Football Science VII: the International Conference in Japan, 2011 Martin Buchheit1, Will G Hopkins2 Sportscience 15, 9-14, 2011 (sportsci.org/2011/mbwgh.htm) 1
Physiology Unit, Aspire Academy for Sports Excellence, Doha, Qatar; Email. 2 Sport Performance Research
institute NZ, AUT University, Auckland, New Zealand; Email. Reviewer: Aaron Coutts, Leisure
Sport & Tourism, University of Technology, Sydney, Australia.
|
|
Staff from the Nagoya University proved friendly and professional in their hosting of the World Congress on Science and Football in Japan, the seventh in the series. Among the 350 or so attendees were some of the most influential researchers in the various football codes. As always, the congress was an excellent opportunity to share recent work and ideas. We're all looking forward to next year's World Conference on Science and Soccer in Ghent, Belgium! The conference organizers have very generously uploaded videos of all the symposia to YouTube. To view each of the 41 invited presentations, go to the end of this article, where you will find a list we have copied from the official program and linked to each video. In this report we have referred to all European football as soccer. The summary of each presentation includes the names of the first one or two authors and the first few words of the title of the abstract, to allow you to find it when the abstracts are eventually published. You can also get a scanned copy of selected abstracts from author MB (who attended the conference) or author WGH (who did not). Talent DevelopmentA few presentations focused on the relative age effect–the over representation of players born in the first half of the year in teams. While the effect occurs in all sports, the impact on anthropometric and performance measures is still unclear [Leite, Relative Age Effect in Football…; Simpson, Is the Relative Age Effect Influenced...; Fernley, Relative Age Effects in Elite Junior…]. A presentation on the application of science to developing young players in an English premier league soccer club listed all the usual sensible strategies [Iga, The Application of Science…]. The training program takes into account age of peak height velocity and prediction of adult stature. The theoretical background is seducing, but how it is implemented on a daily basis is another question. In support to the concept of individualizing training based on growth and maturation, Mendez-Villanueva presented data on spontaneous periods of accelerated improvement in physical fitness in youth soccer players [Physical Fitness Trainability…]. Soccer-specific training tends to improve aerobic fitness in pre-pubertal boys, whereas more mature players show improvements in neuromuscular-related fitness. The changes in sprint performance with growth in young soccer players are associated with different stride patterns: the younger improve more their performance via stride length, while the older, via stride rate [Poon, Age-related Differences in Sprint Performance…]. So what should we train: the parameter that is already improving naturally or the other parameter? Age-related differences in repeated-sprint ability parallel differences in several measures of fitness in soccer [Spencer, Fitness Determinants of Repeated-sprint…]. To optimize the development of repeated-sprint ability, the author suggested that training could therefore focus on different aspects of fitness at different ages. This strategy still lacks evidence, however. After surveying the available evidence, Dawson concluded in a rather cool presentation that performance in tests at rookie draft camps is not a consistent predictor of future career success in various team sports [Draft Camps and Combines…]. Training and MonitoringPlayers for the Australian-rules (AFL) football club Essendon benefit from a "system of load management and fatigue monitoring” using psychological, hormonal and neuromuscular measures, all of which contribute to the overall picture of the players’ shape. [Cormack, Load and Fatigue Monitoring…]. The interesting statistical approach is based on a moving average of each measure in relation to its smallest important change. On similar lines, Buchheit [Monitoring Changes in Performance…] showed that decreases in exercise heart rate and increases in heart-rate variability can track improvements in maximal aerobic speed in young soccer players, yet opposite changes in these indices are not predictive of performance impairment. Perceived exertion in training sessions was also touted as a useful monitoring approach in rugby, in spite of concerns that the impacts in this sport would compromise the accuracy of the measure [Coutts, Efficacy of Using Session-RPE…]. Examples of sport-science support to coaches and players were shown for elite rugby union [Gill, Current Approaches to Sport Science… (the abstract has no useful information)] and soccer [Dijk, Training and Performance Management…]. While everything tends to get monitored, controlled and measured by an increasing number of staff in the Bayern München soccer club, the All Blacks are going in the opposite direction with much stronger relationships and communication between a reduced (but more involved) number of staff. The All Blacks almost always win (the World Cup excepted, but hopefully not this year), so which approach is better? FIFA's injury-prevention training program ("The 11") and a modified version of it performed one or twice a week for 6 months reduced injuries and improved some aspects of fitness in comparison with control training in a trial of 182 Japanese collegiate soccer payers [Saho et al., The Efficacy of The 11…]. The importance of high-intensity aerobic training to develop aerobic fitness in young soccer players was emphasized by Mujika. He presented an interesting (but not new) case study of a young soccer player who managed to improve Yo-Yo performance by 32% in 10 high-intensity sessions [Practical Aspects of High Intensity Training…]. In a randomized controlled trial of 34 male soccer players (age 12-14), 10 wk of training of complex skills produced greater improvements in several test measures than the control traditional training [Ashker & Abdelkader, Impact of Developing Simplex…] In Aussie-rules football and rugby league, Coutts has found that "subtle changes in training load, different periodization strategies and the duration between matches can affect player recovery and subsequent performance" and that "72-h period following matches is critically important for ensuring appropriate recovery for the next game". [Training Factors Affecting…]. He recommends a system for planning and monitoring training at the level of team and individual player. This was the best presentation of all, in the opinion of the author of this article (MB) who attended the conference. A soccer player was trained to increase (by an impressive 3.9 m) the distance he could throw the ball in run-up throw-ins with a staggered stance [Glynn & Kuitunen, Using Motion Analysis…]. More training and research should be done in his area. Unfortunately the abstract reports only on the changes in the player's kinematics, not on how the training was done. A study of the impact of sleep deprivation and alcohol consumption on post-match recovery in rugby league was presented, and guess what? It's not good for your peak power the following day. You knew it, but now you have evidence to show your players that they should go home after the game! [Duffield, Sleep Deprivation and Alcohol…]. Match and Movement AnalysisIn a study of zone-oriented defense, Tenga used a case-control analysis to compare 203 possessions that resulted in goals with 1688 randomly chosen possessions that didn't. He found that imbalanced defensive pressure had little effect, but imbalanced backup and cover were 7.5 and 48 times more likely to result in a conceded goal in the Norwegian soccer league [Zone-oriented Defense…]. This original-research study is the first choice of the absentee author (WGH). The performance physiologist of Bayern München identified four game phases in soccer: own team in possession, losing possession, opposition in possession, and regaining possession [Dijk, Training and Performance Management…]. Technical, tactical, physical and mental requirements of each player in each phase depend on the game plan and the position of the player. He uses the LPM Soccer 3D System to monitor the players’ activity patterns and advanced video-computerized systems to follow the interactions between the players. "Science and innovations are important… to keep the winning edge." Now that technologies are available to track individual players at high spatial and temporal resolution in a match, there are new possibilities for analyzing tactical interactions in soccer, such as symmetry breaking processes in player dyads and attacking/defensive team configurations [Lemmink and Frenkin, Tactical Match Analysis…; see also Dijk's presentation above]. Dynamical systems theory was promoted as a relevant framework with tools that can cope with such data. Duarte et al. have made a start by deriving "collective motion variables" representing various aspects of playing patterns in a single match [In Search of Dynamical Patterns…]. Xie et al. presented a similar limited case study of a "defensive strength map" based on distance between attackers and defenders [A Study on Quantitative Evaluation…]. Some of the information is lost in translation, but Taki & Hasegawa seem to be onto something interesting in their use of time and velocity data to map a soccer player's "dominant region": a region where the player can arrive at earlier than any other players [A Method for Game Analysis…]. When somehow aggregated over all players in each team, there was an obvious pattern favoring the winning team in the one game that was analyzed. In their next abstract they somehow made geometric shapes out of the players' positions, but the method and the result are inscrutable [An Evaluation Method…]. It would great if this approach could be used in real-time or near real-time to provide feedback on tactical strategies. An analysis of movements of 37 elite rugby league players recorded with GPS units in 104 game appearances revealed position-specific movements that should have implications for training [Gabbett et al., Physiological Demands of Professional…]. Two controversial analyses of acceleration patterns in soccer were presented. While agreeing with the general concept, a colleague (Coutts) questioned whether the technology is currently accurate enough, classifications of games and training drills could be made according to the occurrence of acceleration phases [Varley , Accelerations in Football…]. Interestingly, teams are much more likely to win when they show more acceleration phases than the opponent team [Varley, Match Score Influence…]. It is surprising that "despite the strong influence simulation (taking a dive) can have on the outcome of soccer matches, no studies have attempted to quantify and identify factors that can promote or limit this behavior" [David et al., Fighting Back Against Diving…]. The authors did not state the proportion of dives in the 2803 falls they analyzed, but they found a higher proportion closer to the attacking goal, and such falls were more likely to be "rewarded" by the referee. Only 0.2% of dives were punished with a yellow card. No wonder some people don't bother with soccer. Not surprisingly, penalty kicks were less successful when the goalkeeper moved in the right direction in an analysis of 286 penalties in European soccer leagues, but there were no other useful predictors of success [White & O'Donoghue, Factor Affecting Performance…]. A stop-motion lab study of the eye tracking and initial movement of nine experienced goal-keepers confirmed that deception was a successful strategy for penalty kickers but found no useful relationship with eye-tracking of the goalie [Tay, Investigation of the Effectiveness…]. Ball possession is a predictor of winning in English and Spanish soccer leagues, but the opposite obtains in the Italian and some other leagues, and there is no consistent relationship in international games [Collett, The Possession Game…]. Game-winning teams in the soccer World Cup in South Africa on average scored 2.1–0.5, took ~25% more shots, had ~30% more of the shots on target, were in possession 4% more of total match time, had practically no red cards compared with their opponents' 0.3, made 18% more successful passes, and made 50% more successful crosses [Casamichana & Castellano, Differences between Winning…]. These were the statistically significant outcomes of the 64 matches, which is not nearly enough for acceptably low error rates (false detection and failed detection of effects) in this kind of observational study. Useful findings, nevertheless. Tests and TechnologyAnthropometric and performance characteristics of elite Australian rules (AFL) football players were also described. Not that surprisingly, the pros are older, taller and heavier than amateurs, and whole body and segmental lean mass had moderate to large correlations with bench-press, bench-pull and jump-squat performance measures [Bilsborough, Anthropometrical and Performance Characteristics…]. Bangsbo and others again sang the praises of the Yo-yo intermittent recovery test to assess endurance fitness in soccer players [Fitness Testing and Training…]. The Yo-Yo test has limitations in identifying intensities for training prescription, and such tests do not predict running activity in games; see a presentation by MB for more. In an original-research presentation, Mooney et al. identified a positive association between performance in the Yo-yo intermittent recovery test (Level 2) and ball disposals in Australian football players [The Relationship between Physical…]. The kind of multivariate analysis that Wilson et al. applied to 17 skill tasks in semi-professional and elite soccer academy players is probably not appropriate when the sample size was only 110 [Quantifying and Analyzing Skill…]: there are likely to be too many spurious effects. It's also unclear how "overall football ability" was defined. The differences between fast and slow professional soccer players are related to differences in mechanical muscle measures (explosive strength, maximal isokinetic torque), but locomotor differences (technique, coordination) are also important [Palazzi, Muscle Mechanical Determinants…]. So it's not just a matter of developing muscle strength and power. A new method to monitor lower limb muscle flexibility was presented. While usual methods rely on players' tolerance to stretch, which can bias the evaluation, this method is based on the application of a standard torque. The method was shown to be reliable and independent of the assessor, so any skilled staff in a soccer club or academy can make the measurements [Materne, A Novel Method to Monitor…] When you are evaluating performance indicators derived from match analysis, adjusting for the relative strengths of the teams is a good idea [Cullinane & O'Donoghue, Addressing Opposition Quality in Rugby League…]. But using the ranking of the teams at the end of the season is not viable if you want to assess players during the season, and adjustment that allows for non-linear effects with increasing disparity between team strengths may be more appropriate. What sample size (players and matches) will you need to quantify the effects of a treatment on a performance indicator in team games? It depends on the between-player differences and game-to-game within-player variability of the indicator. For the example indicator chosen (high-speed running in English Premier League, as coded by Prozone), it's 122/N players in N matches pre and post a crossover, or four times as many in a parallel-groups trial, for 90% power and 5% significance [Gregson et al., Estimating Sample Size…]. Divide the numbers by 4 when using Type 1 and 2 clinical errors of 0.5% and 25% with sample size for magnitude-based inferences. In a qualitative study of coaches from different football codes and levels in six focus groups, there was a wide range in perceptions of the utility of sport science: "Developmental rugby league coaches and elite soccer coaches had least use for sport research, whereas novice rugby league and developmental soccer coaches were enthusiastic and motivated to use it in practical coaching sessions" [Nash & Martindale, The Usefulness of Sports Science…]. It's hilarious and amazing that the red team wins more matches (about 1 in 10) even in a video soccer game [Ohshimo et al., Uniform Color…]. The study needs more than six presumably Japanese players for the finding to be credible and creditable. What aspects of personality or cognition are responsible? Acknowledgements: The Aspire Academy funded MB to
attend the conference and WGH to collaborate with MB. Both authors also
received institutional salary support. Published July 2011
|