Authors: Chenghao Ma¹

The correlation between weight divisions and methods used by winning mixed martial arts athletes

ABSTRACT

This study analyzes the correlation between weight divisions and the methods used by 174 top-ranking Ultimate Fighting Championship (UFC) elite athletes to victory, thus providing valuable information to help coaches and athletes formulate their training plans and competition strategies. It uses descriptive statistical analysis to present essential data in answer to a number of research questions. A Mann-Whitney U test was conducted to test the difference between male and female athletes within the same weight divisions, with a Kruskal-Wallis test revealing differences among weight divisions for each sex. Spearman’s correlation and Linear Regression tests were then used to analyze the relationship between these weight divisions and the methods used by winning athletes. The results indicated the following ratios for success: Knockout (KO)/Technical Knockout (TKO) (40.21%±22.27), Decision (36.78%±20.88), and Submission (23.01%±18.36). Spearman’s rho bivariate correlation test showed that the weight divisions for male athletes had a positive correlation with KO/TKO and a negative correlation with Decision and Submission regarding the methods leading to their success. However, there was no correlation between weight divisions for female athletes and their winning methods. Linear regression test results indicated linear correlations between the independent and dependent variables and created regression models for the correlation between the weight divisions of male athletes and their methods. The present study aims to provide coaches and athletes with valuable reference points concerning weight divisions and victory, thus enabling them to optimize training plans and competition strategies and change weight divisions to secure competitive advantages. Sports fans would also be able to make more logical predictions concerning the possible victory methods of their favorite athletes.

Key Words: combat sports; UFC; athletic performance; training plan; competition strategy

INTRODUCTION

Mixed Martial Arts (MMA) is a fast-growing sport worldwide. Mixed martial artists usually come from a variety of martial arts disciplines, such as boxing, karate, Brazilian Jiu-Jitsu, Muay Thai, kickboxing, and wrestling (16, 20), and athletes can use different techniques in MMA competitions, such as striking (using hands, elbows, knees, and feet) and grappling (takedown, chokes and joint locks) (4, 6, 15, 18). Most MMA competitions are held in octagonal cages, with the fights being divided into three or five rounds of five minutes with a one-minute interval to rest (9). The wide range of techniques allows athletes to show their unique fighting skills, making competitions entertaining spectacles (13).

From time immemorial, human beings have used fighting techniques to protect themselves from beasts and invaders to survive. One could say, therefore, that fighting has evolved along with human society, with different martial arts being formed in different regions and countries. The origin of MMA can be traced back to 648 BC when at the 33rd Olympic Games, Ancient Greek athletes competed in an arena with their bare hands in what was called Pankration (1). In Pankration, the rules made it difficult to win by mastering only one fighting technique, so the participants were usually proficient in multiple techniques, such as boxing, wrestling, and ground fighting.

Since the 1990s, along with the rapid development of MMA, organizations promoting it have also emerged worldwide. The Ultimate Fighting Championship (UFC) is, in fact, the most influential organization, with its first-ever event in the United States held in 1993 (11). Mixed martial arts is now becoming a mainstream sport, and the UFC has since become the leading organizer of MMA events (29). The UFC adopted the Unified Rules of Mixed Martial Arts in November 2000 to ensure the safety of athletes and fair competition (32). These rules are intended to provide a clear set of regulations governing professional MMA competitions consistent across different athletic commissions and other regulatory bodies. The framework for these rules was proposed and agreed upon by the various athletic commissions and adopted unanimously by the Association of Boxing Commissions (ABC) in July 2009 (32). The establishment and adoption of the Unified Rules of MMA mark the beginning of MMA as a sport accepted by both governing bodies and the public alike and have played an essential role in developing and promoting mixed martial arts. The commercial growth and international expansion have attracted athletes from other fighting categories to join the sport and compete in the UFC. A study on top-ranking UFC athletes is therefore thought to be both valuable and meaningful.

Mixed martial arts fights occur within specific weight divisions, with athletes winning a contest by Knockout, Technical Knockout, Submission, or the Referee’s Decision (21, 27). Divisions are based on body mass, and the athletes are then paired according to weight to prevent heavier athletes from scoring an obvious advantage, just as in other combat sports (8). The weight divisions stipulated by the UFC include eight weight classes for male athletes (Flyweight, Bantamweight, Featherweight, Lightweight, Welterweight, Middleweight, Lightweight Heavyweight, and Heavyweight) and three weight classes for female athletes (Strawweight, Flyweight, and Bantamweight), with the Flyweight division for females having been added (31). MMA athletes are grouped into different weight classes determined by their weight and measured around 24 to 32 hours before the competitions (7). Body mass manipulation through rapid weight gain and rapid weight loss is common among MMA athletes to ensure qualification for the weight class in which the athlete wants to compete (3, 5, 19). Therefore, weight classes are usually a matter of concern for coaches, athletes, and even MMA fans and are a prerequisite for fair competition (22).

UFC fights are governed by the Unified Rules of Mixed Martial Arts in which the fighter wins the fight in the following ways (2, 24, 30): First, for a victory to be secured through KO (Knockout), the referee stops the fight because the athlete cannot defend himself consciously because of striking techniques. The second is the TKO (Technical Knockout), where the referee, doctor, or the athlete’s corner stops the fight because the athlete cannot defend himself or herself if continuing the fight will put the athlete’s health at risk. The third is the Referee’s Decision, in which the referee scores the winner through a ten-point must system. The fourth is won by Submission, in which the athlete controls his opponent through submission techniques, causing that opponent to signal that he cannot continue the fight (10, 23). The diversity of winning styles makes the outcome of MMA fights unpredictable. The unique methods used to secure a win displayed by the different weight classes can provide coaches and athletes with a deeper understanding of their particular weight division and how a move to a new weight class might affect their methods. Thus, coaches and athletes can adjust training plans and competition strategies (25, 26).

The main research questions of this study are: What is the ratio of winning methods used by top-ranking UFC mixed martial artists regarding different gender groups and weight divisions? Secondly, is there a correlation between weight divisions and these winning methods? This study hypothesizes that as the weight divisions increase, there is a greater likelihood for the athletes to win by KO/TKO and a lower chance of winning by Decision and Submission. There is thus a significant positive correlation between weight division and KO/TKO and a significant negative correlation between weight division and Decision or Submission. This study analyzes the methods used to win and the correlation between these weight divisions and approaches to provide valuable information for martial arts coaches, athletes, and fans.

METHODS

Sample

Data concerning weight divisions and the methods used by male and female MMA athletes participating in the UFC to secure a victory (KO/TKO, Decision, and Submission) were collected from publicly available sources. Data were also collected from the official UFC ranking website (https://www.ufc.com/rankings) (31). A total of 174 athletes (male: 127; female: 47) who are champions and rank in the top 15 were involved in the sample data. MMA athletes were divided by gender and weight groups (Strawweight: up to and including 115 lbs, Flyweight: over 115 to 125 lbs, Bantamweight: over 125 to 135 lbs, Featherweight: over 135 to 145 lbs, Lightweight: over 145 to 155 lbs, Welterweight: over 165 to 170 lbs, Middleweight: over 175 to 185 lbs, Lightweight Heavyweight: over 195 to 205 lbs, and Heavyweight: over 225 to 265 lbs).

Archived databases from public access websites have in the past been employed for studies similar to the present research, without ethical issues in the investigation and interpretation of the data, as they were gathered in a secondary form and not developed experimentally (12, 14, 15, 17). The personal identification of individual data has also been avoided in this study, thus ensuring anonymity and confidentiality.

Statistical Analysis

First, a descriptive statistical analysis was conducted using SPSS 24 software. Data were provided as mean, minimum, maximum, and standard deviations. The Mann-Whitney U test was applied to test the difference between male and female athletes within the same weight divisions (Flyweight and Bantamweight), and a Kruskal-Wallis test showed the difference between weight divisions in each gender group. Spearman’s correlation and Linear Regression tests were then used to analyze the relationship between these weight divisions and the methods used by winning athletes within each gender group.

RESULTS

As shown in Table 1, male athletes who won were as follows: by KO/TKO: 43.75%±22.98; Decision: 33.35%±20.75, and Submission: 22.92%±18.70. Female athletes who won were as follows: by KO/TKO: 30.64%±17.04; Decision: 46.04%±18:43, and Submission: 23.23%±17.61. The total for athletes who won was as follows: by KO/TKO: 40.21%±22.27; Decision: 36.78%±20.88, and Submission: 23.01%±18.36. Table 2 shows a descriptive analysis of the methods used to win within different weight divisions for male and female athletes.

Table 1

Analysis of Methods Used to Win

Sex		KO/TKO (%)	Decision (%)	Submission (%)
Male	N	127	127	127
	Mean	43.75	33.35	22.92
	Std. Deviation	22.98	20.75	18.70
	Minimum	.00	.00	.00
	Maximum	93.00	95.00	81.00
Female	N	47	47	47
	Mean	30.64	46.04	23.23
	Std. Deviation	17.04	18.43	17.61
	Minimum	.00	17.00	.00
	Maximum	67.00	85.00	76.00
Total	N	174	174	174
	Mean	40.21	36.78	23.01
	Std. Deviation	22.27	20.88	18.36
	Minimum	.00	.00	.00
	Maximum	93.00	95.00	81.00

Table 2

Descriptive Analysis of the Methods Used to Win Within Different Weight Divisions (Percent)

Sex			N	Mean	Std. Deviation	Std. Error	Minimum	Maximum
Male	KO/TKO	Flyweight	16	32.25	20.03	5.01	4.00	81.00
		Bantamweight	16	33.00	18.21	4.55	2.00	73.00
		Featherweight	16	36.56	16.09	4.02	.00	64.00
		Lightweight	15	44.60	25.72	6.64	6.00	86.00
		Welterweight	16	33.50	18.06	4.52	.00	57.00
		Middleweight	16	48.50	23.50	5.87	8.00	85.00
		Light Heavyweight	16	55.00	19.19	4.80	19.00	86.00
		Heavyweight	16	66.63	20.76	5.19	32.00	93.00
		Total	127	43.75	22.98	2.04	.00	93.00
	Decision	Flyweight	16	30.31	20.45	5.11	6.00	88.00
		Bantamweight	16	49.19	18.79	4.70	20.00	86.00
		Featherweight	16	38.56	12.13	3.03	8.00	53.00
		Lightweight	15	31.47	24.83	6.41	.00	83.00
		Welterweight	16	45.31	24.61	6.15	.00	95.00
		Middleweight	16	31.81	18.11	4.53	.00	63.00
		Light Heavyweight	16	20.63	13.96	3.49	.00	44.00
		Heavyweight	16	19.38	12.16	3.04	.00	37.00
		Total	127	33.35	20.75	1.84	.00	95.00
	Submission	Flyweight	16	37.44	17.79	4.45	6.00	64.00
		Bantamweight	16	17.69	12.86	3.21	3.00	50.00
		Featherweight	16	24.81	19.26	4.82	.00	60.00
		Lightweight	15	23.73	17.69	4.57	4.00	64.00
		Welterweight	16	21.38	17.60	4.40	3.00	61.00
		Middleweight	16	19.75	16.29	4.07	.00	56.00
		Light Heavyweight	16	24.44	24.03	6.01	.00	81.00
		Heavyweight	16	14.19	16.87	4.22	.00	56.00
		Total	127	22.92	18.70	1.66	.00	81.00
Female	KO/TKO	Strawweight	16	27.69	16.26	4.07	.00	58.00
		Flyweight	15	31.53	15.86	4.09	11.00	55.00
		Bantamweight	16	32.75	19.39	4.85	.00	67.00
		Total	47	30.64	17.04	2.49	.00	67.00
	Decision	Strawweight	16	40.75	19.63	4.91	17.00	85.00
		Flyweight	15	50.67	18.12	4.68	27.00	83.00
		Bantamweight	16	47.00	17.25	4.31	22.00	81.00
		Total	47	46.04	18.43	2.69	17.00	85.00
	Submission	Strawweight	16	31.56	22.36	5.59	.00	76.00
		Flyweight	15	17.67	11.51	2.97	.00	38.00
		Bantamweight	16	20.13	14.56	3.64	.00	50.00
		Total	47	23.23	17.61	2.57	.00	76.00

Table 3 shows that U=490, W=1018, and Z=-0.083 for KO/TKO, corresponding to a significant probability of p=0.934>0.05, with the null hypothesis being accepted, showing that there is no significant difference between the KO/TKO rates for men and women within the same two weight divisions (Flyweight and Bantamweight). For the Decision method, U=371, W=867, and Z=-1.720 corresponded to a significant probability of p=0.043<0.05. The null hypothesis is therefore rejected, showing a significant difference in wins by Decision between men and women. For Submission, U=371, W=867, and Z=-1.720 corresponded to a significant probability of p=0.085>0.05, so the null hypothesis was accepted, showing no significant difference between men and women in this respect.

Table 3

The difference within Same Weight Divisions (Male and Female) in Terms of the Methods Used to Win

	KO/TKO	Decision	Submission
Mann-Whitney U	490.000	349.000	371.000
Wilcoxon W	1018.000	877.000	867.000
Z	-0.083	-2.022	-1.720
Asymp. Sig. (2-tailed)	0.934	0.043	0.085

Table 4 shows that the chi-square values for male athletes were: KO/TKO=30.117, Decision=33.209, and Submission=15.815. The corresponding probabilities were: p<0.01 (KO/TKO), p<0.01 (Decision), and p=0.027<0.05 (Submission), showing that there was a significant difference in the methods used by winning male athletes across weight classes. The chi-square values for female athletes were: KO/TKO=0.646, Decision=2.705, Submission=3.552. The corresponding probabilities were: p=0.724>0.05 (KO/TKO), p=0.259>0.05 (Decision), and p=0.169>0.05 (Submission), showing that there was no significant difference in the way female athletes won in different weight classes.

Table 4

Kruskal-Wallis Test to Determine the Approaches Used by Winning Athletes Across Weight Divisions

Sex		KO/TKO	Decision	Submission
Male	Chi-Square	30.117	33.209	15.815
	df	7	7	7
	Asymp.Sig.	0.000	0.000	0.027
Female	Chi-Square	0.646	2.705	3.552
	df	2	2	2
	Asymp. Sig.	0.724	0.259	0.169

 

As shown in Table 5, the results of the Spearman’s rho bivariate correlation test for male athletes were as follows: KO/TKO (r=0.446, p<0.01), Decision (r=-0.311, p<0.01), and Submission (r=-0.255, p=0.004<0.01). These results show that the weight class for male athletes had a positive correlation with KO/TKO and negative correlations with Decision and Submission. The Spearman’s rho bivariate correlation test for female athletes was as follows: KO/TKO (r=0.109, p=0.464>0.05), Decision (r=0.171, p=0.250>0.05), and Submission (r=-0.209, p=0.160>0.05). These results show no significant correlation between the weight classes of female athletes and the three approaches to winning.

Table 5

Spearman’s rho Correlation Test Between Weight Divisions and Winning Methods

Sex		KO/TKO	Decision	Submission
Male	Correlation Coefficient	0.446	-0.311	-0.255
	p (2-tailed)	0.000	0.000	0.004
Female	Correlation Coefficient	0.109	0.171	-0.209
	p (2-tailed)	0.464	0.250	0.160

Because the male athletes showed a correlation between weight class and the methods they used to win, a further linear regression analysis was performed based on the weight (Pounds) and the methods used by male athletes to win. Table 6 shows statistical tests for the linear regression model. The correlation coefficient was R=0.482, R²=0.232, adjusted R²=0.226, and the estimated standard error was 20.21593. The results show a linear correlation between the independent variable weight (Pounds) and the dependent variable KO/TKO.

Table 6

Statistical Tests of Linear Regression Model (KO/TKO)

Model	R	R Square	Adjusted R Square	Std. Error of the Estimate	Change Statistics					Durbin-Watson
					R Square Change	F Change	df1	df2	Sig. F Change
1	.482a	.232	.226	20.21593	.232	37.840	1	125	.000	1.990
a. Predictors: (Constant), Pounds
b. Dependent Variable: KOTKO

Table 7 shows the results of the variance test of the regression statistics, such as the variance, sum of squares, degrees of freedom, mean square value, the value of the statistic F, and its probability of significance for the model. The results show that F=37.840 with a probability of P<0.01, indicating that the regression effect is significant in this case.

Table 7

ANOVA Test (KO/TKO)

Model		Sum of Squares	df	Mean Square	F	Sig.
1	Regression	15464.466	1	15464.466	37.840	.000b
	Residual	51085.471	125	408.684
	Total	66549.937	126
a. Dependent Variable: KOTKO
b. Predictors: (Constant), Pounds

Table 8 shows the test of regression coefficients. The coefficient of the independent variable weight (Pounds) is 0.278, with a companion probability of p<0.01, indicating that the independent variable had a regression coefficient; therefore, there is a linear correlation between the independent and dependent variables. The regression equation is thus: Ŷ = -4.150 + 0.278X.

Table 8

Regression Coefficient Test (KO/TKO)

Model		Unstandardized Coefficients		Standardized Coefficients	t	Sig.
		B	Std. Error	Beta
1	(Constant)	-4.150	7.991		-.519	.604
	Pounds	.278	.045	.482	6.151	.000
a. Dependent Variable: KO/TKO

Table 9 shows the statistical test table of the linear regression model as follows: the correlation coefficient was R=0.329, R²=0.109, adjusted R²=0.101, and the estimated standard error was 19.66908. The results show a linear correlation between the independent and dependent variables.

Table 9

Statistical Tests of Linear Regression Model (Decision)

Model	R	R Square	Adjusted R Square	Std. Error of the Estimate	Change Statistics					Durbin-Watson
					R Square Change	F Change	df1	df2	Sig. F Change
1	.329a	.109	.101	19.66908	.109	15.224	1	125	.000	2.011
a. Predictors: (Constant), Pounds
b. Dependent Variable: Decision

Table 10 shows the results of the variance test of the regression statistics, such as the sum of squares, degrees of freedom, mean square value, and the value of the statistic F and its probability of significance for the model. The results show that F=15.224 with a probability of p<0.01, indicating that the regression effect is significant in this case.

Table 10

ANOVA Test (Decision)

Model		Sum of Squares	df	Mean Square	F	Sig.
1	Regression	5889.677	1	5889.677	15.224	.000b
	Residual	48359.078	125	386.873
	Total	54248.756	126
a. Dependent Variable: Decision
b. Predictors: (Constant), Pounds

Table 11 show the test for regression coefficients. The coefficient of the independent variable weight (Pounds) is -0.172 with a probability of p<0.01, indicating a significant regression coefficient for the independent variable; therefore, linear correlation between the independent and the dependent variables. The regression equation is thus Ŷ = 62.906-0.172X.

Table 11

Regression Coefficient Test (Decision)

Model		Unstandardized Coefficients		Standardized Coefficients	t	Sig.
		B	Std. Error	Beta
1	(Constant)	62.906	7.774		8.091	.000
	Pounds	-.172	.044	-.329	-3.902	.000
a. Dependent Variable: Decision

Table 12 shows the statistical test table for the linear regression model. The correlation coefficient was R=0.222, R²=0.049, adjusted R²=0.042, and the estimated standard error was 18.30691. The results show a linear correlation between the independent and dependent variables.

Table 12

Statistical Tests of Linear Regression Model (Submission)

Model	R	R Square	Adjusted R Square	Std. Error of the Estimate	Change Statistics					Durbin-Watson
					R Square Change	F Change	df1	df2	Sig. F Change
1	.222a	.049	.042	18.30691	.049	6.488	1	125	.012	2.017
a. Predictors: (Constant), Pounds
b. Dependent Variable: Submission

Table 13 shows the results of the variance test of the regression statistics, such as the sum of squares, degrees of freedom, mean square value, and the value of the statistic F and its probability of significance for the model. The results show that F=6.488 with a probability of P=0.012<0.05, indicating that the regression effect is significant in this case.

Table 13

ANOVA Test (Submission)

Model		Sum of Squares	df	Mean Square	F	Sig.
1	Regression	2174.356	1	2174.356	6.488	.012b
	Residual	41892.857	125	335.143
	Total	44067.213	126
a. Dependent Variable: Submission
b. Predictors: (Constant), Pounds

Table 14 shows the test of regression coefficients. The coefficient of the independent variable weight was -0.104 with a probability of p=0.012<0.05, indicating that the regression coefficient of the independent variable was significant; there is a linear correlation between the independent and the dependent variables. The regression equation is thus Ŷ = 40.882-0.104X.

Table 14

Regression Coefficient Test (Submission)

Model		Unstandardized Coefficients		Standardized Coefficients	t	Sig.
		B	Std. Error	Beta
1	(Constant)	40.882	7.236		5.650	.000
	Pounds	-.104	.041	-.222	-2.547	.012
a. Dependent Variable: Submission

DISCUSSION

A comparison of the data shows that the athletes winning by KO/TKO were 40.21%±22.27, Decision was 36.78%±20.88, and Submission was 23.01%±18.36. The percentage of the approaches used by male athletes to win in descending order was: KO/TKO 43.75%±22.98, Decision 33.35%±20.75, and Submission 22.92%±18.70. The rate of female athletes in descending order by the method they used to win was: Decision 46.04%±18:43, KO/TKO 30.64%±17.04, and Submission 23.23%±17.61. The comparison of two typical levels of Flyweight and Bantamweight for both men and women indicated that there was no significant difference between KO/TKO (p=0.934>0.05) and Submission (p=0.085>0.05), but there was a significant difference for Decision (p=0.043>0.05).

The results showed that the most common method used by male athletes to win was KO/TKO, then Decision, and finally Submission. In contrast, the most common method for female athletes was Decision, then KO/TKO, and finally Submission. Since there are only two equal-weight divisions for both male and female athletes, the comparison shows that there is only a significant difference between the Decision wins, with no significant difference between the other two.

The Kruskal-Wallis test showed that there was a significant difference in the styles used by male athletes to win across weight divisions, with p<0.01 (KO/TKO), p<0.01 (Decision), and p=0.027<0.05 (Submission). However, there was no significant difference in the way female athletes won across the different weight divisions, with p=0.724>0.05 (KO/TKO), p=0.259>0.05 (Decision), and p=0.169>0.05 (Submission). The data analysis showed significant differences in the methods used by male athletes to win in each weight class, while there were no significant differences in those employed by female athletes.

The Spearman’s rho bivariate correlation test showed that, for male athletes, the weight class had a significant positive correlation with KO/TKO wins (r=0.446, P<0.01), a significant negative correlation with Decision wins (r=-0.311, P<0.01), and a significant negative correlation with Submission wins (r=-0.255, P=0.004<0.05). Data analysis showed that the higher the weight divisions of the male athletes, the greater the chance of KO/TKO wins and the smaller the chance of Decision and Submission wins. In contrast, the results indicated no significant correlation between the female athletes’ weight divisions and their winning methods. Although there appears to be no significant correlation between weight class and the methods female athletes use to win, this assumption was based on contexts in which there are only three weight class data for female athletes. It is worth noting that data on fewer weight classes for female athletes are likely to influence correlation and that further research is needed to test the association.

Based on the results of the correlation analysis between weight classes and the approaches used to win, a linear regression analysis of the weight and methods of male athletes was conducted in this study. The results indicated a significant linear correlation between the independent variable (weight: pounds) and the dependent variable (KO/TKO, Decision, and Submission). The regression analysis may provide a specific approach for more precise predictions of the correlation effects between weight divisions and the methods used by elite MMA athletes to win. For example, assuming that a male athlete weighs 180 pounds, that brings into the equation Ŷ = -4.150 + 0.278X (KO/TKO), Ŷ = 62.906-0.172X (Decision), Ŷ = 40.882-0.104X (Submission), which corresponds to a KO/TKO probability of winning of about 45.89%, a Decision win probability of approximately 31.95%, and a Submission win probability of roughly 22.16%. It is worth noting here that this linear regression model is not intended to provide coaches or athletes with precise predictions of successful methods in particular weight divisions but rather to provide a reference that must be used in practice, taking other factors into account in the overall judgment.

This paper is not a study of who will win a fight but a projection of the most likely means used by an athlete to win a competition within different weight classes. Its key role is to provide coaches and athletes with additional information to optimize their training programs and competition strategies (28). Neither does this study provide an analysis of other influencing factors. Hence, coaches and athletes need to consider other factors when preparing for a specific competition, for example, the technical and tactical characteristics of the athlete and the opponent, and the particular environment of the competition, instead of considering this as a single constant factor.

Coaches train athletes at different levels of competition, and this study can inform coaches’ training programs for athletes in different weight classes. For example, elite heavyweight MMA fighters are most likely to win by KO/TKO and less likely to win by Decision and Submission. Coaches should consider this factor when scheduling training programs and preparing competitive strategies. MMA athletes should consider their weight divisions as a factor in their training strategies, allowing for more effective training plans. In addition, combat sports athletes will many times move up or down in weight classes to secure a competitive advantage, and coaches and athletes should consider the probability of methods used to win within different weight classes when planning whether to move up or down, thereby increasing the likelihood of success. In addition, MMA fans can also make more reasonable predictions about competition outcomes and the methods used by their favorite athletes to win based on the results of the present study.

The present study has several limitations. The first is that data on the athletes were collected from a single MMA organization, the UFC, even though it is, in fact, the leading organization in promoting the sport. Secondly, this study has only analyzed the factors of weight division with limited R² results,which might limit the linear regression model’s implementations. Therefore, the influence of other factors should also be considered according to the characteristic of mixed martial arts’ diverse techniques and tactics. A more comprehensive analysis will enable effective techniques and tactics to enhance an athlete’s overall performance. Finally, the fact that only three levels of data were available for female athletes’ weight divisions led to no significant correlation between weight classes and the methods used by athletes to secure victory. 

Future studies might therefore be conducted with data for the fourth division for female athletes being added or as more relevant data from other organizations are obtained. In order to examine the generalizations made in the present study or potential trends within the sport itself, future research into other MMA organizations and athletes – for example, amateur organizations and athletes – could be based on similar methods or by observing the same organization at different periods.

CONCLUSIONS

This study investigated the correlation between weight divisions and the methods used by top-ranking UFC mixed martial arts athletes to win. Results provided the following ratios for these methods: KO/TKO (40.21%±22.27), Decision (36.78%±20.88), and Submission (23.01%±18.36). The Mann-Whitney U test results used in the study showed no difference in KO/TKO and Submission as methods used by male and female athletes to secure a victory within the same weight divisions, compared with a difference in Decision. The Kruskal-Wallis test showed a difference in the methods used by winning male athletes across different weight divisions and no difference in those employed by female athletes.

Spearman’s rho bivariate correlation test showed that the weight division of male athletes had a positive correlation with KO/TKO and a negative correlation with Decision and Submission, and no correlation between female athletes’ weight divisions and how they won. Linear regression test results showed linear correlations between male athletes’ independent and dependent variables.

APPLICATIONS IN SPORT

This study may provide coaches and athletes with a valuable reference for the relationship between weight divisions and the methods used by winning athletes when optimizing training plans and competition strategies. Coaches and athletes should also consider the probability of different methods within different weight divisions when planning to move up or down weight classes to secure a competitive advantage. Sports fans may also use this study in order to predict competitive outcomes.

REFERENCES