Identifying the relationship between football match performance indicators and match outcome by mathematical models ,and moving the research of football performance analysis from a comparative and descriptive level to a predictive one .The current study employed the statistical methods of General Linear Model and Multinomial Logistic Regression to explore and identify the winning determinant performance indicators in matches of teams of different levels playing against opponents of different levels in the Chinese Football Association Super League (CSL ) .Match data of all the 240 games of CSL season 2014 were analyzed .The results showed that 1 ) For high‐level teams ,when they played against high‐level oppositions ,per‐formance indicators that determined the match outcome including shot on target ,through ball and red card ;while in the games which they played with low‐level teams ,shot ,shot on target , pass accuracy and tackle are indicators that were significantly related to winning .2 ) For low‐level teams ,when they played against high‐level opponents ,shot ,shot on target ,shot off tar‐get ,cross ,corner ,ball possession ,offside ,tackle ,foul and red card were performance indicators significantly related to winning ;when they played against low‐level teams ,the winning deter‐minant performance indicators included merely shot on target and cross .%通过数学模型界定足球比赛技、战术指标与比赛结果的因果关系，可以使足球比赛表现分析从描述性和对比性研究层面跨越到预测性研究层面。以2014赛季中国足球超级联赛的240场比赛为样本，采用一般线性模型和多元逻辑回归模型，对不同实力参赛球队对阵不同实力对手比赛中决定比赛胜负的关键技、战术指标进行探析和界定。研究结果显示：1）对上游球队而言，在对阵上游球队时，射正球门、直塞球和红牌为决定比赛胜负的显著性关键变量，而在对阵下游球队时，决定比赛结果的显著性关键变量则包括总射门、射正球门、传球成功率和抢断；2）对下游球队而言，当对阵上游球队时，总射门、射正球门、射门偏出、传中、角球、控球率、越位、抢断、犯规和红牌都与比赛结果呈现显著性相关关系，而当对阵下游球队时，决定比赛结果的显著性关键变量只有射正球门和传中。
展开▼
