Effect of quality of opposition on game performance indicators in elite male volleyball

Abstract

Performance in a volleyball match is the result of a dynamic and interactive process between two teams. The current study aimed to investigate the influence of the quality of opposition on skill performance indicators. A total of 550 teams’ performances (N = 550) from 275 sets of men’s European Championship 2019 was recorded and the effectiveness of 12 parameters from 5 basic volleyball skills was analyzed. A two-step cluster analysis was performed to divide 24 teams into three quality groups (upper, middle, lower) and 6 types of match status were created according to the quality of the opponents. Binomial logistic regression showed that for each type of match, the key performance factors that discriminate win and loss are differentiated, while the efficacy of attack win is associated with the success almost with all contexts studied. Other significant parameters were serve aces for matches between upper-quality teams, opponent’s errors for matches between teams of different quality and avoidance of blocked attacks for balanced matches. The findings emphasize the need for coaches to plan strategies that allow players with special serving abilities to risk for an ace, to enhance the side-out skills of their team and to work detailed on attack coverage systems.

Keywords

Performance analysis,skill,team sport

Introduction

In volleyball, a thorough analysis of skills performance is a valuable resource for coaches to help a team’s progress. Information obtained through regular monitoring and statistical analysis can be proved to be an extremely useful way to identify the strengths and weaknesses of a team. Using this information coaches can, on the one hand, give appropriate feedback to their athletes and, on the other hand, plan their strategic or tactical options more effectively. Moreover, it makes it possible to draw information about the characteristics and trends of the playing model and thus compare teams in a tournament.

The effect of skill performance on a team’s success has already been described at a competition and match level.^1–8 Nevertheless, it is important to consider that volleyball is played in independent sets. Sets can be considered as partly independent games with their scoreboard. Previous data show that winning a set relates directly to some performance indicators.^9–12 Understanding the ideal combination of these indicators can help a team achieve success in a given set.

Additionally, volleyball is characterized by alternating ball possession, while each team has a limited number (three) of ball contacts.¹³ Thus, a volleyball match is a direct interaction between two teams as their performance is influenced by the opponent’s performance and hence, one team’s strategy may undergo adjustments depending on the opponent’s strengths.The quality of opposition has been suggested to have an important influence on volleyball teams’ performances.^14–16 Furthermore, differences in the use and performance of volleyball skills were found through age-groups competition categories in terms of teams’ quality and match status.¹⁷ Different performance trends are identified according to these variables such as the setting importance during high ranked teams matches, attack efficiency during low ranked teams matches and serve type, block strategy and serve efficacy in high vs low ranked teams matches.¹⁸ Some studies were related to national championships,^15,16 while others were related to national teams international competitions.^18,14 The latter supported the belief that their results should be in line with the standards of current elite volleyball. Despite their contribution, these studies classified teams in two categories, labelled as high and low quality. This two-level categorisation was restrictive because of the limited participation of teams in the elite tournaments (usually 12 or 16 national teams).

The high-low quality dichotomy lacks the necessary sensitivity to identify changes in performance indicators such as the function of the quality of opposition¹⁹ Besides, this separation limits the concretization of the conclusions because, in reality, teams in male volleyball could be categorized into more categories with clear differences between them.²⁰

Because of all the above, the present study aimed to identify the effect of the quality of opposition on volleyball key technical performance indicators concerning the categorization of national teams competing in an international tournament in more than two quality groups. This information could help coaches to understand how players face different matches during a tournament and how they should prepare matches against teams with a higher, lower or equal quality of opposition.

Methods

The study focused on the Men’s European Volleyball Championship 2019 in which 24 national teams competed enabling a possible division of the teams into more than two categories. The high standard of the tournament is confirmed by the fact that European teams can attain better performance than teams from other continents as at least six European teams were ranked 12th or above in the World Championships since 2010.²¹

A total of 76 matches (at least five matches of each team) were sampled and the data were collected from 550 teams’ performances (N = 550) in 275 sets.

A two-step cluster analysis (distance measure: log-likelihood, clustering criterion: Schwarz’s Bayesian criterion) was used to group teams into competitive subcategories. The number of clusters was fixed at three and the criterion variables were: total points at the end of the tournament (win 3-0 or 3-1 awarded 3 points, win 3-2 awarded 2 points, defeat 2-3 awarded 1point, defeat 1-3 or 0-3 awarded 0 points), the ratio of sets won and lost, percentage of won sets and the ratio of points won and lost. The first cluster consisted of the seven first teams of the tournament (Serbia, Slovenia, Poland, France, Russia, Italy, Ukraine) and the 9^th team of the final ranking (Belgium) was labelled as “upper quality” (UQ). The second cluster was labelled as “middle quality” (MQ) and consisted of the 8^th team (Germany), teams positioned 10^th to 15^th (Nederland, Bulgaria, Turkey, Czech Republic, Finland, Spain), and the 17^th team (North Macedonia). The third cluster included the 16^th team (Greece), the seven lowest-ranked teams, positions 18^th to 24^th (Montenegro, Slovakia, Portugal, Romania, Belarus, Austria, Estonia) and was labelled “lower quality” (LQ).

Differences on quality of opposition led to 6 types of game context were considered as follows: UQ vs UQ (N = 92 teams’ performances, 46 sets, 12 matches), UQ vs MQ (N = 150 teams’ performances, 75 sets, 22 matches), UQ vs LQ (N = 112 teams’ performances, 56 sets, 17 matches), MQ vs MQ (N = 50 teams’ performances, 25 sets, 6 matches), MQ vs LQ (N = 108 teams’ performances, 54 sets, 14 matches) and LQ vs LQ (N = 38 teams’ performances, 19 sets, 5 matches).

The primary recorded and evaluated skills were: 12,243 serves, 9,985 serve’s receptions, 8,497 attacks after serve reception, 5,393 attacks after defence and 6,239 blocks.

For the evaluation scale of each skill, a six-level ordinal scale is employed, with the value of “one” indicating a poorly executed skill and the value of “six” an excellent executed skill (Table 1). Further definitions for the evaluation scale per skill are included in Drikos et al.²⁰ about the serve, Costa et al.²² about the attack, Palao et al.² about the block, and Drikos²³ about the serve’s reception.

Table 1.

Evaluation of ordinal scale for volleyball skills.

Skill/Level	Serve	Attack 1,2	Block	Reception
6	Ace(point)	Win-kill(point)	Win-kill (point)	Excellent reception. All options for an attack without adjustments for the setter
5	OverThe ball to the serving team	The ball to the attacking team with good conditions or to the defending team with bad conditions	The ball to the blocking team with good conditions or to the attacking team with bad conditions	Good reception.All options for the attack
4	One option for the attack for the receiving team	The ball to the attacking or defending team with average conditions	The ball to the blocking or attacking team with average conditions	Two options for the attack from the sidelines
3	Two options for the attack for the receiving team	The ball to the attacking team with bad conditions or to the defending team with good conditions	The ball to the blocking team with bad conditions or to the attacking team with good conditions	One option for the attack or attack out of the system
2	All options for the attack on the receiving team	Stuffed by a Kill block (lost point)	Error on the net (Incorrect touch of the net, lost point)	OverpassThe ball was passed directly to the serving team court.
1	Error (lost point)	Error (lost point)	Error (lost point)	Error (lost point)

Set statistics included variables of efficacy (the number of the categorized events divided by the total number of the skill) for serve, reception, attack 1 (after reception), attack 2 (after defence) and block. A volleyball team can get points in four different ways: serve, block, attack and the points gained from the errors of the opposing team. To materialize a complete view of all the ways teams earned points, a variable of opponents’ errors was also added. As for the opponent errors all the unforced errors (serve errors, attack errors, an illegal touch of the net, mishandling of the ball) of the opponent are included divided by total set points. Hence, the data set consists of the following twelve (12) performance indicators (abbreviations): 1) Serve win (Swin%), 2) Serve Error (Serr%), 3) Reception precise % [Reception perfect %+ Reception excellent %](Rprc%) 4) Reception Error (Rerr%), 5) Attack 1 win (A1win%), 6) Attack 1 Errors (A1err%), 7) Attack 1 blocked (A1blk%), 8) Attack 2 win (A2win%), 9) Attack 2 error (A2err%), 10) Attack 2 blocked (A2blk%), 11) Block win (Bwin%), 12) Opponent errors per set points (Oerr%).

The data were recorded and processed by a volleyball expert scout man of the Greek Volleyball Association using Data Volley software.²⁴ In order to check the intra-observer’s reliability test-retest procedure of 10% of the total sample (28 sets randomly selected) with four weeks interval to avoid any possible adverse learning effects were established. Another independent volleyball expert was asked to observe the same selected 28 sets to check inter-observer reliability. The weighted kappa values for both procedures, intra-observer = 0.92, 0.83, 0.87 and 0.82 and inter observer = 0.89, 0.80, 0.83 and 0.81 for serve, reception, attack and block, respectively showed very good values.²⁵

For statistical analysis, initially, descriptive statistics were calculated for winning and losing teams for each type of game context according to the quality of opposition (upper quality vs upper quality, upper quality vs middle quality, upper quality vs lower quality, middle quality vs middle quality, middle quality vs lower quality and lower quality vs lower quality). For assessment of the magnitude, a change in the mean was expressed as standardized effect size (ES). Differences between winning and losing teams were calculated for each performance indicator according to the quality of opposition. Quantitative evaluation of effect sizes were: 0–0.2 (trivial), >0.2–0.59 (small), ≥0.6–1.19 (moderate), ≥1.2–2 (large), and >2 (very large).²⁶

Secondly, a binomial logistic regression was used to evaluate the association between performance indicators in each match context according to winning and losing the set. The use of this non-linear model of regression allows obtaining the estimated regression coefficients representing the estimated changes in the log-odds. There were calculated the Odds ratios (OR) and their 95% confidence interval (CI). The standard interpretation of the binomial logit is that for a unit change in the predictor variable, given that the other characteristics in the model remain unchanged, the logistic of comparison outcome relative to the base outcome is expected to change by its odds ratio.²⁷ The statistical analyses were performed using the statistical program SPSS for Windows, version 23.0 and statistical significance was set at p < 0.05.

Results

In the data set, there were no missing values, extreme scores, or outliers, and the basic statistical assumptions were tested and met. In particular, there was no multicollinearity between the dependent variables as the simple correlations, presented in Table 2 were all <|.65|). Besides, based on the statistics presented in Table 2 the twelve variables appear not to be affected even by moderate collinearity as tolerances were high (from .668 for A1win to .967 for Swin) and variance inflation factors were lower than 10 (from 1.035 for Swin to 1.498 for A1win). Therefore all dependent variables were appropriate for multivariate analysis.

Table 2.

Collinearity diagnostics and correlations indices among the 12 performance indicators.

Performance indicators	1	2	3	4	5	6	7	8	9	10	11	12	Tol.	VIF
1.Swin%.	1												.967	1.035
2.Serr%	–.098	1											.932	1.073
3.Rprc%	.029	–.102	1										.768	1.302
4.Rerr%	.029	.102	–.411	1									.815	1.228
5.A1win%	.143	–.160	.259	–.114	1								.668	1.498
6.A1err%	–.002	.045	–.106	.013	–.311	1							.870	1.149
7.A1blk%	–.064	–.013	–.089	.000	–.363	–.023	1						.829	1.207
8.A2win%	.040	–.060	.012	–.026	.194	–.092	–.062	1					.767	1.304
9.A2err%	.019	.059	–.052	.006	.000	–.011	–.001	. –299	1				.868	1.152
10.A2blk%	–.042	.038	–.085	–.006	–.113	–.014	.101	–.302	–.062	1			.847	1.181
11.Bwin%	–.012	–.088	.048	–.068	.076	–.051	–.053	.005	.020	–.108	1		.959	1.043
12.Oerr%	–.017	–.074	–.041	.044	–.110	.033	–.042	–.072	.075	–.056	–.092	1	.945	1.058

Tol.: tolerance = 1–R_i². with R_i² = % variance in common with the other variables.

VIF = 1/(1–R_i²) = variance inflation factor.

Table 3 summarizes the descriptive analysis and the magnitude-based differences between a win and loss for teams’ performances related statistics employed in this study for all game context.

Table 3.

Differences between winning and losing teams according to the match status (mean ± standard deviation, significance and effect size).

Type of opposition	UQ vs UQ			UQ vs MQ			UQ vs LQ			MQ vs MQ			MQ vs LQ			LQ vs LQ
Type of result	Loss	Win		Loss	Win		Loss	Win		Loss	Win		Loss	Win		Loss	Win
	MeanSD	MeanSD	Sig/ES	MeanSD	MeanSD	Sig/ES	MeanSD	MeanSD	Sig/ES	MeanSD	MeanSD	Sig/ES	MeanSD	MeanSD	Sig/ES	MeanSD	MeanSD	Sig/ES
Swin%	6.37 ± 5.1	7.58 ± 6.2	.011	4.95 ± 4.8	7.67 ± 5.0	***.072	4.60 ± 4.8	8.10 ± 6.2	***.092	4.69 ± 7.0	8.46 ± 6.2	.077	4.30 ± 4.5	7.64 ± 6.4	**.085	4.17 ± 3.8	5.31 ± 5.9	.014
Serr%	20.49 ± 8.2	17.76 ± 7.5	.030	22.78 ± 11.0	17.29 ± 6.6	***.086	22.59 ± 10.5	15.88 ± 7.5	***.121	19.13 ± 6.6	17.52 ± 7.5	.013	19.08 ± 9.0	16.34 ± 6.8	.029	19.04 ± 8.5	15.32 ± 10.3	.039
Rprc%	45.24 ± 15.4	45.35 ± 16.7	.000	41.16 ± 12.4	48.19 ± 16.7	***.055	45.78 ± 13.7	54.58 ± 17.4	**.075	39.22 ± 14.5	50.54 ± 14.0	**.141	43.41 ± 13.8	53.20 ± 17.2	***.091	53.56 ± 17.4	58.08 ± 15.1	.020
Rerr%	9.30 ± 8.0	8.13 ± 6.1	.007	9.43 ± 6.2	6.39 ± 6.	***.058	9.75 ± 7.6	5.88 ± 6.1	**.074	10.24 ± 7.5	5.95 ± 9.4	.062	9.11 ± 7.3	5.21 ± 5.5	**.084	6.36 ± 7.0	5.35 ± 5.2	.007
A1win%	42.39 ± 11.4	53.17 ± 13.5	***.159	44.02 ± 13.5	60.17 ± 13.1	*** .272	41.73 ± 15.8	62.35 ± 13.7	***.331	45.27 ± 10.0	56.83 ± 12.5	***.214	46.39 ± 11.8	57.39 ± 12.4	***.173	47.53 ± 8.0	57.89 ± 13.0	**.200
A1err%	10.24 ± 8.1	6.79 ± 6.9	.051	8.66 ± 6.6	5.71 ± 6.3	**.050	8.42 ± 7.3	4.53 ± 5.3	**.086	6.89 ± 5.5	5.49 ± 6.2	.015	9.47 ± 7.0	4.50 ± 5.0	***.146	7.01 ± 5.0	4.91 ± 4.7	.047
A1blk%	11.36 ± 9.2	7.02 ± 6.3	.071	10.66 ± .7.5	6.00 ± 6.3	***.101	9.31 ± 8.3	6.12 ± 6.2	*.046	13.57 ± 6.6	5.10 ± 5.6	***.333	11.05 ± 8.5	6.64 ± 5.7	**.085	6.7 ± 5.4	9.53 ± 9.1	.036
A2win%	33.66 ± 16.8	49.90 ± 14.6	***.214	41.61 ± 19.5	51.77 ± 17.5	***.071	35.45 ± 20.4	56.42 ± 1.0	***.259	37.75 ± 18.9	46.72 ± 18.4	.057	39.53 ± 19.1	47.87 ± 17.2	**.051	37.42 ± 17.6	55.53 ± 15.0	**.245
A2err%	9.22 ± 9.6	7.76 ± 8.0	.007	9.08 ± 11.5	7.69 ± 7.8	.005	8.82 ± 10.4	4.72 ± 7.5	*.049	10.75 ± 13.9	7.17 ± 8.4	.025	9.13 ± 10.5	7.06 ± 8.0	.012	5.28 ± 6.4	6.37 ± 7.5	.006
A2blk%	11.94 ± 11.6	6.33 ± 8.3	**.073	13.83 ± 13.8	7.30 ± 8.9	***.075	9.66 ± 12.1	6.55 ± 9.2	.021	11.14 ± 9.5	7.97 ± 8.4	**.033	10.36 ± 10.0	6.85 ± 8.2	*.036	15.07 ± 9.6	1.56 ± 3.7	***.476
Bwin%	13.90 ± 9.8	22.87 ± 11.2	***.157	15.05 ± 13.0	22.88 ± 10.6	***.100	17.96 ± 16.2	22.63 ± 16.8	.020	16.02 ± 13.47	27.13 ± 15.9	**.129	16.88 ± 13.4	24.07 ± 15.5	**.059	14.61 ± 11.5	22.71 ± 8.5	**.146
Oerr%	14.54 ± 4.7	15.73 ± 5.3	.014	13.92 ± 4.9	16.12 ± 5.7	*.042	12.90 ± 4.8	16.16 ± 5.3	***.096	14.19 ± 4.6	14.62 ± 4.6	.002	13.21 ± 4.2	15.39 ± 4.5	**.060	12.86 ± 5.9	13.69 ± 5.0	.006
N	92 (46 + 46)			150 (75 + 75)			112 (56 + 56)			50 (25 + 25)			108(54=54)			38(19 + 19)

Note: Significance *p < 0.05, **p < 0.01, ***p < 0.001; Effect size: 0–0.2 (trivial), >0.2–0.6 (small), ≥0.6–1.2 (moderate), ≥1.2–2 (large), and >2 (very large).

The binomial logistic regression models were significant in all the context studies, indicating that the predictors’ model provides a statistically significant improvement over the constant-only model: Upper quality vs upper-quality team games (91.3% of the cases correctly classified, χ² 6, N = 92, 79.924, p < .001, R² = .77), upper quality vs middle-quality team games (91.3% of the cases correctly classified, χ² 7, N = 150, 127.329, p < .001, R² = .74), upper quality vs lower quality team games (92.8% of the cases correctly classified, χ² 5, N = 112, 113.851, p < .001, R² = .85), middle quality vs middle quality (92.0% of the cases correctly classified, χ² 4, N = 50, 52.916, p < .001, R² = .87), middle quality vs lower quality team games (89.8% of the cases correctly classified, χ² 7, N = 108, 96.132, p < .001, R² = .79) and lower quality vs lower quality team games (86.8% of the cases correctly classified, χ² 1, N = 38, 25.170, p < .001, R² = .65).

On the other hand, the binomial logistic regression detected significant key performance indicators that were related to the set result for all the type of matches. Parameters estimates and odds ratio with their 95% confidence limits are in Table 4 (for the sake of parsimony only statistically significant parameters are reported). For upper quality vs upper-quality team games for a single point of increase in the predictors of serve win, attack 1 & 2 win, block win and opponent errors there is a greater likelihood of winning the set of 1.21, 1.18, 1.18, 1.28, 1.43, respectively. For a single point of decrease in reception errors there is a greater probability of winning (given by the reciprocal of -.228) by 1.26 (95% C.I., 1.052-1.501). For matches competed between upper quality vs middle-quality teams, 7 predictors had significant parameters comparing win with loss of a set. The influence of serve win & error, reception error, attack 1& 2win, block win and opponent errors are relatively equal. Improving by one unit of measurement in each of these performance indicators, increased the probability of winning the set from 1.05 (attack 2 wins) to 1.32 times (opponent errors).

Table 4.

Results of binomial logistic regression.

Type of opponent (% correct classification)	Variable	B	S.E.	Wald	df	Sig.	Exp(B)	95% C.I. for EXP(B)
Type of opponent (% correct classification)	Variable	B	S.E.	Wald	df	Sig.	Exp(B)	Lower	Upper
UQ vs UQ (91.3%)	Swin%	.194	.072	7.223	1	.007	1.215	1.054	1.400
	Rerr%	–.228	.091	6.331	1	.012	.796	.666	.951
	A1win%	.164	.050	10.711	1	.001	1.178	1.068	1.299
	A2win%	.164	.046	12.903	1	.000	1.178	1.077	1.289
	Bwin%	.246	.062	15.518	1	.000	1.279	1.132	1.446
	Oerr%	.359	.107	11.216	1	.001	1.432	1.161	1.766
UQ vs MQ (91.3%)	Swin%	.250	.073	11.843	1	.001	1.284	1.114	1.481
	Serr%	–.078	.036	4.612	1	.032	.925	.861	.993
	Rerr%	–.085	.044	3.729	1	.053	.918	.842	1.001
	A1win%	.136	.028	23.832	1	.000	1.146	1.085	1.210
	A2win%	.051	.016	9.841	1	.002	1.053	1.019	1.087
	Bwin%	.126	.029	18.539	1	.000	1.135	1.071	1.202
	Oerr%	.280	.072	14.932	1	.000	1.323	1.148	1.524
UQ vs LQ (93.8%)	Swin%	.313	.104	9.116	1	.003	1.368	1.116	1.677
	Serr%	–.140	.057	6.060	1	.014	.869	.777	.972
	A1win%	.163	.048	11.812	1	.001	1.177	1.073	1.292
	A2win%	.124	.033	14.353	1	.000	1.133	1.062	1.208
	Oerr%	.444	.115	14.792	1	.000	1.559	1.243	1.955
MQ vs MQ (92.0%)	A1win%	.204	.119	2.914	1	.088	1.226	.970	1.549
	A1blk%	–.563	.228	6.112	1	.013	.570	.365	.890
	A2win%	.186	.095	3.831	1	.050	1.205	1.000	1.452
	Bwin%	.283	.116	5.976	1	.015	1.328	1.058	1.666
MQ vs LQ (89.8%)	Swin%	.292	.096	9.146	1	.002	1.339	1.108	1.617
	Rerr%	–.250	.083	9.161	1	.002	.779	.662	.916
	A1win%	.145	.042	11.851	1	.001	1.156	1.064	1.255
	A1err%	–.210	.070	9.138	1	.003	.810	.707	.929
	A1blk%	–.114	.062	3.409	1	.065	.892	.790	1.007
	A2win%	.091	.026	12.237	1	.000	1.095	1.041	1.152
	Bwin%	.127	.038	10.920	1	.001	1.135	1.053	1.224
	Oerr%	.420	.140	8.984	1	.003	1.522	1.156	2.003
LQ vs LQ (86.8%)	A2blk%	–.306	.097	10.050	1	.002	.736	.610	.890

Concerning, upper quality vs lower quality team games, five predictors had significant parameters. One unit of increase in serve win, attack 1 & 2win, and opponent errors increased the odds of a win by 1.37, 1.18, 1.14, 1.56 times respectively, while for one unit of decrease in serve errors the odds of winning the set increased (given by the reciprocal of −.140) by 1.15 (95% C.I., 1.029-1.287). Four predictors had significant parameters for games between middle-quality teams. Attack 1 &2 win and opponent errors increased the odds to win set by 1.26, 1.20 and 1.33 times if increased by one unit, while one unit of decrease in attack 1 blocked to increase the odds of winning (given by the reciprocal of −.563) by 1.75 (95% C.I., 1.124–2.742). Eight predictors had significant parameters when comparing won to lost sets for middle quality vs lower quality team games including serve win, reception error, attack 1&2 win, attack 1 error and blocked, block win and opponent error. So, according to the standard interpretation of the binomial logit, the influence of the selected parameters is relatively equal. The improvement of each one of them by one unit ceteris paribus the others increased the odds of a set’s win by 1,09 (attack 2 win) to 1.52 (opponent errors). About games between lower quality teams, only one predictor had a significant parameter. Improvement (reduction) by one unit of the variable attack 2 blocked would be expected to increase the odds to win a set (given by the reciprocal of -.306) by 1.36 (95% C.I., 1.124-1.641).

Discussion

This study aimed to identify the effects of the quality of opposition on volleyball key technical performance indicators that contributed to team success in a set. A three-grade categorization of the teams of an elite volleyball tournament was established and six types of matches were created. The suggestion of Taylor et al.¹⁹ that it is preferable to determine the quality of opposition taking into consideration performance measures than categorizing based on their ranking within a particular tournament was worthy in the case of Volleyball European Championship 2019 because the final ranking of the teams, at least in four cases did not identify with the results of the clusters analysis of the competitive subcategories.

The results showed that the effect of quality of the opposition has a relative influence on performance indicators, at the behavioural and outcome level. In line with the findings of Garcia de Alcaraz & Marcelino¹⁷ and Marcelino et al.^18,14 the results of the present study support the notion that critical performance indicators in volleyball are influenced by the quality of opponent. Therefore, for each type of match, the key performance factors that discriminate win and loss are differentiated. Additionally, the weight of each performance factor when improving by one unit of measurement differentiates according to the quality of the opponent.

The main finding of the study showed various performance indicators differentiating winning and losing the set in each match context with two variables associated with the success in all contexts studied apart of one (LQ vs LQ): the efficacy of attack win 1 & 2. As attack has been regularly identified as the main volleyball skill to predict the outcome of the games.^3,4,11,28,29 This finding is in agreement with those for the men’s game where the effectiveness of attack is the basic correlate with the victory.⁵,³⁰

Additionally, this study revealed that in all contexts of the game the improvement of a unique performance indicator is not adequate to increase significantly the odds ratio to win a set. As all the odds ratios in this study were between 1.053 and 1.75 and represent a weak relationship (a strong one is over 3.0),³¹ it should be emphasized the need to combine improvement in various performance indicators to increase significantly the probability of winning a set.

Particularly, the results for balanced matches, between upper-quality teams, identified the importance of efficacy of attack win 1 & 2, serve win, reception errors, block win and opponent errors. In these type of matches, the importance of serve and reception is extremely high and has a double fold dimension. The success of a serve ace which increases the probability of win the set by 1.2 times and the ability to avoid an ace from the opponents serve which increase the probability to win the set by 1.3 times. According to many researchers^32–34 the serve is reported as a disadvantage for the team that executes it in elite men’s volleyball. Nevertheless, the ability to serve an ace in a balanced match between high-quality teams is priceless. In line with this, a slight increase in % ace serve occurred in the 2019 edition of the Volley Nations League tournament compared with the previous ones as is stated in the technical report of the F.I.V.B. committee.³⁵ This result of the present study is in accordance with the findings of Zetou et al.³ who stated that a serve direct point is the best predictor among skills of complex II (serve, block, attack 2). For the practical implementation of the above elements, coaches should spend more training time in practising players in the power serve and should allow the players with special serving abilities to work hard for an ace during high-quality matches.¹⁴

Contrary to the notion of some researchers that reception’s efficacy is a predictor of competitive success^33,36,37 in this study the errors in reception appear to be the unique factor affecting the probability of winning a set. This finding is in line with the results of other studies²⁰,^38,39 which proposed that the avoidance of reception errors is an important factor. However, considering that volleyball actions are performed in a specific order forming a chain of events that affect one another, it is logical to assume that quality of reception influences the predictors of attack after the reception,¹,³⁴ even though the quality of reception is not a significant parameter on its own.

Regarding the block, it is one of the game’s most important actions in elite volleyball^7,40,41 and establishes the difference among top-level male teams and it is the most important skill in opposition to attack.² Following previous studies, the results of this study indicate that the 1% increase of block win actions increases the probability to win a set between two upper-quality teams by 1.3 times.

The general trend that the lower quality teams perform poorly in all skills^6,7,12 is visible in the matches between different quality teams (upper vs middle, upper vs lower). Therefore, in the present study, the upper-quality teams score serve, attack and block points with higher efficiency and perform less spike and serve errors when playing against lower quality teams. This skill-scoring advantage combined with the more unforced errors made by weaker teams affects success in upper-quality teams due to the relationship between skill performance and set outcome.^7,42 The main performance indicator that weaker teams need to improve to increase their chances of winning against higher quality teams is the reduction of unforced errors. For a one-unit reduction of unforced errors, the probability of winning a set versus a high-quality team increased by 1.3 times and by 1.6 times for middle and lower quality teams, respectively.

Continuing with unbalanced matches and more specifically with middle quality vs lower quality teams the importance of all performance indicators connecting with attack 1 is highlighted. This finding is in partial agreement with those of Stutzig et al.⁶ who found the impact of attack after reception was marginal for matches between elite national teams ranking 1^st - 8^th position in various top-level tournaments. A novel finding of the present study is the importance of attack after reception as win, errors and blocked attacks are significant parameters for winning a set between teams of the 2^nd and 3^rd quality cluster of this analysis. As a consequence, coaches of middle and lower quality teams have to emphasize the enhancement of the side out skills of their team through all the procedures, like players’ selections, improving individual skills of players and scheduling of training units.

Returning to balanced matches between middle vs middle-quality teams and lower vs lower quality teams the performance indicator of block avoidance is highlighted as a crucial factor.

The results of the current analysis show that for middle-quality teams the decrease of one unit in blocked attacks after reception increases the probability of winning a set against an equal quality team by 1.8 times. Concerning lower quality teams the decrease of one unit in blocked attacks after defense increases the probability of winning a set against an equal quality team by 1.4 times ceteris paribus the other variables. The block avoidance is related not only to the physical characteristics and technical abilities of the attackers^43–45 but also to attack coverage. Attack coverage is a pre-contact defensive action that coincides with the team’s attack.⁴⁶ During attack coverage, the attacking team manages to volley the blocked attack before it lands in its court and to organize the counter-attack.⁴⁷ According to Hilenio et al.⁴⁸ attack coverage is more effective in the counterattack phase (attack after defence). Supplementary to this, results of this study indicate that coaches have to focus on attack coverage systems separated into coverage after reception and coverage after defence aiming to reduce the number of blocked attacks.

Regarding the evaluation of the proposed models, prediction success is extremely high (≥90%) for all type of matches and the percentage of the variance of the dependent variable explained by the significant performance indicators labelled as Nagelkerke R² is between 65% (LQ vs LQ) and 87% (MQ vs MQ). This suggests that the set of predictors discriminates between winners and losers successfully for each type of match.

Conclusions

Overall, the two-step cluster analysis applied to this study grouped 24 national teams in three subcategories and as a consequence led to six types of match contexts depending on the quality of oppositions. For each type of match, the effect of quality of the opposition has a relative influence on performance indicators as the key performance factors that discriminate win and loss are differentiated. In the majority of the game contexts (5 out of 6), the effectiveness of attack after reception and after defence is the most important skill. In addition to the previous, the effectiveness in skills of complex II as serve and block increases the probability of winning in balanced high-quality matches. Furthermore, teams, which are lower ranked, should reduce unforced errors to claim a win from higher quality teams, while in matches between medium and lower quality teams the block avoidance is an influential factor for success. The level of the European teams in the World Championship of 2018 (10 European teams participate, all of them were ranked above 16th position in a total of 24 teams) would allow an empirical generalization of the results and of the conclusions also in elite-level male volleyball worldwide. Coaches should concentrate on those skills when designing training plans and tactical scenarios for the competition. The importance of these factors is reflected in changes in team strategy as a response to the quality of the opponent.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Sotirios Drikos

Dimitrios G Balasas

References

Eom

Schutz

NR.

Statistical analysis of volleyball team performance. Res Q Exerc Sport 1992; 63: 11–18.

Palao

Santos

Ureña

Effect of team level on skill performance in volleyball. Int J Perform Anal Sport 2004; 4: 50–60.

Zetou

Tsigilis

Moustakidis

, et al. Playing characteristics of men’s Olympic volleyball teams in complex II. Int J Perform Anal Sport 2006; 6: 172–177.

Zetou E, Moustakidis A, Tsigilis N, Komninakidou A. Does Effectiveness of Skill in Complex I Predict Win in Men’s Olympic Volleyball Games? J Quant Anal Sport 2007; 3(4): 1–11.

Monteiro

Mesquita

Marcelino

Relationship between the set outcome and the dig and attack efficacy in elite male volleyball game. Int J Perform Anal Sport 2009; 9: 294–305.

Stutzig

Zimmermann

Büsch

, et al. Analysis of game variables to predict scoring and performance levels in elite men’s volleyball. Int J Perform Anal Sport 2015; 15: 816–829.

Marcelino

Mesquita

Afonso

The weight of terminal actions in volleyball. Contributions of the spike, serve and block for the teams’ rankings in the world league 2005. Int J Perform Anal Sport 2008; 8: 1–7.

Drikos

Vagenas

Multivariate assessment of selected performance indicators in relation to the type and result of a typical set in men’s elite volleyball. Int J Perform Anal Sport 2011; 11: 85–95.

Marcelino

Mesquita

Palao

, et al. Home advantage in high-level volleyball varies according to set number. J Sport Sci Med 2009; 8: 352–356.

10.

Rodriguez-Ruiz

Quiroga

Miralles

, et al. Study of the technical and tactical variables determining set win or loss in top-level European men’s volleyball. J Quant Anal Sport 2011; 7: 1–13.

11.

Ciemiński

The efficiency of executing technical actions in volleyball and the teams’ gender and sports level. Trends Sport Sci 2018; 25: 159–165.

12.

Drikos

Sotiropoulos

Barzouka

, et al. The contribution of skills in the interpretation of a volleyball set result with minimum score difference across genders. Int J Sports Sci Coach 2020; 15: 542–551. 31

13.

F.I.V.B. Official volleyball rules. Anaheim, CA, USA: FIVB, 2012.

14.

Marcelino

Sampaio

Mesquita

IM.

Attack and serve performances according to the match period and quality of opposition in elite volleyball matches. J Strength Cond Res 2012; 26: 3385–3391.

15.

Campos

FAD

Stanganélli

LCR

Campos

LCB

, et al. Performance indicators analysis at Brazilian and Italian women’s volleyball leagues according to game location, game outcome, and set number. Percept Mot Skills 2014; 118: 347–361.

16.

García-De-Alcaraz

Cui

, et al. Interactive effects of home advantage and quality of opponent in Chinese women’s volleyball association league. Int J Perform Anal Sport 2020; 20: 107–117. 2

17.

García-de-Alcaraz A, Marcelino R. Influence of match quality on men’s volleyball performance at different competition levels. Int J Perform Anal Sport 2017; 17(4): 394–405.

18.

Marcelino

Mesquita

Sampaio

Effects of quality of opposition and match status on technical and tactical performances in elite volleyball. J Sports Sci 2011; 29: 733–741.

19.

Taylor

Mellalieu

James

, et al. The influence of match location, quality of opposition, and match status on technical performance in professional association football. J Sports Sci 2008; 26: 885–895. Jul

20.

Drikos

Sotiropoulos

Papadopoulou

, et al. Multivariate analysis of the success factors in high-level male volleyball: a longitudinal study. Trends Sport Sci 2019; 177: 177–185.

21.

Konaka

A quantitative method for evaluating the skills of national volleyball teams: prediction accuracy comparisons of the official ranking system in the worldwide tournaments of 2010s. In: Karlis

Ntzoufras

Drikos

(eds) MathSport international 2019. Athens: Propobos Publications, 2019, pp.202–216.

22.

Costa

GDCT

Afonso

Barbosa

, et al. Predictors of attack efficacy and attack type in high-level Brazilian women’s volleyball. Kinesiology 2014; 46: 242–248.

23.

Drikos S. Pass level and the outcome of attack for age categories in male Volleyball. J Phys Act Nutr Rehabil 2018; 3(1): 428–38.

24.

Data Project. Data volley. Bologna, Italy, www.dataproject.com/EU/en/Volleyball (2017, accessed 3 November 2017)

25.

Altman

Ludbrook

Altman

DG.

Practical statistics for medical research. ANZ J Surg 1991; 61: 963–964.

26.

Hopkins

WG.

Measures of reliability in sports medicine and science measurement error & reliability. Sport Med 2000; 30: 1–15.

27.

Tabachnick

Fidell

LS.

Using multivariate statistics. 5th ed. Pearson, Allyn and Bacon, 2007, pp.60–116.

28.

Castro

Souza

Mesquita

Attack efficacy in volleyball: elite male teams. Percept Mot Skills 2011; 113: 395–408.

29.

Barzouka

Comparison and assessment of the setting zone choices by elite male and female volleyball setters in relation to the reception quality. J Phys Educ Sport 2018; 18: 2014–2021.

30.

Patsiaouras

Charitonidis

Moustakidis

, et al. Comparison of technical skills effectiveness of men’s national volleyball teams. Int J Perform Anal Sport 2009; 9: 1–7.

31.

Chen

Cohen

Chen

How big is a big odds ratio? Interpreting the magnitudes of odds ratios in epidemiological studies. Commun Stat – Simul Comput 2010; 39: 860–864. 31

32.

Kountouris

Drikos

Aggelonidis

, et al. Evidence for differences in men’s and women’s volleyball games based on skills effectiveness in four consecutive Olympic tournaments. Compr Psychol 2015; 4: 1–7.

33.

Pena

Rodriguez-Guerra

Busca

, et al. Which skill and factors better predict winning and losing in high-level men’s volleyball? J Strength Cond Res 2013; 27: 487–493.

34.

Drikos

Ntzoufras

Apostolidis

Bayesian analysis of skills importance in world champions men’s volleyball across ages. Int J Comput Sci Sport 2019; 18: 24–44.

35.

F.I.V.B. Picture of the game. 2019 Volleyball Nations League [Internet]. 2019. Available from: https://www.fivb.com/-/media/2020/fivb corporate/volleyball/rules of the game/72019pictureofthegamereportfinal version.pdf?la=en.

36.

Silva

Lacerda

João

PV.

Game-related volleyball skills that influence victory. J Hum Kinet 2014; 41: 173–179.

37.

Bergeles

Barzouka

Nikolaidou

ME.

Performance of male and female setters and attackers on Olympic-level volleyball teams. Int J Perform Anal Sport 2009; 9: 141–148.

38.

Florence

Vehrs

Fellingham

, et al. Skill evaluation in women’s volleyball. J Quant Anal Sport 2008; 4.

39.

Lobietti R, Michele R, Merni F. Relationships between performance parameters and final ranking in professional Volleyball. In: Dancs H, Hughes M, Ekler J, editors. World Congress of Performance Analysis in Sports 7. Szombathely: Berzenyi College; 2006. p. 474–83.

40.

Araújo R, Castro J, Marcelino R, Mesquita IR. Relationship between the Opponent Block and the Hitter in Elite Male Volleyball. J Quant Anal Sport 2011 May 4; 6(4): 1–12.

41.

Marelic

Rešetar

Jankovic

Discriminant analysis of the sets won and the sets lost by one team in a1 Italian Volleyball League – a case study. Kinesiology 2004; 36: 75–82.

42.

Yu Y, García-De-Alcaraz A, Wang L, Liu T. Analysis of winning determinant performance indicators according to teams level in Chinese women’s volleyball. Int J Perform Anal Sport 2018; 18(5): 750–63.

43.

García-De-Alcaraz

Ortega

Palao

JM.

Effect of age group on male volleyball players´ technical-tactical performance profile for the spike. Int J Perform Anal Sport 2015; 15: 668–686.

44.

Forthomme

Croisier

J-L

Ciccarone

, et al. Factors correlated with volleyball spike velocity. Am J Sports Med 2005; 33: 1513–1519. Oct

45.

Marques

Van Den Tillaar

Gabbett

, et al. Physical fitness qualities of professional volleyball players: determination of positional differences. J Strength Cond Res 2009; 23: 1106–1111. Jul

46.

Laporta L, Nikolaidis P, Thomas L, Afonso J. Attack Coverage in High-Level Men’s Volleyball: Organization on the Edge of Chaos? J Hum Kinet. 2015; 47(1): 249–57.

47.

Laporta

Afonso

Valongo

, et al. Using social network analysis to assess play efficacy according to game patterns: a game-centred approach in high-level men’s volleyball. Int J Perform Anal Sport 2019; 19: 866–877.

48.

Hileno

Garcia de Alcaraz

Busca

, et al. What are the most widely used and effective attack coverage systems in men’s volleyball? J Hum Kinet 2018; 62: 111–118.