Monday, I talked about decisions. I really enjoy the decision-making process. And not to beat a dead horse, but I want to continue talking about decisions. Oh, and if I was not clear on Monday, I am going to try and post on Monday-Wednesday-Friday. We will see if this schedule is more conducive.
Looks like it will be Ohio State and LSU playing for the national championship in January. The big BCS super computer algorithm thingie pumps out a couple of names and they play for the national championship. But it is not the best way – I mean, the ranking of teams is not the best way.
For such a complex analysis (20 to 30 teams that legitimately compete for the national championship, and most of them don't play but a few of the teams each year), a rank analysis is not the best approach to take, but we do it anyway. A better approach is to use the decision making technique known as paired comparison analysis.
Let me explain about paired comparison analysis, not using football.
Let's say we want to rank hot women. We have four contestants: Shakira, ~Deb, Britney Spears and Drew Barrymore. Our goal here is to rank these ladies from one to four in terms of sexiness. This example simplifies what I am writing about – with only four women, ranking may be easy enough because some people perform paired comparison analysis with small sets of data anyway. But I am just trying to illustrate the point.
One looks at Shakira and ~Deb, first, then asks the question, "which woman is hotter?" Winner gets a point. Let's say the answer is Shakira. After all, Shakira is the only women I know that have lots of men and women lusting after her. I wrote about it a long time ago. Shakira and ~Deb are both sexy and smart, but the edge goes to Shakira. Shakira also beats Drew Barrymore – she has a sexier voice, can dance better and is way smarter. Also, Drew looses points because of her association with Cameron Diaz, a woman who just looks goofy, while dancing or otherwise. I really don't know why men find Ms. Diaz attractive. Well, all beat Britney Spears. Some hypothesize that Al-Qaeda is trying to obtain a pap smear from Britney in their efforts to obtain a new biological weapon. Not saying it is true. Just a rumor at this point.
So the scoring may be as follows:
Shakira – 3 points
~Deb – 2 points
Britney – 0 points
Drew – 1 point
You rearrange to get the ranking: Shakira, ~Deb, Drew, Britney. So that is paired comparisons.
The last time I wrote about football, I went to the ESPN website, and ESPN started asking me questions: Which team is better, A or B. And it kept on asking me. Paired comparisons.
Well, they did this to basketball. Problem is, I knew/know nothing about NCAA basketball. Ohio State lost to an SEC school last year. That's about all I know. Well, okay, the ACC is really good, but since I don't like basketball (even though Georgia Tech is good), I don't follow it.
I know why they don't use paired comparisons in the coach's polls: coach's don't have the time to answer a bunch of questions. Plus, I think there is a lot of manipulations going on.
Me, I don't really care who wins the national championship: Georgia Tech is out of the running. The coach is gone. Boo hoo. Besides, another Big Ten team will loose to some SEC team again. Right? Isn't that's what's going to happen. Notice the three contractions in a row – I did that on purpose.
Perhaps something better would be to take the hottest women in the world, and have them pick a winner. Paired comparisons. I believe that these women are familiar with this type of analysis. After all, I have heard many men say, "Nice pair." I assume the comment has to do with this topic.