1 利用SAS对各个赛季的比分作描述性统计.DOC

对NBA比赛比分和猜中比分概率的统计分析 华南师范大学 李琪、刘冠超、庄合芳 摘要：现今我国的彩票业蒸蒸日上，体育彩票更是蓬勃发展。特别足球单场竞猜异常火爆，反观全国联网篮彩的发展却一波三折，挽救低迷中的篮彩已经是迫在眉睫。而篮彩单场竞猜的出现让篮彩看到了一线生机，故本文首先利用描述性统计对NBA 06-10四个赛季的主客队比分作出了客观分析，提议将篮彩的竞猜内容改为NBA单场比赛主客队的比分，再根据描述性统计的结果对06-10四个赛季的数据进行分组并建立了二维正态分布概率模型，从而估计出最大的中奖概率，最后通过仿真模型和实证分析检验该概率的合理性，为篮彩设置比分竞猜提供有益的参考。 关键词： The Analysis of the NBA Scores and the Proper Score-guessing Probability of NBA Match Abstract: With the rapid development of lottery industry in China, the sports lottery, especially Football Score Quiz, has become more and more popular in the public. Under this situation, the basketball lottery, however, is not as welcomed as it should be for the unreasonable setting. Thus we need to calculate a Largest Probability of proper guessing of the score to develop a new game about NBA Score Quiz to help the recovery of basketball lottery. In this article, the score data of 06-10 NBA seasons is being analyzed by the Descriptive Statistics Method. Accordingly, we then divide the NBA teams into different groups in different seasons in order to form Two-dimension Normal Distributions for the score of a match between teams from different groups and obtain the Largest Probability of proper guessing of the score of an NBA match. Finally, to prove the correctness of that Largest Probability, a Simulation Model is built up and an Empirical Analysis is been used. Keywords: Descriptive Statistics, Variance Analysis, Two-dimension Normal Distribution，Bootstrap Resampling Method, Simulation Model, Empirical Analysis 目录 1 引言 1 2 数据整理 1 3 基本模型的建立 3 3.1 建立二维正态模型 3 3.2 模型检验 4 3.2.1 比较每个得分区间的得分百分比 4 3.2.2 比较需要打加时赛的概率 6 3.3 估计实际中奖概率 6 4 四个赛季的比较分析 7 4.1 平均得分对比 9 4.1.1 无加时 9 4.1.2 有加时 10 4.2 标准差对比 11 4.2.1 无加时 11 4.2.2 有加时 12 4.3 主客队的得分频率分布对比 13 4.4 实际加时情况对比 15 4.5 实际中奖概率对比 15 5 仿真 17 5.1 仿真的目的 17 5.2 仿真的方法 17 5.2.1 模型假设 17 5.2.2 假设的合理性 17 5.2.3 仿真的具体步骤 18 5.3 仿真结果分析（以06-07赛季为例） 20 6 实证分析 22 6.1 前期准备 22 6.2 理论分析 23 6.3 调查分析 23 6.4 结果与讨论 25 7 讨论与结论 26 附录 28 参考文献 37 致谢 38 引言 博彩业自古以来便深受人们追捧，无论出自什么