Spss for windows statistical software package version 22

The Kolmogorov–Smirnov (general results) revealed that match running performance data were not normally distributed for some variables (p < 0.05). Thus, to avoid textual confusion, all the data are described by the median (interquartile range). The comparisons of running outputs between bottom- and top-ranked teams (general results) were assessed using the Mann–Whitney test. The ANOVA two-way was used to compare the interactive effects of independent measures on running performance: (1) playing position according to ranking-teams; (2) match location according to ranking-teams; (3) quality of opposition according to team’s ranking; (4) match outcome according to ranking-teams. The significance level was set at p < 0.05. Data were analyzed using the SPSS for Windows statistical software package version 22.0 (SPSS Inc., Chicago, IL, USA). Additionally, effect sizes (ES) for non-parametric data (general results) were calculated for pairwise comparisons (ES = z/√n) and classified as negligible (< 0.1), small (0.1–0.29), medium (0.3–0.49), and large (> 0.5) [27 (link)]. The ES for parametric data (playing positions, match location, quality of opponents, match outcome) were assessed using partial eta squared (η2), and classified as: > 0.01(small), > 0.06 (moderate), and > 0.15 (large) [28 ].

All the statistical analyses were conducted using SPSS for Windows statistical software package version 22.0 (SPSS Inc., Chicago, IL, USA). Initially, descriptive statistics were used to describe and characterize the sample. Shapiro–Wilk and Levene’s tests were conducted to determine normality and homoscedasticity, respectively. One-way ANOVA was used with Scheffe’s post-hoc method. One-way analyses of variance were used to compare all the dependent variables (external match load measures) across the playing positions. Student’s t-test was also used to compare data by match location (home/away contextual factor) and championship phase (first and second phases). The effect size with 95% confidence interval (ES 95% CI) statistic was calculated to determine the magnitude of effects. Furthermore, Hopkins’ thresholds for the effect size statistics were used as follows: ≤0.2, trivial; >0.2, small; >0.6, moderate; >1.2, large; >2.0, very large; and >4.0, nearly perfect [33 (link)]. Alpha was set at p ≤ 0.05.

Data were analysed using the SPSS for Windows statistical software package version 22.0 (SPSS Inc., Chicago, IL, USA). Initially descriptive statistics were used to describe and characterize the sample. Shapiro-Wilk and Mauchly’s tests were used to verify the assumption normality and sphericity, respectively. Repeated measures ANOVA was used with Bonferroni post hoc, once variables obtained normal distribution (Shapiro-Wilk > 0.05) and it was used ANOVA Friedman and Mann-Whitney tests for the variables that not obtained normal, to compare different match results. Also, paired sample t-test was used to compare data from first half with second half according to the final result of the match. Results were significant with p ≤ 0.05. The effect-size (ES) statistic was calculated to determine the magnitude of effects by standardizing the coefficients according to the appropriate between subject’s standard deviation and was assessed using the following criteria: <0.2 = trivial, 0.2 to 0.6 = small effect, 0.6 to 1.2 = moderate effect, 1.2 to 2.0 = large effect and >2.0 = very large [24 ].