Reviewing The 2006 Recruiting Class: Top 50

is that ESPN's rankings were shaky around this time but have improved the last few years. Would love to know if that's actually true, but I would have expected Rivals and Scout to have outperformed ESPN in 2006 (with Rivals maybe outperforming Scout).  

..."actual research" is nice to see.

That should be a new recruiting metric: "Propensity to Rob Liquor Store."

"Jimmy has exceptional speed, hips on swivel and excellent pad level.  Major concern is the unregistered hand gun he likes to carry and the abundance of liquor stores in his neighborhood."

how the hits/misses progress through an era where video taking and posting to the web becomes more wisespread. With phones coming out now that can take 1080p video and edit on the fly, its not going to be just camps where the student athletes can be evaluated. Anyone can have a bad/good weekend but over a 10ish game season plus camps it'll be much easier to grade players out. 

What is much harder to grade a player out on is the system he gets put in and how it accentuates his abilities or hinders them. Sometimes a player is punished by bad coaching/scheme and can lead to a false miss. I'm sure Toby Gerhart is not a Heisman finalist if he's playing for Oregon....

Like you suggested, without seeing the top 50 from Scout & Rivals, I don't think you can say much in terms of comparisons. 

Here, we see some of the guys who ESPN thought they'd take a shot on even though Scout & Rivals weren't as crazy about them.  You wouldn't expect those kids to perform as well as the across-the-board, unanimous stars.  It'll be interesting to see how the other two services did when they took shots with kids that ESPN and/or the other service weren't as crazy about.

There was a lot of teeth-gnashing when Eugene dropped Michigan in favor of LSU.  Hopefully people feel better about missing out on him now (circumstances of the change aside).

This kinda shows what I generally see in these "look back" pieces that I love.   It seems that the really top kids the (say top 20) most of the services agree on and they generally become very successful.   Once you get away from the real freaks it becomes much more of a crap shoot. 

This says to me that these guys don't scout as much as just look at the kids body and his measurables and guess.  

We only had 1 of the top 50?

UNACCEPTABLE!!!!!!!!!!

/s

USC has had so much "talent".

I've always believed ESPN's rankings are the worst of the 3 sites (4 now with 24/7, but there rankings seem to be all over the place to0).  They tend to grade guys going to the SEC or USC always a little higher than they should.  This breakdown just gives more proof to that argument.  They were incorrect big time on quite a few guys.

But the ranking differential on Gerald McCoy is the one that stood out to me. Also, to a lesser extent, CJ Spiller. Major ESPN biffs.

This is the type of info that makes me a regular visitor on Mgoblog and puts this web site in an elite group for the multiple Michigan fans throughout the world, starving for information on our our great university, who seek sports information  . I have often wondered what the followup on common rankings is.   Nice job.  Keep up the in depth analysis.

This is an example of one of the reasons why I am a regular visitor on mgoblog.  Good analysis of what the rankings mean.  Keep up the good work.

But I'm not sure what one can make of it, other than a snapshot in time, partly because of a lack of a control group and partly because of confounding variable (i.e. "coaching").  Another confounding variable is that players do not choose teams at random (a "highly ranked" linebacker may be more likely to choose to go to Linebacker U).  There seems to be several ways to look at the "rankings" from a fan perspective, each probably fairly painful to analyze, but my guess are these are what most people care about (very similar to what Brian posted):

Outcome #1: I want my college team to do great

Outcome #2: High school kid makes it to the NFL

Outcome #3: High school kid becomes NFL stud

So, the hope when you have a highly ranked recruiting class is that it will materialize into wins and mythical national championship.  To analyze that you could look at a specific year for a team (or league), go down the roster and average the rankings (similar to what Tim graciously provides) then look at the number of wins.  To get a really good measure, it may be helpful to do this over 4 or 5 years for a team/league.  That may even be enough data to gauge whether there is a recruiting service ranking that is a better predictor than others.

The other outcome is of course whether a kid goes pro.  One way to do this is looking at the NFL draft over a few years and seeing what each player was ranked by each recruiting service (if even ranked).  One could probably parse out which service better predicts pro potential.

Lastly, one could do the same thing for pro teams that I mentioned doing for college teams.  One could look at playoff teams, super bowl winners, or all comers.  Since "coaching" is a confounding variable, maybe just looking at pro bowl players over a few years makes sense.

Apologies for the long post, and doubly for presenting a methodology without actually doing the work.  Hopefully, time will allow me to actually follow-up on this...