One of the fundamental topics in sports analytics is the science of success, i.e., the study of the correlation between players’ performances and their success. This is a very challenging task especially in the case of team sports, among which football is a prominent example. This paper is concerned with uncovering a dangerous bias that is present in most of the approaches proposed in the literature that apply statistical techniques or machine learning models to study the correlation between team performances and match outcome. In particular, we find out that players’ behavior on a time interval is more and more correlated with the match outcome as the 90 minutes elapse. As an extreme example, we show that we can predict the output of a match with high confidence simply by looking at the last 15 minutes of the game. We call this effect elapsing-time bias. We conduct a quantitative analysis that proves the existence of this phenomenon and shows its consequences. We then propose a novel way to address the problem. Namely, we design a new machine learning task that is not affected by elapsing-time bias. All the experiments are conducted on a large corpus of finely annotated football matches of European leagues.

体育分析的基本主题之一是成功的科学，即研究球员的表现和他们的成功之间的相关性。这是一项非常具有挑战性的任务，尤其是在团队运动的情况下，其中足球就是一个突出的例子。本文关注的是发现文献中提出的大多数方法中存在的危险偏见，这些方法应用统计技术或机器学习模型来研究团队表现和比赛结果之间的相关性。特别是，我们发现随着 90 分钟的过去，球员在某个时间间隔内的行为与比赛结果越来越相关。作为一个极端的例子，我们展示了只需查看比赛的最后 15 分钟，我们就可以高度自信地预测比赛的输出。我们称这种效应为经过时间偏差. 我们进行了定量分析，证明了这种现象的存在并显示了其后果。然后我们提出了一种新的方法来解决这个问题。也就是说，我们设计了一个不受时间偏差影响的新机器学习任务。所有的实验都是在一个大型的欧洲联赛足球比赛语料库上进行的。