Rivals Recruiting Rankings (4 years) vs. The Top 25

[golfballs]

Chris Low has a blog post about it:
BCS standings and recruiting rankings - College Football Nation Blog - ESPN

Check that chart at the bottom. It's a pretty good graphic

 

[thebjd]

Take an intro logic class... This is about the second thing they'll teach you. Equating coincidence with causation is probably the most common error we make in thinking.

BUT, of course we all want the best players we can get...
Posted via VolNation Mobile

Are you saying a correlation can't be causation?

 

[thebjd]

But even if it isn't mere correlation, you would have to ask what direction the causation is going. Maybe they're getting #1 recruiting classes because they've already shown to be national championship contenders. Or maybe they're national champions because they have #1 recruiting classes. Or maybe there's a common cause to both.

This makes sense.

 

[FREE EARL]

No, I think it's pretty clear that whatever cause/effect sequence we wish to exist is the one that must be in affect when we analyse a causal relationship.

Clearly there is no need for experimentation, or peer review, or any of that other useless drivel we learn in school. Just come up with a theory. Give it the eyeball test, and run with that. There are no other variables than our own wishes.

Yep. Sounds good to me. :dance:

LSU hadn't won a NC until they got their first #1 class. Same with Florida, USC, Alabama, and Texas.

 

[glhal2814]

No, I think it's pretty clear that whatever cause/effect sequence we wish to exist is the one that must be in affect when we analyse a causal relationship.

Clearly there is no need for experimentation, or peer review, or any of that other useless drivel we learn in school. Just come up with a theory. Give it the eyeball test, and run with that. There are no other variables than our own wishes.

Not sure what to tell you if you aren't willing to count the tendency for all the winningest coaches to constantly chase a similar list of highly rated players as peer review.

Not sure what to tell you if you don't want to count the past six years of national championship results, which are the result of 1000's of sporting contests involving the test subjects - to say nothing of practice - as experimentation. Ditto for the following drafts which show the highest correlation of players taken in the early rounds to be from the highest rated players.

Competive sports by their very nature provide a reference point for peer review and experimentation. The players aren't signing Letters of Intent to see who the best recruiter is, then never actually playing the game.

That said, I agree with your main point, that its not a defitively settled debate. Primarly for two reasons.

1) Recruiting services only rate classes back about 10 years. This isn't a lot when it comes to determining National Champions. The Sample Size is very small.
2) The rating system used by the services have been significantly overhauled over the past few years. The way the data was compiled has been changing.

 

[checkerboard03]

Why all the fascination with #1 classes? There are much more interesting ways to look at the data.

I disagree (a little bit) with the idea that these data suggest that recruiting rankings don't matter.

If you exclude the non-BCS schools that don't play a tough schedule, there is a statistically significant relationship between average recruiting ranking and position in the BCS. For those who care, about 17% of variation in the BCS standings is due to recruiting rankings over the last few years.

Blue dots are BCS schools, red dots are non-BCS schools (Boise, TCU, Utah).

It would be better to do this with BCS values (.9038 or whatever) instead of ranks but I don't want to spend a lot of time on this.