[Husker] *Matt Hayes* discovers discrepencies between recruiting rankings and results

[Steve Reichenbach]

I think you're confusing statistics with the inferences that you are
drawing.  You seem to draw conclusions that the journalist who posted
this article didn't.  For the most part, the journalist reported the
computed difference between recruiting rank and winning rank.  The
journalist didn't say it was clear what the numbers mean, but they are
interesting and worthy of reporting and discussion.

You seem to be looking at the numbers to see if recruiting rankings
are meaningful, but if that were the point, then the list would
have ranked teams by number of wins and not by differential.  For
an example of that, see: http://recruiting.scout.com/2/836034.html.
>From comparing recruiting rankings with success, as the Scout
article does, it is clear that there is a fairly high correlation
between recruiting rankings and success, but that correlation is not
perfect.  I think this ranking by the differential tries to look at
possible reasons for why recruiting rankings are not perfect predictors
of winning, i.e., what other factors are involved.

Certainly one is that the recruiting rankings aren't perfect, which you
note as a concern.  This article acknowledges that issue in the notes
in the top ten specifically identifying underrated recruits.  But, I
don't think that's the point of the rankings either.  After all, who
cares if the recruiting rankings didn't give enough credit to
Cincinnati's recruits (even if that is a possible conclusion from
looking at these numbers).

By ranking the teams by differential, the ranking appears intended
to measure the success of the coaches at (a) knowing better than the
recruiting rankings and/or (b) getting the most out of the players that
they recruited.  And, I think the emphasis probably is on (b).  Hence,
the interest in the differential between recruiting rankings and
success.  (Here, I too am going beyond what Hayes wrote, as he
described being "curious to see what schools do the most with
least --- and which ones can't turn impressive recruiting classes
into winning teams".  Their numbers are for "schools" not coaches
so their description is more accurate than mine, but ultimately
the coaches at those schools are most directly responsible.)

I think the conclusion that Cincinnati's coaches did a better job than
Washington's coaches did during the period of analysis is well
supported by this data.  Looking at the some of the teams low on the
list --- Washington, Texas A&M, Notre Dame, and even Nebraska --- you
have evidence to support the conclusion that the coaches don't deserve
much credit.  Similarly, the coaches of teams high on the list probably
deserve quite a bit of credit.  Closest to home, in the Big 12, you
have to marvel at Mike Leach's success as a coach at Texas Tech.

So, as a scientist and sports fan interested in statistics, I heartily
disagree with the assertion that science and statistics don't mix with
journalism.

> On Wed, Feb 18, 2009 at 10:39 PM, Mike Jaixen <mikejaixen at yahoo.com> wrote:
> 
> This is an interesting article, and it certainly is additional fodder
> for sports lists, but its not exactly clear what all of this means.
> Rivals was the only ranking service used and billions upon billions of
> electrons have been redirected from useful purposes in the debate over
> whether Rivals means anything at all.  It would have been a bit more
> useful if more than just Rivals had been used.  But a main criticism
> of the article is that comparing success of a recruiting class to wins
> in the same year is really meaningless, since there is almost always
> (not a perfect correlation) a lag of a couple of years before recruits
> become juniors and seniors and their true abilities can be seen in
> team performance.  A second concern is whether success can only be
> measured by number of wins.  We won more games this year than last,
> but was that due to the recruits that came in this last year, or the
> year before, or the year before that?  I sorta thought it may have had
> something to do with how that recruited talent was put into play on
> the field, but I could be all wet.
> 
> One person commented on that article that journalism and statistics
> shouldn't mix.  I'd agree with that.  Just like what Scott Adams said:
> "Science is a good thing. News reporters are good things too. But it's
> never a good idea to put them in the same room."
>