Diving competition athlete score data quality evaluation method

Abstract

The invention discloses a diving competition athlete score data quality evaluation method. According to the method, score data of diving competition athletes on an international swimline official website are obtained, and the score data of the diving competition athletes are obtained from the four aspects of the whole diving competition, each round of diving competition, each athlete and each diving action; in each aspect, a Kendall coordination coefficient and a Kendall correlation coefficient are used as evaluation indexes from three dimensions of data quality of the overall score, data quality between every two multiple measurement scores and data quality between the multiple measurement scores and the final measurement score; evaluating the score data quality of the athletes in the diving competition according to the Kendall coordination coefficient and the Kendall correlation coefficient, and finally achieving the comprehensive and systematic evaluation of the score data quality of the athletes in the diving competition. According to the invention, on the basis of evaluating the score data quality of the athletes from multiple angles, the defects and improvement directions of the score data quality of the athletes can be determined.

Description

Technical Field

[0001] This invention relates to a method for evaluating the quality of diving competition athlete performance data. Background Technology

[0002] To ensure the healthy development of sports, athlete performance should be an objective measurement of an athlete's ability. However, in many sports competitions, athlete performance is often disputed due to various factors, primarily manifested in questions about the quality of the performance data. Therefore, evaluating the quality of athlete performance data is crucial for maintaining fairness and impartiality in sports competitions.

[0003] Currently, the main methods for evaluating the quality of athlete performance data include analysis of variance, Spearman's rank correlation coefficient, Kendall's coefficient of concordance, ICC intragroup correlation coefficient, and Cronbach's coefficient.

[0004] In his article "An Analysis of the Level of Singers and Judges in Singing Competitions," Gao Shunkang used two-way ANOVA (with the two factors being singers and judges) to evaluate the quality of data on the performance of young singers in competitions. Wang Baohe et al., in their article "A Mathematical Model for Evaluating the Level of Judges in Sports Competitions and Its Application," used Spearman's rank correlation coefficient to evaluate the quality of data on the performance of athletes in a national sports competition. He Jiangchuan, in his article "A Study on Non-parametric Evaluation Methods of Subjective Scoring of Judges' Levels," used Spearman's rank correlation coefficient and Kendall's coefficient of concordance to evaluate the quality of data on the performance of gymnasts in competitions. In their article, "Is the Quality of Judging in Women's Artistic Gymnastics Equivalent at Major Competitions of Different Levels?", Pajek et al. evaluated the data validity of gymnastics athletes' performance using the mean deviation of athlete performances, Kendall's coefficient of concordance, and partial Eta-squared analysis of variance. They assessed the reliability and consistency of the data by calculating the mean absolute deviation between athlete performances and final scores, Cronbach's alpha coefficient, ICC intragroup correlation coefficient, and Armor's theta coefficient. Jerneja... In their paper "Reliability of Judging in DanceSport," Deng used Kendall correlation coefficient and ICC within-group correlation coefficient to analyze the data quality of athletes' performance in a dance competition. Nahoko Sato, in her paper "Improving reliability and validity in hip-hop dance assessment: Judging standards that elevate the sport and competition," analyzed the reliability and validity of scores from a hip-hop competition system across five performance categories: creativity, expression and interpretation, impression, technical quality, and synchronicity, conducted between 2014 and 2019.

[0005] However, the main problem with the above-mentioned methods for evaluating the quality of athlete performance data is:

[0006] The data quality evaluation is based on a single perspective and dimension, only analyzing the overall data quality of athletes' performance throughout the entire competition, without comprehensively and systematically analyzing the data quality of athletes' performance. Therefore, it is impossible to determine the deficiencies in the data quality of athletes' performance and the direction for improvement.

[0007] This invention proposes a method for evaluating the quality of diving competition athlete performance data. It evaluates the data quality from four aspects: the entire diving competition, each round of diving, each athlete, and each type of diving movement. Each aspect is further evaluated from three dimensions: the overall performance data quality, the data quality between pairs of multiple measurements, and the data quality between multiple measurements and the final measured result. This method can comprehensively and systematically grasp the data quality of athlete performance, identify deficiencies in the data quality, and determine directions for improvement. Summary of the Invention

[0008] The technical problem solved by this invention is to overcome the shortcomings of existing analysis and provide a method for evaluating the quality of diving athlete performance data.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a method for evaluating the quality of diving athlete performance data, comprising the following steps:

[0010] S1: Data acquisition, obtaining diving competition athlete performance data from the FINA official website;

[0011] S2: Data organization. For b judges and k athletes, and the athlete's performance data for m rounds of diving competition for each athlete, organize the athlete's performance data into k×m rows and b columns, and save it as an Excel file.

[0012] S3: Data Quality Evaluation Indicators Calculation. Kendall's coefficient of concordance and Kendall's correlation coefficient are used as data quality evaluation indicators for diving athletes' performances. Evaluation is conducted from four aspects: the overall diving competition, each round of diving, each athlete, and each type of diving maneuver. Each aspect is further evaluated from three dimensions: overall performance data quality, pairwise data quality between multiple measurements, and data quality between multiple measurements and the final measured result. Specifically, Kendall's coefficient of concordance is used to evaluate the overall performance data quality of the entire diving competition; Kendall's correlation coefficient is used to evaluate the pairwise data quality between multiple measurements of the entire diving competition; Kendall's correlation coefficient is used to evaluate the data quality between multiple measurements of the entire diving competition and the final measured result; and Kendall's coefficient of concordance is used to evaluate the overall performance data quality of each round of diving. The evaluation methods included: using Kendall correlation coefficients to assess the data quality between pairs of multiple measurements in each round of diving competition; using Kendall correlation coefficients to assess the data quality between multiple measurements and the final measured result in each round of diving competition; using Kendall concordance coefficients to assess the data quality of the overall performance of each athlete; using Kendall correlation coefficients to assess the data quality between pairs of multiple measurements for each athlete; using Kendall correlation coefficients to assess the data quality between multiple measurements and the final measured result for each athlete; using Kendall concordance coefficients to assess the data quality of the overall performance of each type of diving movement; using Kendall correlation coefficients to assess the data quality between pairs of multiple measurements for each type of diving movement; and using Kendall correlation coefficients to assess the data quality between multiple measurements and the final measured result for each type of diving movement. Kendall's covariance coefficient is calculated as shown in equation (1): In the formula, b is the rank determined by the average rank rule; k is the number of referees; t is the number of athletes; h g is the length of the h-th node group; i Let be the number of clusters in the k samples of the i-th variable; Among them, rank refers to the position or order of each specific data in the entire data arrangement (starting from the smallest number) after the data is arranged in order of size. If there are equal samples, the same samples are put together and called a knot. The number of samples in the knot is called the length of the knot. When the number of knots is greater than 1, the rank of the sample is the average of the ranks of the same samples. This method is called the average rank method. Kendall correlation coefficient is calculated as shown in equation (2): In the formula, t x It is the length of each node in the variable X; t y It is the length of each node in the Y variable;

[0013] S4: Data quality assessment. Data quality is evaluated based on the Kendall coefficient of concordance and the Kendall correlation coefficient. The Kendall coefficient of concordance ranges from [0,1], with values ​​closer to 0 indicating lower data quality and values ​​closer to 1 indicating higher data quality. The Kendall correlation coefficient ranges from [-1,1], with values ​​closer to -1 indicating lower data quality and values ​​closer to 1 indicating higher data quality.

[0014] Preferably, in step 3, the data quality evaluation of the overall performance of the entire diving competition is based on b judges as b variables, the i-th variable is the i-th score of m rounds of diving by k athletes, a total of k×m data, and the value of i is in the range of [1, b]. The Kendall coefficient of concordance of the b variables is calculated according to formula (1) and used as the data quality evaluation index of the overall performance of the entire diving competition. The closer the Kendall coefficient of concordance is to 0, the lower the data quality of the overall performance of the entire diving competition. The closer the Kendall coefficient of concordance is to 1, the higher the data quality of the overall performance of the entire diving competition.

[0015] Preferably, in step 3, the data quality evaluation of the pairwise measurements of multiple diving results throughout the entire diving competition is performed using b judges as b variables, with the i-th variable being the i-th result of m dives by k athletes, resulting in a total of k×m data points, and the value of i ranging from [1, b]. The Kendall correlation coefficient between the pairwise measurements of the b variables is calculated according to formula (2) and used as the data quality evaluation index for the pairwise measurements of multiple diving results throughout the entire diving competition. The closer the Kendall correlation coefficient is to -1, the lower the data quality of the pairwise measurements of multiple diving results throughout the entire diving competition; the closer the Kendall correlation coefficient is to 1, the higher the data quality of the pairwise measurements of multiple diving results throughout the entire diving competition.

[0016] Preferably, in step 3, the data quality evaluation between multiple measurements and the final measured result of the entire diving competition is carried out by taking b judges as b variables, the i-th variable as the i-th result of m rounds of diving by k athletes, a total of k×m data, and the value of i in the range of [1, b]. The overall performance of the athletes is taken as the b+1-th variable. The Kendall correlation coefficient between the b variables and the b+1-th variable is calculated according to formula (2) and used as the data quality evaluation index between multiple measurements and the final measured result of the entire diving competition. The closer the Kendall correlation coefficient is to -1, the lower the data quality between multiple measurements and the final measured result of the entire diving competition. The closer the Kendall correlation coefficient is to 1, the higher the data quality between multiple measurements and the final measured result of the entire diving competition.

[0017] Preferably, in step 3, the data quality evaluation for the overall performance of each round of diving competition uses b judges as b variables, the i-th variable is the i-th score of the j-th round of diving by k athletes, a total of k data points, i is in the range of [1, b], and j is in the range of [1, m]. The Kendall coefficient of concordance of the b variables is calculated according to formula (1) m times, and is used as the data quality evaluation index for the overall performance of each round of diving competition. The closer the Kendall coefficient of concordance is to 0, the lower the data quality of the overall performance of each round of diving competition; the closer the Kendall coefficient of concordance is to 1, the higher the data quality of the overall performance of each round of diving competition.

[0018] Preferably, in step 3, the data quality evaluation for pairwise measurements of multiple divergence results in each round of diving competition uses b judges as b variables, the i-th variable is the i-th result of the j-th dive of k athletes, a total of k data points, i ranges from [1, b], and j ranges from [1, m]. The Kendall correlation coefficient between each pair of b variables is calculated according to formula (2), and is calculated m times in total, serving as the data quality evaluation index for pairwise measurements of multiple divergence results in each round of diving competition. The closer the Kendall correlation coefficient is to -1, the lower the data quality between pairwise measurements of multiple divergence results in each round of diving competition; the closer the Kendall correlation coefficient is to 1, the higher the data quality between pairwise measurements of multiple divergence results in each round of diving competition.

[0019] Preferably, in step 3, the data quality evaluation between multiple measured scores and the final measured score for each round of diving competition is performed using b judges as b variables, the i-th variable being the i-th score of k athletes in the j-th round of diving, for a total of k data points, i ranging from [1, b] and j ranging from [1, m], with the overall athlete scores as the b+1-th variable; the Kendall correlation coefficient between the b variables and the b+1-th variable is calculated according to formula (2) m times, serving as the data quality evaluation index between multiple measured scores and the final measured score for each round of diving competition; the closer the Kendall correlation coefficient is to -1, the lower the data quality between multiple measured scores and the final measured score for each round of diving competition, and the closer the Kendall correlation coefficient is to 1, the higher the data quality between multiple measured scores and the final measured score for each round of diving competition.

[0020] Preferably, in step 3, the data quality evaluation for the overall performance of each athlete is based on b judges as b variables, with the i-th variable being the i-th score of the p-th athlete in m rounds of diving, for a total of m data points, where i ranges from [1, b] and p ranges from [1, k]. The Kendall coefficients of the b variables are calculated according to formula (1) and are calculated k times, serving as the data quality evaluation index for the overall performance of each athlete. The closer the Kendall coefficient is to 0, the lower the data quality of the overall performance of each athlete; the closer the Kendall coefficient is to 1, the higher the data quality of the overall performance of each athlete.

[0021] Preferably, in step 3, the data quality evaluation for pairwise measurements of each athlete's multiple measurements uses b judges as b variables, with the i-th variable being the i-th result of the p-th athlete's m-th round of diving, for a total of m data points. The value of i ranges from [1, b], and the value of p ranges from [1, k]. The Kendall correlation coefficient between each pair of the b variables is calculated according to formula (2), and is calculated k times in total, serving as the data quality evaluation index for pairwise measurements of each athlete's multiple measurements. The closer the Kendall correlation coefficient is to -1, the lower the data quality between pairwise measurements of each athlete's multiple measurements; the closer the Kendall correlation coefficient is to 1, the higher the data quality between pairwise measurements of each athlete's multiple measurements.

[0022] Preferably, in step 3, the data quality evaluation between the multiple measurements and the final result for each athlete is performed using b judges as b variables. The i-th variable is the i-th result of the m-th dive of the p-th athlete, with a total of m data points. The value of i ranges from [1, b], and the value of p ranges from [1, k]. The judges' scores are used as the b+1-th variable. The Kendall correlation coefficient between the b variables and the b+1-th variable is calculated according to formula (2) and is calculated k times. This coefficient serves as the data quality evaluation index between the multiple measurements and the final result for each athlete. The closer the Kendall correlation coefficient is to -1, the lower the data quality between the multiple measurements and the final result for each athlete. The closer the Kendall correlation coefficient is to 1, the higher the data quality between the multiple measurements and the final result for each athlete.

[0023] Preferably, in step 3, the data quality evaluation for the overall performance of each type of diving action uses b judges as b variables, the i-th variable is the i-th performance of w athletes selected for the q-th type of diving action, and there are a total of w data points, where i ranges from [1, b], q ranges from [1, n], and w ranges from [1, k]. The Kendall coefficients of the b variables are calculated according to formula (1) and are calculated n times, serving as the data quality evaluation index for the overall performance of each type of diving action. The closer the Kendall coefficient is to 0, the lower the data quality of the overall performance of each type of diving action; the closer the Kendall coefficient is to 1, the higher the data quality of the overall performance of each type of diving action.

[0024] Preferably, in step 3, the data quality evaluation for pairwise measurements of multiple diving performances for each type of diving action is performed using b judges as b variables. The i-th variable is the i-th performance of w athletes selected by the i-th judge for the q-th diving action, with a total of w data points. The range of i is [1, b], the range of q is [1, n], and the range of w is [1, k]. The Kendall correlation coefficient between each pair of the b variables is calculated according to formula (2) and is calculated n times. This is used as the data quality evaluation index for pairwise measurements of multiple diving performances for each type of diving action. The closer the Kendall correlation coefficient is to -1, the lower the data quality between pairwise measurements of multiple diving performances for each type of diving action. The closer the Kendall correlation coefficient is to 1, the higher the data quality between pairwise measurements of multiple diving performances for each type of diving action.

[0025] Preferably, in step 3, the data quality evaluation between multiple measurements and the final measured result for each type of diving action is performed using b judges as b variables. The i-th variable is the i-th result of w athletes who selected the q-th type of diving action, with a total of w data points. The range of i is [1, b], the range of q is [1, n], and the range of w is [1, k]. The judges' scores are used as the b+1-th variable. The Kendall correlation coefficient between the b variables and the b+1-th variable is calculated according to formula (2) and is calculated n times in total. This serves as the data quality evaluation index between multiple measurements and the final measured result for each type of diving action. The closer the Kendall correlation coefficient is to -1, the lower the data quality between multiple measurements and the final measured result for each type of diving action; the closer the Kendall correlation coefficient is to 1, the higher the data quality between multiple measurements and the final measured result for each type of diving action. Attached Figure Description

[0026] Figure 1 This is a diagram illustrating the system structure for evaluating the quality of athlete performance data in diving competitions.

[0027] Figure 2 This is a flowchart for evaluating the quality of athletes' performance data in diving competitions.

[0028] Figure 3 This is a data structure diagram for this diving competition. Detailed Implementation

[0029] 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 examples are only some embodiments of the present invention, and 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.

[0030] Figure 1 This diagram illustrates the system architecture for evaluating the quality of diving athlete performance data. It includes a diving athlete performance data quality evaluation system, a computer, and software for evaluating the diving athlete performance data quality based on Matlab R2020b programming. The diving athlete performance data quality evaluation system employs… Figure 2 The process shown uses a Lenovo YangTian S540-14-IWL computer. Athlete performance data obtained from FINA is saved as an Excel spreadsheet, input into the computer, and the computer uses diving performance data quality evaluation software to evaluate the quality of the diving performance data and outputs an athlete performance data quality evaluation report. For example... Figure 2As shown, the specific implementation of the evaluation method for diving athletes' performance data is as follows:

[0031] S1: Data Acquisition: Obtain athlete performance data for the 2022 Diving World Cup - Women's 3m Springboard Diving competition from the FINA official website;

[0032] S2: Data organization. For 7 judges and 12 athletes, each athlete's diving performance data (5 rounds) is organized into 12×5 rows and 7 columns, with the 12 athletes' 5 rounds of diving as the rows and the 7 judges' 5 rounds as the columns. Figure 3 As shown, save them as an Excel file;

[0033] S3: Data Quality Evaluation Indicators Calculation. Kendall's coefficient of concordance and Kendall's correlation coefficient are used as data quality evaluation indicators for diving athletes' performances. Evaluation is conducted from four aspects: the overall diving competition, each round of diving, each athlete, and each type of diving maneuver. Each aspect is further evaluated from three dimensions: overall performance data quality, pairwise data quality between multiple measurements, and data quality between multiple measurements and the final measured result. Specifically, Kendall's coefficient of concordance is used to evaluate the overall performance data quality of the entire diving competition; Kendall's correlation coefficient is used to evaluate the pairwise data quality between multiple measurements of the entire diving competition; Kendall's correlation coefficient is used to evaluate the data quality between multiple measurements of the entire diving competition and the final measured result; and Kendall's coefficient of concordance is used to evaluate the overall performance data quality of each round of diving. The evaluation methods included: using Kendall correlation coefficients to assess the data quality between pairs of multiple measurements in each round of diving competition; using Kendall correlation coefficients to assess the data quality between multiple measurements and the final measured result in each round of diving competition; using Kendall concordance coefficients to assess the data quality of the overall performance of each athlete; using Kendall correlation coefficients to assess the data quality between pairs of multiple measurements for each athlete; using Kendall correlation coefficients to assess the data quality between multiple measurements and the final measured result for each athlete; using Kendall concordance coefficients to assess the data quality of the overall performance of each type of diving movement; using Kendall correlation coefficients to assess the data quality between pairs of multiple measurements for each type of diving movement; and using Kendall correlation coefficients to assess the data quality between multiple measurements and the final measured result for each type of diving movement. Kendall's covariance coefficient is calculated as shown in equation (1): In the formula, b is the rank determined by the average rank rule; k is the number of referees; t is the number of athletes; h g is the length of the h-th node group; i Let be the number of clusters in the k samples of the i-th variable; Among them, rank refers to the position or order of each specific data in the entire data arrangement (starting from the smallest number) after the data is arranged in order of size. If there are equal samples, the same samples are put together and called a knot. The number of samples in the knot is called the length of the knot. When the number of knots is greater than 1, the rank of the sample is the average of the ranks of the same samples. This method is called the average rank method. Kendall correlation coefficient is calculated as shown in equation (2): In the formula, t x It is the length of each node in the variable X; t y It is the length of each node in the Y variable;

[0034] S4: Data quality assessment. Data quality is evaluated based on the Kendall coefficient of concordance and the Kendall correlation coefficient. The Kendall coefficient of concordance ranges from [0,1], with values ​​closer to 0 indicating lower data quality and values ​​closer to 1 indicating higher data quality. The Kendall correlation coefficient ranges from [-1,1], with values ​​closer to -1 indicating lower data quality and values ​​closer to 1 indicating higher data quality.

[0035] Preferably, in step 3, the data quality evaluation of the overall performance of the entire diving competition uses 7 judges as 7 variables, and the i-th variable takes the i-th score of the 12 athletes in 5 rounds of diving, for a total of 12×5 data points, with i ranging from [1, 7]. The Kendall coefficient of concordance of the 7 variables is calculated according to formula (1) and used as the data quality evaluation index of the overall performance of the entire diving competition. The closer the Kendall coefficient of concordance is to 0, the lower the data quality of the overall performance of the entire diving competition; the closer the Kendall coefficient of concordance is to 1, the higher the data quality of the overall performance of the entire diving competition.

[0036] Preferably, in step 3, the data quality evaluation of the pairwise measurements of multiple diving results throughout the entire diving competition is conducted using 7 judges as 7 variables, with the i-th variable taking the i-th result of 5 rounds of diving by 12 athletes, resulting in a total of 12×5 data points, and the value of i ranging from [1, 7]. The Kendall correlation coefficient between the pairwise measurements of the 7 variables is calculated according to formula (2) and used as the data quality evaluation index for the pairwise measurements of multiple diving results throughout the entire diving competition. The closer the Kendall correlation coefficient is to -1, the lower the data quality of the pairwise measurements of multiple diving results throughout the entire diving competition; the closer the Kendall correlation coefficient is to 1, the higher the data quality of the pairwise measurements of multiple diving results throughout the entire diving competition.

[0037] Preferably, in step 3, the data quality evaluation between the multiple measurements and the final measured result of the entire diving competition is conducted using 7 judges as 7 variables, with the i-th variable taking the i-th result of the 5 rounds of diving by 12 athletes, for a total of 12×5 data points, and i taking the range of [1, 7]. The overall results of the athletes are taken as the 8th variable. The Kendall correlation coefficient between the 7 variables and the 8th variable is calculated according to formula (2) and used as the data quality evaluation index between the multiple measurements and the final measured result of the entire diving competition. The closer the Kendall correlation coefficient is to -1, the lower the data quality between the multiple measurements and the final measured result of the entire diving competition; the closer the Kendall correlation coefficient is to 1, the higher the data quality between the multiple measurements and the final measured result of the entire diving competition.

[0038] Preferably, in step 3, the data quality evaluation for the overall performance of each round of diving competition uses 7 judges as 7 variables. The i-th variable takes the i-th score of the j-th round of diving by 12 athletes, for a total of 12 data points. The value range of i is [1, 7], and the value range of j is [1, 5]. The Kendall coefficient of concordance of b variables is calculated according to formula (1) and calculated 5 times in total, which serves as the data quality evaluation index for the overall performance of each round of diving competition. The closer the Kendall coefficient of concordance is to 0, the lower the data quality of the overall performance of each round of diving competition. The closer the Kendall coefficient of concordance is to 1, the higher the data quality of the overall performance of each round of diving competition.

[0039] Preferably, in step 3, the data quality evaluation for pairwise measurements of multiple dive results in each round of diving competition uses 7 judges as 7 variables. The i-th variable takes the i-th result of the j-th dive by 12 athletes, for a total of 12 data points. The value range of i is [1, 7], and the value range of j is [1, 5]. The Kendall correlation coefficient between each pairwise of the 7 variables is calculated according to formula (2), and is calculated 5 times in total. This is used as the data quality evaluation index for pairwise measurements of multiple dive results in each round of diving competition. The closer the Kendall correlation coefficient is to -1, the lower the data quality between pairwise measurements of multiple dive results in each round of diving competition. The closer the Kendall correlation coefficient is to 1, the higher the data quality between pairwise measurements of multiple dive results in each round of diving competition.

[0040] Preferably, in step 3, the data quality evaluation between the multiple measured scores and the final measured score in each round of diving competition is performed using 7 judges as 7 variables. The i-th variable takes the i-th score of the j-th round of diving by 12 athletes, for a total of 12 data points. The value range of i is [1, 7], and the value range of j is [1, 5]. The overall performance of the athletes is taken as the 8th variable. The Kendall correlation coefficient between the 7 variables and the 8th variable is calculated according to formula (2) five times, which serves as the data quality evaluation index between the multiple measured scores and the final measured score in each round of diving competition. The closer the Kendall correlation coefficient is to -1, the lower the data quality between the multiple measured scores and the final measured score in each round of diving competition. The closer the Kendall correlation coefficient is to 1, the higher the data quality between the multiple measured scores and the final measured score in each round of diving competition.

[0041] Preferably, in step 3, the data quality evaluation for the overall performance of each athlete is based on 7 judges as 7 variables. The i-th variable is the i-th score of the p-th athlete in 5 rounds of diving, with a total of 5 data points. The value of i is in the range of [1, 7] and the value of p is in the range of [1, 12]. The Kendall coefficient of concordance of the 7 variables is calculated according to formula (1) and is calculated 12 times in total, which serves as the data quality evaluation index for the overall performance of each athlete. The closer the Kendall coefficient of concordance is to 0, the lower the data quality of the overall performance of each athlete. The closer the Kendall coefficient of concordance is to 1, the higher the data quality of the overall performance of each athlete.

[0042] Preferably, in step 3, the data quality evaluation for pairwise measurements of each athlete's multiple measurements uses 7 judges as 7 variables. The i-th variable is the i-th result of the p-th athlete's 5 rounds of diving, with a total of 5 data points. The value of i ranges from [1, 7], and the value of p ranges from [1, 12]. The Kendall correlation coefficient between each pairwise of the 7 variables is calculated according to formula (2) and is calculated 12 times in total. This is used as the data quality evaluation index for pairwise measurements of each athlete's multiple measurements. The closer the Kendall correlation coefficient is to -1, the lower the data quality between pairwise measurements of each athlete's multiple measurements. The closer the Kendall correlation coefficient is to 1, the higher the data quality between pairwise measurements of each athlete's multiple measurements.

[0043] Preferably, in step 3, the data quality evaluation between the multiple measurements and the final measured score of each athlete is conducted using 7 judges as 7 variables. The i-th variable is the i-th score of the p-th athlete in 5 rounds of diving, with a total of 5 data points. The value of i ranges from [1, 7], and the value of p ranges from [1, 12]. The judges' scores are the 8th variable. The Kendall correlation coefficient between the 7 variables and the 8th variable is calculated according to formula (2) and is calculated 12 times. This is used as the data quality evaluation index between the multiple measurements and the final measured score of each athlete. The closer the Kendall correlation coefficient is to -1, the lower the data quality between the multiple measurements and the final measured score of each athlete. The closer the Kendall correlation coefficient is to 1, the higher the data quality between the multiple measurements and the final measured score of each athlete.

[0044] Preferably, in step 3, the data quality evaluation for the overall performance of each type of diving action uses 7 judges as 7 variables. The i-th variable takes the i-th score of w athletes selected for the q-th type of diving action, for a total of w data points. The value of i ranges from [1, 7], the value of q ranges from [1, 5], and the value of w ranges from [1, 12]. The Kendall coefficient of concordance of the 7 variables is calculated according to formula (1) and is calculated 5 times in total, serving as the data quality evaluation index for the overall performance of each type of diving action. The closer the Kendall coefficient of concordance is to 0, the lower the data quality of the overall performance of each type of diving action; the closer the Kendall coefficient of concordance is to 1, the higher the data quality of the overall performance of each type of diving action.

[0045] Preferably, in step 3, the data quality evaluation for pairwise measurements of multiple diving performances for each type of diving action uses 7 judges as 7 variables. The i-th variable is the i-th performance of w athletes selected by the i-th judge for the q-th diving action, with a total of w data points. The value range of i is [1, 7], the value range of q is [1, 5], and the value range of w is [1, 12]. The Kendall correlation coefficient between each pair of the 7 variables is calculated according to formula (2) and is calculated 5 times. This is used as the data quality evaluation index for pairwise measurements of multiple diving performances for each type of diving action. The closer the Kendall correlation coefficient is to -1, the lower the data quality between pairwise measurements of multiple diving performances for each type of diving action. The closer the Kendall correlation coefficient is to 1, the higher the data quality between pairwise measurements of multiple diving performances for each type of diving action.

[0046] Preferably, in step 3, the data quality evaluation between multiple measurements and the final measured result for each type of diving action uses 7 judges as 7 variables. The i-th variable is the i-th result of w athletes who selected the q-th type of diving action, with a total of w data points. The value of i ranges from [1, 7], the value of q ranges from [1, 5], and the value of w ranges from [1, 12]. The judges' score is the 8th variable. The Kendall correlation coefficient between the 7 variables and the 5th variable is calculated according to formula (2) 8 times, which serves as the data quality evaluation index between multiple measurements and the final measured result for each type of diving action. The closer the Kendall correlation coefficient is to -1, the lower the data quality between multiple measurements and the final measured result for each type of diving action. The closer the Kendall correlation coefficient is to 1, the higher the data quality between multiple measurements and the final measured result for each type of diving action.

[0047] The following are the results of the quality evaluation of athlete performance data for the 2022 Diving World Cup - Women's 3m Springboard Diving Competition:

[0048] ① Regarding the data quality of the overall match results, Kendall's coefficient of concordance (W*) is 0.927, indicating that the overall data quality of the match results is relatively high. Looking at the pairwise data quality between multiple measurements, Kendall's correlation coefficients are all above 0.7, indicating high data quality across multiple measurements. However, regarding the data quality between multiple measurements and the final result, the fifth measurement has the worst data quality compared to the other six measurements, with a Kendall correlation coefficient of 0.8314, while the fourth measurement has the best data quality, with a Kendall correlation coefficient of 0.8814.

[0049] ② Regarding the data quality of each round of diving competition results, from the perspective of the overall data quality, the Kendall correlation coefficient (W*) calculation results are all greater than 0.85, which indicates that the overall data quality of each round of diving competition results is relatively high. Among them, compared with the data quality of the first four rounds, the data quality of the fifth round is the worst, and the Kendall correlation coefficient calculation result is 0.857.

[0050] Looking at the pairwise data quality of multiple measurements, in the first round of diving competition, the Kendall correlation coefficients between the fifth and first measurements, and between the second and sixth measurements, were less than 0.7. In the fourth round of diving competition, the Kendall correlation coefficient between the fifth and sixth measurements was also less than 0.7. In the fifth round of diving competition, the Kendall correlation coefficients between the fifth and first, third, and sixth measurements were all less than 0.7. These results further reflect that the data quality of the fifth measurement was the worst compared to the other six measurements.

[0051] Looking at the data quality between the multiple measurements and the final result, the fifth measurement was the worst among the first, fourth, and fifth rounds of diving, indicating that its data quality was inferior to the other six measurements. Furthermore, the second measurement had the worst data quality in the second and third rounds, but the best in the fourth round, reflecting that while its data quality was relatively good among the seven measurements, it was inconsistent.

[0052] ③ Regarding the data quality of each athlete's performance, considering the overall data quality, the pairwise data quality between multiple measurements, and the data quality between multiple measurements and the final measured result, the data quality of the sixth athlete's performance was the worst compared to the other eleven athletes, with a Kendall correlation coefficient of 0.515. Furthermore, the data quality of the highest and lowest ranked athletes was relatively high, with Kendall correlation coefficients mostly above 0.85. Conversely, the data quality of the middle-ranked athletes was relatively low, with Kendall correlation coefficients mostly below 0.85.

[0053] ④ Regarding the data quality of each type of diving performance, from the perspective of overall data quality, the 5152B dive has the worst overall data quality compared to the other four dives. Since most athletes chose the 5152B dive in the fifth round of the competition, the overall data quality of the fifth round was worse than that of the first four rounds, which is consistent with the analysis results from the perspective of each round of the competition, further proving the effectiveness of the present invention.

[0054] Looking at the data quality between pairs of multiple measurement scores, the Kendall correlation coefficients between the fifth measurement score and the first, second, third, fourth, and sixth measurement scores are all less than 0.7, which again reflects that the data quality of the fifth measurement score is the worst compared to the other six measurement scores.

[0055] From the perspective of data quality between multiple measurements and the final result, for diving move 107B, the third measurement had the worst data quality compared to the other six measurements, with a Kendall correlation coefficient of 0.8343; for diving move 205B, the first measurement had the worst data quality compared to the other six measurements, with a Kendall correlation coefficient of 0.8668; for diving move 305B, the first measurement had the worst data quality compared to the other six measurements, with a Kendall correlation coefficient of 0.8353; for diving move 405B, the first and sixth measurements had the worst data quality compared to the other five measurements, with a Kendall correlation coefficient of 0.8485; and for diving move 5152B, the fifth measurement had the worst data quality compared to the other six measurements, with a Kendall correlation coefficient of 0.7267.

[0056] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for evaluating the quality of diving competition athlete performance data, characterized in that, Includes the following steps: S1: Data acquisition, obtaining diving competition athlete performance data from the FINA official website; S2: Data organization. For b judges and k athletes, and the athlete's performance data for each athlete in m rounds of diving competition, organize the athlete's performance data into k×m rows and b columns, and save it as an Excel file. S3: Data Quality Evaluation Indicators Calculation. Kendall's coefficient of concordance and Kendall's correlation coefficient are used as data quality evaluation indicators for diving athletes' performances. Evaluation is conducted from four aspects: the overall diving competition, each round of diving, each athlete, and each type of diving maneuver. Each aspect is further evaluated from three dimensions: overall performance data quality, pairwise data quality between multiple measurements, and data quality between multiple measurements and the final measured result. Specifically, Kendall's coefficient of concordance is used to evaluate the overall performance data quality of the entire diving competition; Kendall's correlation coefficient is used to evaluate the pairwise data quality between multiple measurements of the entire diving competition; Kendall's correlation coefficient is used to evaluate the data quality between multiple measurements of the entire diving competition and the final measured result; and Kendall's coefficient of concordance is used to evaluate the overall performance data quality of each round of diving. The evaluation methods included: using Kendall correlation coefficients to assess the data quality between pairs of multiple measurements in each round of diving competition; using Kendall correlation coefficients to assess the data quality between multiple measurements and the final measured result in each round of diving competition; using Kendall concordance coefficients to assess the data quality of the overall performance of each athlete; using Kendall correlation coefficients to assess the data quality between pairs of multiple measurements for each athlete; using Kendall correlation coefficients to assess the data quality between multiple measurements and the final measured result for each athlete; using Kendall concordance coefficients to assess the data quality of the overall performance of each type of diving movement; using Kendall correlation coefficients to assess the data quality between pairs of multiple measurements for each type of diving movement; and using Kendall correlation coefficients to assess the data quality between multiple measurements and the final measured result for each type of diving movement. Kendall's covariance coefficient is calculated as shown in equation (1): In the formula, b is the rank determined by the average rank rule; k is the number of referees; t is the number of athletes; h g is the length of the h-th node group; i Let be the number of clusters in the k samples of the i-th variable; Among them, rank refers to the position or order of each specific data in the entire data arrangement (starting from the smallest number) after the data is arranged in order of size. If there are equal samples, the same samples are put together and called a knot. The number of samples in the knot is called the length of the knot. When the number of knots is greater than 1, the rank of the sample is the average of the ranks of the same samples. This method is called the average rank method. Kendall correlation coefficient is calculated as shown in equation (2): In the formula, t x It is the length of each node in the variable X; t y It is the length of each node in the Y variable; S4: Data quality assessment. Data quality is evaluated based on the Kendall coefficient of concordance and the Kendall correlation coefficient. The Kendall coefficient of concordance ranges from [0,1], with values ​​closer to 0 indicating lower data quality and values ​​closer to 1 indicating higher data quality. The Kendall correlation coefficient ranges from [-1,1], with values ​​closer to -1 indicating lower data quality and values ​​closer to 1 indicating higher data quality.

2. The method for evaluating the quality of diving competition athlete performance data according to claim 1, characterized in that: To evaluate the overall data quality of the diving competition, b judges are used as b variables, and the i-th variable is the i-th score of m dives by k athletes, for a total of k×m data points, with i ranging from [1, b]. The Kendall coefficients of the b variables are calculated according to formula (1) and used as the data quality evaluation index for the overall diving competition. The closer the Kendall coefficient is to 0, the lower the data quality of the overall diving competition; the closer the Kendall coefficient is to 1, the higher the data quality of the overall diving competition.

3. The method for evaluating the quality of diving competition athlete performance data according to claim 1, characterized in that: To evaluate the data quality between pairs of multiple measurements in the entire diving competition, b judges are used as b variables, and the i-th variable is the i-th result of m dives by k athletes, for a total of k×m data points, with i ranging from [1, b]. The Kendall correlation coefficient between each pair of the b variables is calculated according to formula (2) and used as the data quality evaluation index between pairs of multiple measurements in the entire diving competition. The closer the Kendall correlation coefficient is to -1, the lower the data quality between pairs of multiple measurements in the entire diving competition; the closer the Kendall correlation coefficient is to 1, the higher the data quality between pairs of multiple measurements in the entire diving competition.

4. The method for evaluating the quality of diving competition athlete performance data according to claim 1, characterized in that: To evaluate the data quality between multiple measurements and the final result of the entire diving competition, b judges are used as b variables, the i-th variable is the i-th result of m rounds of diving by k athletes, a total of k×m data, the value of i is [1, b], and the overall performance of the athletes is used as the b+1-th variable. According to formula (2), calculate the Kendall correlation coefficient between the b variables and the b+1th variable respectively, and use it as the data quality evaluation index between the multiple measurements and the final measurement of the whole diving competition. The closer the Kendall correlation coefficient is to -1, the lower the data quality between the multiple measurements and the final measurement of the whole diving competition. The closer the Kendall correlation coefficient is to 1, the higher the data quality between the multiple measurements and the final measurement of the whole diving competition.

5. The method for evaluating the quality of diving competition athlete performance data according to claim 1, characterized in that: For the data quality evaluation of the overall results of each round of diving competition, b judges are used as b variables, the i-th variable is the i-th result of the j-th round of diving of k athletes, and there are a total of k data, i is in the range of [1, b], and j is in the range of [1, m]. The Kendall coefficients of b variables are calculated according to formula (1) and calculated m times in total. They serve as the data quality evaluation index for the overall performance of each round of diving. The closer the Kendall coefficient is to 0, the lower the data quality of the overall performance of each round of diving competition. The closer the Kendall coefficient is to 1, the higher the data quality of the overall performance of each round of diving competition.

6. The method for evaluating the quality of diving competition athlete performance data according to claim 1, characterized in that: For the data quality evaluation of the pairwise measurements of the results in each round of diving competition, b judges are used as b variables, and the i-th variable takes the i-th result of the j-th round of diving of k athletes, for a total of k data, i takes the value range of [1, b], and j takes the value range of [1, m]. The Kendall correlation coefficients between each pair of b variables are calculated according to formula (2), and the calculation is performed m times. This serves as an evaluation index of the data quality between each pair of multiple measurements in each round of diving competition. The closer the Kendall correlation coefficient is to -1, the lower the data quality between each pair of multiple measurements in each round of diving competition. The closer the Kendall correlation coefficient is to 1, the higher the data quality between each pair of multiple measurements in each round of diving competition.

7. The method for evaluating the quality of diving competition athlete performance data according to claim 1, characterized in that: For the data quality evaluation between multiple measurements and the final result in each round of diving competition, b judges are used as b variables, the i-th variable is the i-th result of the j-th round of diving of k athletes, a total of k data, the value of i is [1, b], the value of j is [1, m], and the overall performance of the athletes is used as the b+1-th variable. According to formula (2), the Kendall correlation coefficient between b variables and the (b+1)th variable is calculated respectively, and a total of m calculations are performed. This is used as the data quality evaluation index between the multiple measurements and the final measurement results in each round of diving competition. The closer the Kendall correlation coefficient is to -1, the lower the data quality between the multiple measurements and the final measurement results in each round of diving competition. The closer the Kendall correlation coefficient is to 1, the higher the data quality between the multiple measurements and the final measurement results in each round of diving competition.

8. The method for evaluating the quality of diving competition athlete performance data according to claim 1, characterized in that: For the data quality evaluation of each athlete's overall performance, b judges are used as b variables, and the i-th variable is the i-th score of the p-th athlete in m rounds of diving, with a total of m data points. The value of i is in the range of [1, b] and the value of p is in the range of [1, k]. The Kendall coefficient of concordance of the b variables is calculated according to formula (1) and is calculated k times in total, which serves as the data quality evaluation index for each athlete's overall performance. The closer the Kendall coefficient of concordance is to 0, the lower the data quality of each athlete's overall performance. The closer the Kendall coefficient of concordance is to 1, the higher the data quality of each athlete's overall performance.

9. The method for evaluating the quality of diving competition athlete performance data according to claim 1, characterized in that: For the data quality evaluation of each athlete's multiple measurements, b judges are used as b variables. The i-th variable takes the i-th score of the p-th athlete in m rounds of diving. There are a total of m data points, with i ranging from [1, b] and p ranging from [1, k]. Calculate the Kendall correlation coefficient between each pair of b variables according to formula (2), and calculate it k times in total. This serves as the data quality evaluation index between each pair of multiple measurements of each athlete. The closer the Kendall correlation coefficient is to -1, the lower the data quality between each pair of multiple measurements of each athlete. The closer the Kendall correlation coefficient is to 1, the higher the data quality between each pair of multiple measurements of each athlete.

10. The method for evaluating the quality of diving competition athlete performance data according to claim 1, characterized in that: For the data quality evaluation between the multiple measurements and the final result for each athlete, b judges are used as b variables. The i-th variable takes the i-th result of the m-th dive of the p-th athlete, with a total of m data points. The value of i is in the range of [1, b] and the value of p is in the range of [1, k]. The b+1-th variable is the score given by the judges. The Kendall correlation coefficient between the b variables and the (b+1)th variable is calculated according to formula (2) and is calculated k times in total. It serves as the data quality evaluation index between the multiple measurements and the final measurement of each athlete. The closer the Kendall correlation coefficient is to -1, the lower the data quality between the multiple measurements and the final measurement of each athlete. The closer the Kendall correlation coefficient is to 1, the higher the data quality between the multiple measurements and the final measurement of each athlete.

11. The method for evaluating the quality of diving competition athlete performance data according to claim 1, characterized in that: For the data quality evaluation of the overall performance of each type of diving action, b judges are used as b variables. The i-th variable takes the i-th score of w athletes who selected the q-th type of diving action. There are a total of w data points, with i ranging from [1, b], q ranging from [1, n], and w ranging from [1, k]. The Kendall coefficients of b variables are calculated according to formula (1) and calculated n times in total. They serve as the data quality evaluation index for the overall performance of each type of diving action. The closer the Kendall coefficient is to 0, the lower the data quality of the overall performance of each type of diving action. The closer the Kendall coefficient is to 1, the higher the data quality of the overall performance of each type of diving action.

12. The method for evaluating the quality of diving competition athlete performance data according to claim 1, characterized in that: For the data quality evaluation of multiple measurements of each type of diving action, b judges are used as b variables. The i-th variable is the i-th result of w athletes selected by the i-th judge for the q-th diving action. There are a total of w data points, with i ranging from [1, b], q ranging from [1, n], and w ranging from [1, k]. The Kendall correlation coefficients between each pair of b variables were calculated according to formula (2), and the calculation was performed n times in total. This was used as a data quality evaluation index for the pairwise measurements of the results of each type of diving action. The closer the Kendall correlation coefficient is to -1, the lower the data quality between pairs of multiple measurements of each type of diving action; the closer the Kendall correlation coefficient is to 1, the higher the data quality between pairs of multiple measurements of each type of diving action.

13. The method for evaluating the quality of diving competition athlete performance data according to claim 1, characterized in that: For the data quality evaluation between multiple measurements and the final measured result for each type of diving action, b judges are used as b variables. The i-th variable is the i-th result of w athletes who selected the q-th diving action. There are a total of w data points, with i ranging from [1, b], q ranging from [1, n], and w ranging from [1, k]. The judges' scores are used as the b+1-th variable. According to formula (2), calculate the Kendall correlation coefficient between b variables and the (b+1)th variable respectively, and calculate it n times in total. It is used as the data quality evaluation index between the multiple measurement results and the final measurement results of each type of diving action. The closer the Kendall correlation coefficient is to -1, the lower the data quality between the multiple measurement results and the final measurement results of each type of diving action. The closer the Kendall correlation coefficient is to 1, the higher the data quality between the multiple measurement results and the final measurement results of each type of diving action.

Images