so i was reading the blog of the owner of my beloved dallas mavericks, mark cuban and one of his posts uses statistical analysis to prove that one of the league officials statistically has an effect on the outcome of games. he quotes some minnesota timberwolves' fan's data analysis of the outcome of games when joey crawford is one of the officials, he quotes:
> Outcome 1 Outcome 2 Total
> Group 1 7 1 8
> Group 2 0 3 3
> Total 7 4 11
> Fisher’s exact test
> The two-tailed P value equals 0.0242
> The association between rows (groups) and columns (outcomes)
> is considered to be statistically significant.
granted he's quoting someone else, and granted he's crazy, i just thought that i'd use this as a perfect example of how statistics are so misused in sports. anyone who know's anything about basic statistics can tell you, not enough data. you really can't make any good inferences based on three freakin observations. but the worst thing of all is that i checked the comments and only one person commented on the problem with using this as evidence to support the argument.
i din't mean for this post to do this, but this gives me a perfect opportunity to announce to all five of you who read this blog, that i officially got the teaching gig for this summer. i'll be teaching statistics for the social sciences M thru F from 9:30AM - 10:45AM. it's a crazy thought to think that i will be responsible for providing young men and women with the tools that they need to read the stupid USAToday graphics, but the soc department has given me the keys to their car. i only hope that i don't drive it into a ditch. in any case, you can bet that i'll be using sports a whole lot to make my point about statistics. also, if anyone has any suggestions as to how i might word the "sex for grades" policy for my syllabus, i'd be eternally grateful.
only P1's will get this next joke, but if i could somehow work it where i gave extra credit to some hot chick for sleeping with me, in effect, i'd be giving her points for sex...baby arm!