adoption rate not only proves the accuracy of the recommendation system, but also reflects that the system can well understand and adapt to the strategic needs of players.

[0117] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for recommending cards in a card game, characterized in that: include, During the game, the player's current game information is collected in real time; Determine the player's game style and targeted strategy in the current game based on the player's historical game data and the current game information; Predicting the most suitable card for the current situation based on the current game information and the targeted strategy; After the player selects a recommended card from the card pool, the combination effect of the newly added card and the current deck is analyzed, and the effectiveness of the recommendation is measured.

2. The card recommendation method in a card game as claimed in claim 1, characterized in that: The real-time collection of the player's current game information includes collecting card group information, tracking the drawn cards, monitoring the remaining card pool cards, analyzing the opponent's card group and recording the game status.

3. The card recommendation method in a card game as claimed in claim 2, characterized in that: The process of determining the player's game style includes: Historical game records are extracted from the player database. Based on the historical game records, the overall style of the players is preliminarily classified using a classification algorithm. Assuming that the player's style is divided into offensive, defensive and controlling, the classification formula is: Among them, G p is the player’s style classification result, G agg , G def , G ctrl Respectively represent offensive, defensive, and controlling styles; ω j is the weight of the jth game in the historical data, B j is the proportion of cards of a certain style used by the player in the jth game, and n is the number of historical game rounds of the player.

4. The card recommendation method in a card game as claimed in claim 3, characterized in that: The process of determining the targeted strategy in the current game includes: Compare the current deck structure with the player's historical decks, and preliminarily determine the current game strategy based on the consistency of the player's commonly used strategies; Determine the strategy composition of the current deck by analyzing the proportion of various cards in the current deck; Based on the cost distribution of the current deck, determine whether the player's strategy for this game is early attack or late burst, and comprehensively obtain a targeted strategy for the current game.

5. The card recommendation method in a card game as claimed in claim 4, characterized in that: The cards predicted to be most suitable for the current situation include the following actions: Constructing a decision tree prediction model for card recommendation, the prediction model comprehensively considers the synergy of the card with the existing deck, the preliminary marked player style and the strategic response of the opponent's deck; The decision tree breaks down the card recommendation process into multiple decision nodes. Each node makes a decision based on the current input data, including: whether the cost distribution in the current deck is low, whether the opponent's strategy tends to be defensive, and whether the existing hand needs to replenish resources or defense; The prediction model will judge each variable in turn, and through a series of "yes / no" judgments, match the player's needs with the most appropriate card recommendation. The final node of the decision tree is the recommended card; Score each recommended card, where the score SC of each recommended card is i The calculation formula is: SC i =α×F i +β×P st +γ×P op +δ×P ne Among them, α, β, γ and δ are the weight coefficients of each parameter respectively; F i is the card functionality score, P st For the game stage; P op To evaluate the opponent's deck type, a specific score is given; P ne To set the current functional requirement score for the card according to the situation; Establish priorities based on the ratings, and select the 2-3 highest-rated cards as recommended options.

6. The card recommendation method in a card game as claimed in claim 5, characterized in that: The analyzing the combined effect of the newly added card and the current deck includes analyzing the overall synergy between the new card and the existing deck and performing synergy analysis based on the new card and key cards in the existing deck; The analysis of the overall synergy between the new card and the existing deck includes the following steps: Evaluate the synergy of new cards with the current deck: Among them, C sy Score synergy; ω h is the weight of the hth card in the deck, E h is the synergy score between the new card and the hth card in the deck, and N is the number of cards in the current deck; Calculate the energy consumption balance and evaluate changes after the introduction of new cards: Among them, D var is the standard deviation of the energy consumption of the deck, D avg is the average energy consumption of the deck, indicating the average energy consumption of the deck per round, D h is the energy consumption value of the hth card; Combined with the impact on the overall energy consumption fluctuation of the deck: Where, ΔD im This is the effect of adding new cards on the stability of the deck's energy consumption.

7. The card recommendation method in a card game as claimed in claim 6, characterized in that: The collaborative analysis based on the new card and the key cards in the existing deck includes: Evaluate the skill triggering probability of key cards after the introduction of new cards: Among them, P tri is the probability of triggering the chain skill of the new card and the key card, K is the number of key cards, PS k is the probability of triggering a single skill of the kth key card, M is the number of additional conditions that need to be met in the skill chain, m is the mth additional condition, Q m is the probability of triggering the mth additional condition.

8. A card recommendation system in a card game, based on the card recommendation method in a card game according to any one of claims 1 to 7, characterized in that: Also includes, Real-time data collection module, used to collect players' current game information in real time, including collecting deck information, tracking drawn cards, monitoring remaining card pool cards, analyzing opponents' decks, and recording game status; A game style analysis module, used to determine the player's game style and targeted strategy in the current game based on the player's historical game data and the current game information; A card prediction module, used to predict the most suitable card for the current situation based on the current game information and the targeted strategy; Combination effect analysis module, used to analyze the combination effect of newly added cards and the current deck, and measure the effectiveness of the recommendation; The recommendation feedback module is used to track the frequency of players adopting recommendations and the win rate of games, measure the effectiveness of recommendations, and provide feedback to optimize the recommendation algorithm.

Images