摘要
以入选2014年巴西世界杯决赛圈的736名球员为研究对象,运用卡方检验,发现球员出生日期存在显著的相对年龄效应。进一步利用排序选择模型考察了相对年龄效应对比赛成绩的影响,实证结果表明,上场球员的相对年龄均值越大,球队获胜的概率越小,这样的证据不足以从球队角度反对按照相对年龄优势选拔球员的做法。另一个反映相对年龄构成的变量,主队四季度与客队一季度出场球员占大名单比例的差,其系数为正,说明直接对位的球员中,主队四季度出场的球员更多时,球队获胜的概率越大,这在一定程度上抵消了相对年龄均值对比赛成绩的负作用。同时,模型成功预测了64场比赛中43场的结果,准确率达到67%,模型的预测功能表现良好。
We take the 736 players of 2014 Brazil World Cup finals for research objects, and find the relative age effect by using of Chi-square test. Then, we adopt the ordered choice model to investigate the relative age effect further. The empirical results show that the greater the relative average age, the smaller the probability of victory. Meanwhile, we believe another variable which is the difference accounted by the home team players appearing who born in the fourth quarter and the visiting team players appearing who born in the first quarter between the squad players, is closely associated with the match result. When there are more players of home team born in the fourth quarter, the parameter becomes positive, that means the probability of victory is greater. In this case, it partly offsets the negative effect by the relative average age. The ordered choice model's prediction ability is very good. We successfully predicted 43 games" outcomes in 64 games with an accuracy of 67% by the use of this model.
出处
《统计与信息论坛》
CSSCI
北大核心
2015年第2期88-94,共7页
Journal of Statistics and Information