[R] Data source for American college football rankings?

[Bill Pikounis]

Hi Doug,
An interesting site I stumbled on long ago related to your intial
question of data download is at

http://www.mratings.com/cf/compare.htm

which also contains a CSV file version at

http://www.mratings.com/cf/compare.csv

of *64* different ranking systems.  Other pages within the site
discuss analyses including correlation and summaries like mean and
median.

Hope that helps.
Bill



On Tue, Sep 29, 2009 at 13:55, Douglas Bates <bates at stat.wisc.edu> wrote:
> On Tue, Sep 29, 2009 at 7:56 AM, Douglas Bates <bates at stat.wisc.edu> wrote:
>> An interesting, and topical, example of multivariate data for
>> classroom illustrations are the American college football rankings.
>> Starting at the end of October (or "week 8", the 8th week of the
>> football season) a set of rankings called the BCS (Bowl Championship
>> Series) will be published.  This is a composite ranking based on two
>> subjective polls, the Harris and USA Today polls, and a trimmed mean
>> of 6 objective scores, the so-called computer rankings.  The Harris
>> and USA Today polls are currently available along with several others
>> (the best known and most often quoted is the AP poll but, for
>> complicated reasons, it was replaced in the BCS rankings by the Harris
>> poll).
>>
>> I enjoy using these as classroom examples but I haven't found a web
>> site from which I can download the data directly and am reduced to
>> "screen scraping" the HTML from popular sites like ESPN.  Does anyone
>> know of a site from which one can download the current poll results?
>
> To follow up on my own posting, these data are interesting in part
> because they are presented as tables in many, many places and almost
> never presented graphically.  Some graphical analysis from the
> rankings on October 21, 2007 is shown in
>
> http://www.stat.wisc.edu/~bates/BCS-2007-10-21.pdf
>
> It's not often that an example that many students find interesting can
> be used to illustrate trimmed means, various correlation measures,
> scatterplot matrices, parallel coordinate plots, principal components
> and biplots.
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>