Take a look. I took the Massey Composite top 25 for 2014-15, and those schools' average Rivals recruiting rankings from 2010-14 (i.e., five years to account for redshirts), and then subtracted their average ranking from their actual ranking... the greater the negative integer, the more the school outperformed their average recruiting rank...

(02-04-2015 09:41 PM)_sturt_ Wrote:  Take a look. I took the Massey Composite top 25 for 2014-15, and those schools' average Rivals recruiting rankings from 2010-14 (i.e., five years to account for redshirts), and then subtracted their average ranking from their actual ranking... the greater the negative integer, the more the school outperformed their average recruiting rank...

I heard a stat on the radio today that, for as long as they have been doing recruiting rankings, no school has won a national championship without at least one top 10 recruiting class in the previous four years. Oregon would have been the first had they won this year. Good coaching and the ability to develop players, I think, are more important, but recruiting does still matter. Even the best coach can take a team only so far without a strong base to build from.

Recruiting matters, but recruiting rankings are at best a rough estimate of recruiting. The problem with the rankings is that they are largely a reflection of popularity of recruits in the eyes of the traditionally strong schools. It's a good way to base rankings as it follows that the "best" programs are going to offer and land the "best" recruits. But it means that those schools are by definition always going to have the highest rated classes. I don't think it would be possible to do the rankings any other way, though. No way the recruiting services have the ability or time to accurately rank every prospect independently. Relying on the evaluations of the coaches makes more sense. But it is still a self fulfilling prophecy.

But it shouldn't be a surprise when there are teams or players that outperform their rankings. There are so many strong prospects that get overlooked by the traditionally strong programs. Prospect evaluations and projections are difficult- just look at the NFL draft who miss often with way more information and data. Teams who know the kind of players that are undervalued and can develop those players are going to outperform their recruiting "rankings"

Agreed, Frog.

I suppose what is somewhat stunning to me is 40% of those teams ended up exceeding their 5-year average recruiting ranking by, not just a few, but 20 slots or more.

Before I started into the spreadsheet, I would have predicted maybe 20% at most...

And what is absolutely stunning is that less than half of the actual top 25 averaged a top 25 recruiting ranking the previous 5 years.

Ignore the big name programs and go from about 10 and down. Matching Coaching Styles and Players are far more important than stars. Too many times Coaches are more concerned about stars than actual fit. Granted if a coach can get a 4-5 star that fits into his scheme, it is a no brainer. However, the good coaches might see that three star but knows that kid is a far better fit into his system than some four stars. I think a prime example of this is in the Triangle area. Duke with Coach Cutcliff and Coach McNeil at ECU are very good at this.. It appears NCSU Coach Doeren may have this philosophy looking at his past two classes, but it is still too early to determine yet. Coach Fedora in my opinion, wants the highest number of stars and appears not to be as concerned about fit. Not saying it is wrong, Mack Brown made a living off getting top players.

Interesting analysis, however there are too many other variables, the most important being strength of schedule.

I believe that any delta less than 10 means the team is performing where expected; 10-20, slightly better and more than 20, outstanding.

Would be nice to see a similar analysis of SOS vs ranking (P5 only).

Here are the grains of salt to take with recruiting rankings.

#1. Rankings are based on a recruit's basic value versus other members of that group of high school seniors.
#2. The progression of rankings is linear. In reality the talent and ability gap between the #1 rated QB and the #5 rated QB can be much greater than the gap between the #5 QB and the #20 QB. The school signing the #5 QB will be more rewarded in rankings than the school signing #20 compared to the ranking advantage of signing #1 instead of #5.
#3. Rankings cannot effectively adjust for a player's fit within a system. The player you need for Hugh Freeze's offense isn't identical to what you need to fit withing Saban's system. The rankings cannot assess the mental aspects effectively. Not just the kid's attitude but whether he will respond to the style of the staff he signs with.
#4. Rankings only assess the player vs other players and not their utility to a team. If a school thinks it will be loaded for bear in 2016 and the missing piece of the puzzle is a Mike LB, then signing two top LB's and having an average class all around means it is an average class in rankings but quite possibly it could be the most important class the school has ever signed.
#5. There is always the risk of a scout second-guessing himself. If he thinks a kid is a three star and suddenly Bama and Auburn are pulling out all the stops, he may go back and reevaluate his assessment wondering what he missed that the coaches saw.

The rankings do have some useful benchmarking and are worth consideration they just aren't the final answer on a class.

(02-05-2015 10:42 AM)mac6115cd Wrote:  Interesting analysis, however there are too many other variables, the most important being strength of schedule.

Let me say from the top that I'm not sure that I'm getting your drift. But assuming I am...

I think it's fair to assume that most of the computer-based rankings of the 75-80 inputs to the Massey Composite have some mechanism in their formulas that accounts for SOS, and that, inherently, when voters vote, they make decisions based on their own SOS perceptions. So, I'm not sure why it's suggested that SOS isn't considered.

Ultimately, this is one of those "it is what it is" things. What explains why it is how it is can be (and should be) discussed. But the math is the math is the math, and clearly, a major conclusion is that on-the-field success is not exclusively a function of having your recruiting classes routinely ranked high.

(02-05-2015 11:06 AM)_sturt_ Wrote:  

(02-05-2015 10:42 AM)mac6115cd Wrote:  Interesting analysis, however there are too many other variables, the most important being strength of schedule.

Let me say from the top that I'm not sure that I'm getting your drift. But assuming I am...

I think it's fair to assume that most of the computer-based rankings of the 75-80 inputs to the Massey Composite have some mechanism in their formulas that accounts for SOS, and that, inherently, when voters vote, they make decisions based on their own SOS perceptions. So, I'm not sure why it's suggested that SOS isn't considered.

Ultimately, this is one of those "it is what it is" things. What explains why it is how it is can be (and should be) discussed. But the math is the math is the math, and clearly, a major conclusion is that on-the-field success is not exclusively a function of having your recruiting classes routinely ranked high.

Strength of schedule is obviously factored in the computer rankings. That is obvious from the fact that an undefeated G5 is regularly rated below teams with one and two losses.

But here is the interesting fact of recruiting.

Go back for 15 years and list the number of teams that finished with national championship. Then look at their recruiting classes.

I cannot thing of a SINGLE team that won a national championship and did not have a top five recruiting classes within two years of winning it.

Recruiting only does not really matter at the G5 level. You cannot really accurately determine what makes someone a 3 star vs a 2 star or a 2 star vs a 1 star.

(02-05-2015 10:42 AM)mac6115cd Wrote:  Interesting analysis, however there are too many other variables, the most important being strength of schedule.

I believe that any delta less than 10 means the team is performing where expected; 10-20, slightly better and more than 20, outstanding.

Would be nice to see a similar analysis of SOS vs ranking (P5 only).

Agreed. Also think that in judging recruiting ranking you really need to look at patterns more than individual results. You can easily change a few plays in a year and effect several teams ranking considerably. That isn't really a reflection on recruiting ranking, just the wildness of college football.

It would probably work better if, rather than just using this years data for the ranking, if they did the same thing for a set of 5 years (using the 4 classes relevant to those particular years) and then figured an average final ranking in both the season and the recruiting (although this would require something besides the AP or Coaches polls as you'd need to go beyond top 25).

All that said, very interesting none-the-less.

I bet BYU in 1984 did not have a top ten class, and maybe Miami of Florida's early classes. But then again, that was a long time ago

(02-05-2015 10:44 AM)arkstfan Wrote:  Here are the grains of salt to take with recruiting rankings.

#1. Rankings are based on a recruit's basic value versus other members of that group of high school seniors.
#2. The progression of rankings is linear. In reality the talent and ability gap between the #1 rated QB and the #5 rated QB can be much greater than the gap between the #5 QB and the #20 QB. The school signing the #5 QB will be more rewarded in rankings than the school signing #20 compared to the ranking advantage of signing #1 instead of #5.
#3. Rankings cannot effectively adjust for a player's fit within a system. The player you need for Hugh Freeze's offense isn't identical to what you need to fit withing Saban's system. The rankings cannot assess the mental aspects effectively. Not just the kid's attitude but whether he will respond to the style of the staff he signs with.
#4. Rankings only assess the player vs other players and not their utility to a team. If a school thinks it will be loaded for bear in 2016 and the missing piece of the puzzle is a Mike LB, then signing two top LB's and having an average class all around means it is an average class in rankings but quite possibly it could be the most important class the school has ever signed.
#5. There is always the risk of a scout second-guessing himself. If he thinks a kid is a three star and suddenly Bama and Auburn are pulling out all the stops, he may go back and reevaluate his assessment wondering what he missed that the coaches saw.

The rankings do have some useful benchmarking and are worth consideration they just aren't the final answer on a class.

As I posted on the other thread, the Georgia Tech coach was telling how they were recruiting a 2 star running back. Then a top 5 school started recruiting him and 3 days later, he's a 4 star running back. So there's some self-fulfilling prophecy on the recruiting rankings.

(02-05-2015 01:28 PM)bullet Wrote:  

(02-05-2015 10:44 AM)arkstfan Wrote:  Here are the grains of salt to take with recruiting rankings.

#1. Rankings are based on a recruit's basic value versus other members of that group of high school seniors.
#2. The progression of rankings is linear. In reality the talent and ability gap between the #1 rated QB and the #5 rated QB can be much greater than the gap between the #5 QB and the #20 QB. The school signing the #5 QB will be more rewarded in rankings than the school signing #20 compared to the ranking advantage of signing #1 instead of #5.
#3. Rankings cannot effectively adjust for a player's fit within a system. The player you need for Hugh Freeze's offense isn't identical to what you need to fit withing Saban's system. The rankings cannot assess the mental aspects effectively. Not just the kid's attitude but whether he will respond to the style of the staff he signs with.
#4. Rankings only assess the player vs other players and not their utility to a team. If a school thinks it will be loaded for bear in 2016 and the missing piece of the puzzle is a Mike LB, then signing two top LB's and having an average class all around means it is an average class in rankings but quite possibly it could be the most important class the school has ever signed.
#5. There is always the risk of a scout second-guessing himself. If he thinks a kid is a three star and suddenly Bama and Auburn are pulling out all the stops, he may go back and reevaluate his assessment wondering what he missed that the coaches saw.

The rankings do have some useful benchmarking and are worth consideration they just aren't the final answer on a class.

As I posted on the other thread, the Georgia Tech coach was telling how they were recruiting a 2 star running back. Then a top 5 school started recruiting him and 3 days later, he's a 4 star running back. So there's some self-fulfilling prophecy on the recruiting rankings.

And it is also likely that the kid had 2 stars assigned based on his performance as a junior and he improved more than expected which led to other schools that were not interested in what they saw of him as a junior and the kid was re-evaluated based on his senior performance which did warrant being a four star.

The change in rating could well not be an issue of the top 5 school coming in but rather kid's caliber of play improved to a four star level which is what brough that school in to recruit him.

You can't win in P-5 football without talent. Talent can overcome mediocre coaching, but mediocre coaching will ruin talent.

Where the talent is also matters. A 5-star quarterback, running back, defensive tackle, or defensive end can change the game. An offensive lineman not so much. A great offensive line can not make a bad qb look good.

(02-05-2015 02:30 PM)lumberpack4 Wrote:  You can't win in P-5 football without talent. Talent can overcome mediocre coaching, but mediocre coaching will ruin talent.

Where the talent is also matters. A 5-star quarterback, running back, defensive tackle, or defensive end can change the game. An offensive lineman not so much. A great offensive line can not make a bad qb look good.

Positions affect results differently.

With OL, it isn't the best player, but the weakest one which determines effectiveness.

Defense is similar - if there is one great CB and one bad one, the opposing offense will simply throw at the weakest link.

QB, RB and WR are positions where one great player CAN make a difference.

(02-05-2015 01:16 PM)TrojanCampaign Wrote:  But here is the interesting fact of recruiting.

Go back for 15 years and list the number of teams that finished with national championship. Then look at their recruiting classes.

I cannot thing of a SINGLE team that won a national championship and did not have a top five recruiting classes within two years of winning it.

Recruiting only does not really matter at the G5 level. You cannot really accurately determine what makes someone a 3 star vs a 2 star or a 2 star vs a 1 star.

I think that is true, but I wonder if it might not be correlation and not causation. A couple honest questions in my mind:

1. Is it possible recruiting rankings simply restate who the traditionally top teams are? Since the traditionally top teams are the most likely teams to have future success, ranking the recruits from those programs highly is always going to "predict" success. And if it doesn't for one of the traditionally strong team, odds are for another it will.

2. Is the continued success of the traditionally strong teams due to superior recruiting or could it be due to other factors- superior facilities, coaches, player development, schemes?

3. And then there is the question of opportunity. In the past did the system give the traditional powers more access to the national championship than non-traditional powers with lower ranked recruits?

(02-05-2015 02:42 PM)Frog in the Kitchen Sink Wrote:  

(02-05-2015 01:16 PM)TrojanCampaign Wrote:  But here is the interesting fact of recruiting.

Go back for 15 years and list the number of teams that finished with national championship. Then look at their recruiting classes.

I cannot thing of a SINGLE team that won a national championship and did not have a top five recruiting classes within two years of winning it.

Recruiting only does not really matter at the G5 level. You cannot really accurately determine what makes someone a 3 star vs a 2 star or a 2 star vs a 1 star.

I think that is true, but I wonder if it might not be correlation and not causation. A couple honest questions in my mind:

1. Is it possible recruiting rankings simply restating who the traditionally top teams are? Since the traditionally top teams are the most likely teams to have future success, ranking the recruits from those programs highly is always going to "predict" success. And if it doesn't for one of the traditionally strong team, odds are for another it will.

2. Is the continued success of the traditionally strong teams due to superior recruiting or could it be due to other factors- superior facilities, coaches, player development, schemes?

3. And then there is the question of opportunity. In the past did the system give the traditional powers more access to the national championship than non-traditional powers with lower ranked recruits?

Obviously if a coach at say Texas does a poor job evaluating talent the services could skew results, however you rarely see a Texas/Oregon/USC/Alabama being the only big name offering a kid. You don't often see Alabama sign a kid who has two other offers, one from Middle Tennessee and one from Arkansas State.

I suspect if you compared a low rated recruit who signs with Arkansas State and makes the NFL to a high rated recruit who signs with Alabama and misses on the NFL that you find (assuming no academic or discipline issues) that the lower regarded player probably played more snaps at Arkansas State and started more games. The difference in depth gave that player a chance to grow and develop more and it isn't like a Sun Belt player who plays four years isn't going to play 8-12 games against top schools so they can gain some truly valuable experience (AState had a true freshman noseguard start against Tenn and Miami this year because of injuries to the line).

The other thing to remember.

Recruiting rankings assess the player at age 17 or 18.
Four years (or even five) is a long time when you consider that you are dealing with people who are not yet at their physical prime.

The extreme example is Scottie Pippen. He was a little regarded six foot one player out of Hamburg, Arkansas who landed at the University of Central Arkansas and graduated as a six foot eight pro prospect.

Did the recruiting services "miss" on Pippen? I would argue not. Even his college coach wasn't projecting him to add seven inches in college. Coming out of high school he was rated about right (unrated). What happened in college changed his stock.

If the recruiting services were projecting where players would be in four or five years, then they are mediocre at the task but they are only assessing them where they are at that point in time.

(02-05-2015 01:55 PM)arkstfan Wrote:  

(02-05-2015 01:28 PM)bullet Wrote:  

(02-05-2015 10:44 AM)arkstfan Wrote:  Here are the grains of salt to take with recruiting rankings.

#1. Rankings are based on a recruit's basic value versus other members of that group of high school seniors.
#2. The progression of rankings is linear. In reality the talent and ability gap between the #1 rated QB and the #5 rated QB can be much greater than the gap between the #5 QB and the #20 QB. The school signing the #5 QB will be more rewarded in rankings than the school signing #20 compared to the ranking advantage of signing #1 instead of #5.
#3. Rankings cannot effectively adjust for a player's fit within a system. The player you need for Hugh Freeze's offense isn't identical to what you need to fit withing Saban's system. The rankings cannot assess the mental aspects effectively. Not just the kid's attitude but whether he will respond to the style of the staff he signs with.
#4. Rankings only assess the player vs other players and not their utility to a team. If a school thinks it will be loaded for bear in 2016 and the missing piece of the puzzle is a Mike LB, then signing two top LB's and having an average class all around means it is an average class in rankings but quite possibly it could be the most important class the school has ever signed.
#5. There is always the risk of a scout second-guessing himself. If he thinks a kid is a three star and suddenly Bama and Auburn are pulling out all the stops, he may go back and reevaluate his assessment wondering what he missed that the coaches saw.

The rankings do have some useful benchmarking and are worth consideration they just aren't the final answer on a class.

As I posted on the other thread, the Georgia Tech coach was telling how they were recruiting a 2 star running back. Then a top 5 school started recruiting him and 3 days later, he's a 4 star running back. So there's some self-fulfilling prophecy on the recruiting rankings.

And it is also likely that the kid had 2 stars assigned based on his performance as a junior and he improved more than expected which led to other schools that were not interested in what they saw of him as a junior and the kid was re-evaluated based on his senior performance which did warrant being a four star.

The change in rating could well not be an issue of the top 5 school coming in but rather kid's caliber of play improved to a four star level which is what brough that school in to recruit him.

If that's the case, that would certainly indicate that there's no point in looking at the recruiting rankings until late in the game (if ever), when every major school would have looked at the film from their senior year. I think it's a stretch though. When do these jumps in star rating actually occur?

(02-04-2015 10:15 PM)prp Wrote:  

(02-04-2015 09:41 PM)_sturt_ Wrote:  Take a look. I took the Massey Composite top 25 for 2014-15, and those schools' average Rivals recruiting rankings from 2010-14 (i.e., five years to account for redshirts), and then subtracted their average ranking from their actual ranking... the greater the negative integer, the more the school outperformed their average recruiting rank...

I heard a stat on the radio today that, for as long as they have been doing recruiting rankings, no school has won a national championship without at least one top 10 recruiting class in the previous four years. Oregon would have been the first had they won this year. Good coaching and the ability to develop players, I think, are more important, but recruiting does still matter. Even the best coach can take a team only so far without a strong base to build from.

It seems like the biggest name schools do relatively poorly by this measure. To me that indicates a little overzealousness in giving high ranks to those players that get offers to the top schools. Just a thought.