A matching method for optimizing a 100-player four-row online competitive game

Abstract

The present invention provides a method for optimizing the matching mechanism in a 100-player four-row online competitive game, and relates to a game matching mechanism technology. The method uses the KD (number of defeats / number of eliminations) data of 100 players in nearly 20 games within 30 seconds after the start of the game, and divides the 100 players into groups of 4, a total of 25 groups, by linear regression, normal distribution, confidence interval and other methods. The average score of each group is similar, so that the level of each group of players is in a specific interval, thereby ensuring the fairness of the game. Using the method of the present invention, on the one hand, the level of each group of players is similar, the game experience is increased, and there is no situation where any team is particularly strong or weak; on the other hand, it is conducive to promoting the development of the game industry and having a good game experience.

Description

Technical Field

[0001] The present invention relates to the field of optimization of game matching mechanisms, and in particular to player information collection and data processing, and specifically refers to a method for optimizing a matching mechanism in an online competitive game of 100 players in four rows. Technical Background

[0002] With the advancement of technology, more and more entertainment facilities have become an indispensable part of people's lives as a way to increase fun. Games have obviously become a part of the daily life of young people. Appropriate games are beneficial to physical and mental health, but some game matching mechanisms do make people feel that the gap in strength between themselves and their opponents is too large, leading to unpleasantness in the game.

[0003] Therefore, how to give us a better gaming experience and solve problems such as unbalanced matching mechanisms has become a widely concerned issue among young people in China. Some methods have also been proposed, but there are no specific implementation measures. At present, how to optimize the matching mechanism, solve the problem of unbalanced player strength, and improve the player's gaming experience has become a problem that game developers must solve. Summary of the invention

[0004] The purpose of the present invention is to provide a method for optimizing the matching mechanism in an online competitive game of 100 players in four rows, which can effectively solve the problem of unbalanced matching mechanism and improve the game experience of players.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0006] A method for optimizing a matching mechanism in a 100-player four-row online competitive game, characterized in that it comprises the following steps:

[0007] 1) Collection of player information

[0008] Collect KD data of nearly 20 games of N players who entered the game in the first 30 seconds.

[0009] 2) Processing of player information

[0010] Use the linear regression method to write a linear regression function for each player's 20-game KD data and the distance in time, draw a linear regression graph, and use this to predict the next player's KD data to get the S value.

[0011] 3) Selection of scoring range

[0012] The S values ​​of 100 players are randomly selected from the matched players with a confidence level of 68.27%, and the final score interval is calculated using the confidence interval calculation method.

[0013] 4) Generate the original matrix

[0014] 100 players whose S values ​​are within the scoring range are randomly selected from the matched players and randomly arranged into an original matrix of 4 rows and 25 columns. The S value of the original matrix forms a one-to-one correspondence with each player.

[0015] 5) Final player combination

[0016] The four players corresponding to the four S values ​​in each column of the original matrix form a combination, namely S1, S2, S3, and S4, forming a total of 25 combinations. The game starts, and the remaining players are assigned to other intervals.

[0017] In step 2), the player information is processed by making a linear regression equation between the 20 sets of KD and time for each player, and a corresponding linear regression graph is drawn, with the horizontal axis being time and the vertical axis being KD. The next KD is predicted to obtain the S value, and each calculation result corresponds one-to-one with the player information.

[0018] In step 3), the selection of the scoring interval is specifically as follows: randomly select 100 players to obtain their S values, calculate the mean value u and standard deviation δ of these 100 sets of data, select the confidence level as 68.27%, and finally obtain the scoring interval as [u-δ, u+δ].

[0019] In step 4), the original matrix is ​​generated by extracting players whose S values ​​are in the scoring range. Since a large number of people are matched at the same time, multiple scoring ranges can be generated, so the players in the scoring range are matched to form the original matrix.

[0020] In step 5), the final player combination is generated. Multiple scoring intervals are calculated each time, and 100 players are randomly selected in each scoring interval to generate a matrix of 4 rows and 25 columns. The players in each column form a combination, which can ensure that the strength difference of each team is not too large. After completion, another 100 players are drawn and the allocation is repeated. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the operation flow of the present invention.

[0022] Figure 2 Schematic diagram of linear regression formula of embodiment.

[0023] Figure 3 Schematic diagram of the normal distribution of KD data of 100 players extracted in the embodiment.

[0024] Figure 4 Schematic diagram of the confidence interval calculation formula of the embodiment.

[0025] Figure 5 This is a schematic diagram of the player's S value in row 4 and column 25 of Example 1.

[0026] Figure 6 Schematic diagram of player team formation in an embodiment. DETAILED DESCRIPTION

[0027] The following focuses on three key steps in the technical solution of the present invention in combination with the accompanying drawings and examples, which is conducive to a better understanding of the technical solution of the present invention.

[0028] Example

[0029] A method for optimizing a matching mechanism in a 100-player four-row online competitive game, characterized in that it comprises the following steps:

[0030] 1) Collection of player information

[0031] Collect KD data of nearly 20 games of N players who entered the game in the first 30 seconds.

[0032] 2) Processing of player information

[0033] Use the linear regression method to write a linear regression function for each player's 20-game KD (number of defeats / number of eliminations) data and the distance in time, draw a linear regression graph, and use this to predict the next player's KD data to get the S value.

[0034] 3) Selection of scoring range

[0035] The S values ​​of 100 players are randomly selected from the matched players with a confidence level of 68.27%, and the final score interval is calculated using the confidence interval calculation method.

[0036] 4) Generate the original matrix

[0037] 100 players whose S values ​​are within the scoring range are randomly selected from the matched players and randomly arranged into an original matrix of 4 rows and 25 columns. The S value of the original matrix forms a one-to-one correspondence with each player.

[0038] 5) Final player combination

[0039] The four players corresponding to the four S values ​​in each column of the original matrix form a combination, namely S1, S2, S3, and S4, forming a total of 25 combinations. The game starts, and the remaining players are assigned to other intervals.

[0040] Each step is described in detail below.

[0041] In step 2), the linear regression method is used to process the player information. Here, it is necessary to discard the extreme values, find the average value of the 20 data, find the points with large differences and remove them, record the time of each player playing as x, and record the KD data of this time as y. The linear regression equation formula (such as Figure 2As shown in the figure, we can get the linear regression function y=ax+b, draw the corresponding curve on the coordinate axis and find the y value corresponding to x=21, which is the predicted KD value we will get this time.

[0042] In step 3), the scoring interval is selected by using the normal confidence interval method. The S values ​​calculated for the 100 randomly selected players follow the normal distribution, such as Figure 3 As shown, we can get the mean and standard deviation of this set of data, recorded as u and δ respectively. Since the standard deviation is known, let its confidence level 1-a be 0.6827. Therefore, the confidence interval of the 68.27% credibility of these 100 players is (as shown in Figure 4 (shown)

[0043]

[0044] The corresponding Z value at the confidence level of 0.6827 is 1, so the above formula can be written as:

[0045]

[0046] This is the final scoring range.

[0047] In step 4), the corresponding original matrix is ​​generated. Here, 100 players in the scoring range are randomly selected and divided into the following Figure 5 The original matrix shown in the figure, if there are less than 100 players, the draw will continue. If 100 players in the scoring range have not been drawn within 30 seconds, only players with an integer multiple of 4 will be drawn, and the rest will be replaced by human-machine players, which are located in the last few columns of the matrix. These players are numbered from 1 to 100 in each column.

[0048] In step 5), the final game players are obtained. Since the original matrix is ​​already given, we only need to group the four players into teams by column, such as Figure 6 As shown, since the players are drawn within the scoring range, their strength is roughly equal (except for humans and machines). Humans and machines are in the last few columns of the matrix. Since only players with an integer multiple of 4 are drawn, they can just form a team of four. The addition of humans and machines can make the game more fun.

Claims

1. A method for optimizing matching in a 100-player four-row online competitive game, characterized in that: The following steps are involved: 1) Collection of player information Collect KD data of the last 20 games of N players who entered the game in the first 30 seconds; 2) Processing of player information Use the linear regression method to write a linear regression function of each player's 20-game KD data and the time distance, draw a linear regression graph, and use it to predict the next player's KD data to get the S value, where KD is the number of defeats / eliminations; 3) Selection of scoring range Randomly select the S values ​​of 100 players from the matched players, with a confidence level of 68.27%, and then use the confidence interval calculation method to calculate the final score interval; 4) Generate the original matrix Randomly select 100 players whose S values ​​are within the scoring range from the matched players and randomly arrange them into an original matrix of 4 rows and 25 columns. The S value of the original matrix forms a one-to-one correspondence with each player. 5) Final player combination The four players corresponding to the four S values ​​in each column of the original matrix form a combination, forming a total of 25 combinations, and the game begins; In step 2), the player information is processed by making a linear regression equation between the 20 sets of KD and time for each player, and a corresponding linear regression graph is drawn, with the horizontal axis being time and the vertical axis being KD, and the next KD is predicted to obtain the S value, and each calculation result corresponds to the player information one by one; In step 3), the selection of the scoring interval is as follows: randomly select 100 players to obtain their S values, calculate the mean value u and standard deviation δ of these 100 sets of data, select the confidence level as 68.27%, and finally obtain the scoring interval as [u-δ, u+δ]; In step 4), the original matrix is ​​generated, and players whose S values ​​are in the scoring interval are extracted. Since there are many people matching at the same time, multiple scoring intervals are generated, so the players in the scoring intervals are matched to form the original matrix; In step 5), the final player combination is generated; each time multiple scoring intervals are calculated, 100 players are randomly selected in each scoring interval to generate a matrix of 4 rows and 25 columns, and the players in each column form a combination. After completion, another 100 players are drawn and the allocation is repeated.

Images