There are two researchers called Stephen Dobson and John Goddard concerning a interesting question of sports fans in 2003. In academic words, it is about the nature of persistence in sequences of consecutive match results. In non-academic words, the questions are:
(在2003年, 有两位研究者称为Dobson和约翰戈达德处理一个体育迷有趣的问题的。在学术上,它的性质是持久性的序列的连续比赛的结果。用非学术话来说,问题是:)
1. Does a sequences of wins tend to build a team's confidence and morale, increasing the probability the the next match will also be won?; or Does it tend to create pressures or breed complacency, increasing the likelihood that the next match will be drawn or lost?(难道一个胜方序列倾向于建立一个团队的信心和士气, 这增加下一场比赛也将赢了的概率?;还是会倾向于建立压力或滋生自满情绪,越来越多的可能性,下一场比赛将是失败呢?)
2. Does a sequence of losses tend to sap confidence or morale, increasing the probability of a further loss in the next match? Or does it tend to inspire greater effort, increasing the likelihood that the next match will be won or drawn?(难道一个失败序列往往会消磨其意志和士气,增加下一场比赛进一步损失的概率?还是倾向于激发更大的努力,增加下一场比赛是赢得或平手的可能性呢?)
Dobson and Goddard has analyse 30 years of match data for football from the English Premier League and Football League. They have setup an hypothesis based on a computational test of Monte Carlo analysis. The assumption behind the Monte Carlo simulation is parameter constancy and zero persistence. The result is then compared with the real results. (Dobson和戈达德用了从英超联赛和足球联赛30年的对足球比赛分析数据。他们设置一个假设的基础上计算测试 -- 蒙特卡罗分析。假设背后的蒙特卡罗模拟参数稳定性和持久性为零。然后比较其结果与实际结果。)
They got an interesting observation that the actual probability a reversal occurs is higher than the simulated probability under assumptions of no persistence. What does it mean in non-academic wordings? They summed up in the statement "Empirically, the conditional probabilities of a good result are found to decline with the duration of a poor spell, and the conditional probabilities of a poor result decline with the duration of a good spell." (他们得到了一个有趣的观察,实际发生的"反转概率"是高于在假设不持久性的条件下的模拟概率。在非学术字眼,这是什么意思?他们在声明中总结了“根据经验,一个很好的结果的条件概率是随着一个负面的时间下降的,和一个负面结果的条件概率会随着一个良好的时间下降。”)
Reference:
Dobson and Goddard (2003), "Persistence in sequences of football match results: A monte Carlo analysis", European Journal of Operational Reseaarch, 148, 247-256
No comments:
Post a Comment