## Sunday, November 06, 2011

### Using Sports To Teach Stats

There are lots of real-life examples to be used in stats classes, sports among them. However, I'm not sure that focusing only on sports is the best way to teach stats, primarily because there are plenty of people for whom sports is a meaningless topic:
At colleges and universities, however, baseball and other sports have become fertile grounds for teaching formal statistics to otherwise wary undergraduates. Pascal’s triangle helps explain why the favored team often loses the World Series. Expected values develop rather unexpected value when they unmask football coaching blunders. And when it comes to Bell curves, which would you rather plot — your school’s N.C.A.A. basketball scores, or snowfalls in Saskatoon?
If your course is "Sports Stats", that's great. But for a general course I see two problems with this. First, not everyone enjoys sports, and second, there are so many other topics that can demonstrate the applicability of stats.

“Intro Stat as a course is frustrating because students don’t care about the things you talk about — economics, politics — but everyone can relate to sports in some way, either as participant or fan,” says Jim Albert, author of the 2003 textbookTeaching Statistics Using Baseball” and a professor of statistics at Bowling Green State, in Ohio. “The students are interested, and that’s all we’re looking for. They’re willing.”
I disagree. Personally, I don't even like baseball much. It's one thing to be there in the park enjoying a game, but I can't watch it on tv or read the paper about it or anything. And there are plenty of people out there who like economics and/or politics.
Flipping a coin twice is a mundane thought experiment; having Steve Nash of the Phoenix Suns attempt two free throws with a game on the line is like watching ESPN in the classroom.
Can't stand basketball at all. And everyone can understand a coin flip.

Again, my point isn't to trash this idea--it certainly appeals to a certain segment of the population. But if you want a course that's universally applicable, you should probably include more than one topic. Why avoid weather, medicine, investing, gambling, psychology, test scores, manufacturing, "bar tricks", equal opportunity/discrimination challenges, weaponry, genetics, global warming (hehe), opinion polling, food production, red-light waiting time, and zillions of other topics, and focus only on sports?

American football is perhaps the greatest sport ever invented. Before my son was born I would spend all weekend on the couch, watching every game I could. I would run numbers through my brain constantly, predicting outcomes. Sometimes I'd even "reverse engineer" games, looking at game stats and trying to predict the finals scores of the games from them (it was more art than science, at least consciously, but I got to be very accurate at it). Fatherhood, though--especially single fatherhood--has reduced my football down to the point where I don't even know who's quarterbacking the Raiders today against Denver.

My point is that while I see the utility of using sports examples, and can even understand why many would take a "Math of Sports" class (I might enjoy a Math of Football class), such can never be more than a niche class.

#### 1 comment:

Left Coast Ref said...

As a sports guy, I understand the logic of having these courses available, but they should by no means replace Intro Stats. I know WAY more people than you would think that have absolutely zero interest in ANY sports, but they want to understand patterns, analyze data and other topics integral to stats. "Math of Sports" as you call it could be a Math elective geared towards Jocks and "Box Score Geeks".