WyScout Open Access Dataset

该数据集包含了来自世界五大足球联赛的赛事数据，详细记录了球员的动作信息，包括动作类型、球场位置以及动作结果。此外，该数据集涵盖了21种动作类型和78种子类型，球员的动作以(x,y)坐标形式记录，并按区域分组以便于分析。这是目前可供研究使用的最大规模的足球赛事事件数据集，其任务是对足球赛事进行分析和预测。

This dataset comprises match event data from the world's top five football leagues, thoroughly documenting players' movement details including movement type, pitch location, and movement outcome. Additionally, the dataset covers 21 main movement categories and 78 sub-categories, with each player's movement recorded in (x,y) coordinate format and grouped by pitch regions to facilitate analysis. Currently, this is the largest-scale football match event dataset available for research, with its target tasks being football match analysis and prediction.

搜集汇总

背景与挑战

背景概述

该数据集来源于WyScout开放访问数据集，用于支持Transformer-Based Neural Marked Spatio Temporal Point Process (NMSTPP)模型的训练和测试，专注于足球比赛事件序列的分析和预测。