Performance-related variables for succeeding in the UEFA Champions League What performance-related variables best differentiate between eliminated and qualified teams for the knockout phase of UEFA Champions League? Carlos H. Almeida 1 1 CIDEF, Instituto Superior Manuel Teixeira Gomes, Portim˜ ao, Portugal Association football | Match analysis | Team performance | Competitive success Headline T he UEFA Champions League is the most prestigious foot- ball club competition worldwide (1). Reaching the knock- out phase is paramount for participant clubs to warrant more revenues and international reputation. Although performance profiles vary among teams (2, 3), match events as total at- tempts, shots on target and passing accuracy have been pre- viously associated with success in this tournament (1, 3, 4). However, since other performance-related variables (e.g., ball possession or yellow cards) displayed contradictory results (3, 4), further research is required to clarify the contemporary trends of successful teams’ performance in elite football (5, 6). Aim. This study aimed to identify the team performance vari- ables of total attempts, shots on target, goalkeeper’s (Gk) saves, ball possession, completed passes, passing accuracy, dis- tance covered, balls recovered, fouls committed, yellow and red cards that best differentiate between eliminated and qual- ified teams for the knockout phase of 2017/2018 Champions League. Methods Sample. The sample consisted of all matches (n = 98) con- tested in the 32-team group stage of UEFA Champions League (season 2017/2018). At this stage clubs are divided into eight groups of four teams, playing a total of six matches against their pool opponents (1). Performance-related statis- tics of each match report were derived from the official web- site of UEFA (www.uefa.com) considering both teams. The study conformed to the recommendations of the Declaration of Helsinki. Design. A retrospective observational study was conducted with a comparative data analysis. Matches were contested between September and December of 2017, whereas data col- lection and analysis were undertaken between February and March of 2018. Methodology. For the independent variable ‘group stage out- come’, two groups of teams were created based on the final classifications of group stage (0: eliminated teams; 1: qual- ified teams). The two top teams of each group qualified for the knockout phase (n = 16). The third and fourth-place fin- ishers were eliminated from the UEFA Champions League (n = 16). In line with previous research (1, 3), 11 performance- related statistics were selected for analysis: (1) variables re- lated to goal-scoring: total attempts and shots on target; (2) variables related to passing and organising: ball possession (%), completed passes and passing accuracy (%); (3) vari- ables related to defending: Gk saves, balls recovered, fouls committed, yellow and red cards; (4) variable related to phys- ical performance: distance covered. The dataset was prepared on a Microsoft Excel sheet containing the calculated means of selected performance-related variables for each team. Statistical Analysis Non-clinical magnitude-based inferences were employed using a spreadsheet arranged by Hopkins (7). Effects of group stage outcome (eliminated vs. qualified) on performance-related variables were evaluated by using the smallest worthwhile change (SWC) - that is - 0.2 multiplied by the pooled between- group standard deviations (SD), in accordance with the Co- hen’s d principle. To reduce bias arising from non-uniformity error, log-transformed values were used to calculate effect sizes (ES) based on standardised mean differences with 90% confi- dence intervals (CI) (8). Magnitudes of clear effects were de- scribed according to the following scale: 25-75%, possible; 75- 95%, likely; 95-99.5%, very likely; >99.5%, almost certain (9). If the probabilities of the effect being substantially positive (or beneficial) and negative (or harmful) were both >5%, the effect was deemed unclear (8). Standardised differences were classified as trivial (<0.2), small (0.2-0.6), moderate (0.6-1.2) and large (>1.2) (9). Results Despite the effects presented in table 1, when it comes to dif- ferences between-group means analysed in standardised units (with 90% CI) considering the SWC (Figure 1), the team per- formance variables that best differentiated between eliminated and qualified teams for the knockout phase were shots on tar- get, total attempts, ball possession (almost certainly bene- ficial) and completed passes (very likely beneficial). Whilst passing accuracy was likely beneficial, likely harmful effects were found for Gk saves, fouls committed, yellow and red cards. The other differences were unclear. Discussion The comparative analysis of team performance profiles associ- ated with success may reveal which variables are more deter- minant in contemporary elite football (3, 6, 10). Particularly in the group stage of 2017/2018 UEFA Champions League, the team performance variables that best differentiated be- tween eliminated and qualified teams for the knockout phase were shots on target, total attempts, ball possession and com- pleted passes. These results support previous findings high- lighting the goal-scoring variables shots on target and total at- tempts (shots) as major predictors of success in UEFA Cham- pions League (1, 3, 4) and in the last three editions of FIFA World Cup (10, 11). Also in line with a couple of studies sportperfsci.com 1 SPSR - 2018 | Apr | 23 | v1