Tactical Analysis Through Objective Data in Football
Tactical Analysis Through Objective Data in Football
复制标题
通过足球客观数据进行战术分析
DOI：
10.18282/iss.v2i1.257
复制
发表时间：
2020
期刊：
影响因子：
--
通讯作者：
M. Sugiyama
中科院分区：
文献类型：
--
作者：
A. Yamanaka;H. Otsuka;T. Deguchi;K. Okihara;Dota Otsuka;Shinya Uchida;Jorge Diaz;S. Khoo;D. Stroud;M. Sugiyama
足球中的战术分析很少使用带有数字数据的数学模型(尽管通过客观数据进行战术分析更多地被使用)。因此，本研究通过建立基于球员和球的位置的数学模型，建立了利用数字数据进行团队运动战术分析的原则。拍摄了2011年日本天皇杯第三轮广岛队(主场)和爱美队(客场)之间的一场竞争性比赛，并将其用于比赛分析。观察是由一名球队分析师进行的，并摘自官方比赛记录。研究流程中的主要程序是根据足球中的战术概念建立一个数学进攻/防守模型，用于确定球员的位置，然后从视频图像中进行量化，以便对球队的战术表现(与进攻或防守有关)进行分类。此外，作者将重点放在进攻类别上，并确定了特定时期内不同类型的传球，并将这些发现与实际比赛视频进行了比较。从应用进攻/防守模型得出的数字数据得出的结果与团队分析员所作的战术分析结果相同。此外，对传球类型进行分类时的结果(通过数学模型提取)再次反映了从实际比赛视频中检索到的结果。这导致了这样的结论，即进攻/防御模式可以提供对攻击类型的相关洞察。数据显示，可以使用数值模型成功地执行足球战术分析，这可能使足球比赛的自动战术分析不需要比赛分析员。
Tactical analysis in football has rarely been conducted using a mathematical model with numerical data (although tactical analysis through objective data has been used more often). Therefore, this study establishes principles for tactical analysis in team sports using numerical data, through a mathematical model based on the location of the players and the ball. A competitive match between Sanfrecce Hiroshima (home) and Ehime FC (away) in the third round of the Japanese Emperor’s Cup 2011 was filmed and used for the match analysis. Observations were made by a team analyst as well as extracted from official match records. The main procedure in the research flow was to establish a mathematical offence/defence model based on tactical concepts in football, which was applied for the location of players, which, in turn, was quantified from video images in order to categorise a team’s tactical performance (in relation to attacking or defending). Furthermore, the authors focused on attacking categories and identified different types of passes during a specific period, as well as comparing these findings with an actual match video. The results obtained from the numerical data derived from applying the offence/defence model led to the same overview as the tactical analysis produced by a team analyst. In addition, the results when categorising types of passes (as extracted through the mathematical model) again mirrored those retrieved from an actual match video. This leads to the conclusion that the offence/defence model could provide relevant insight into types of attacks. The data revealed that football tactical analysis can be successfully performed using a numerical model, which might possibly enable automatic tactical analysis of football games without a match analyst.