Recruiting Potential

[GT_EE78]

Last year we finished 27th per composite. That's a reasonable floor.
i think 16-25 should be achievable

 

[Coloradojacket]

Average star rating is probably the worst way to compare recruiting lol

Ok, so what metric would you use. Points as in what Rivals uses? If so the right now Rutgers is out recruiting us. Along with some other schools that would raise your eye brows....

 

[Oakland]

And what do you consider elite recruiting?
Specifically what range do we need to be in to compete at the level CGC has stated for us?

I don't think we'll really obtain the so called elite level that CGC dreams of. I think we can get good recruits, but not on the level that the factory schools can get. That doesn't mean we can't have competitive football and challenge for the ACC championship, just not on a yearly level. For Georgia Tech 25 to 35 is doing pretty good.

 

[JacketOff]

Ok, so what metric would you use. Points as in what Rivals uses? If so the right now Rutgers is out recruiting us. Along with some other schools that would raise your eye brows....

Points are how the rankings are determined... so yeah that’s the better metric. Average stars isn’t really useful because 1 player can heavily skew the average.

The average star metric includes every player in a signing class past the 20 that the points metric counts. So if a team signs 20 3-star players a 1 2-star player, their average stars drop to 2.95. Does that mean a team who signs 20 3-stars and 1 2-star had a worse recruiting class than a team who just signs 20 3-stars? No.

The average star system also doesn’t account for the discrepancies between a player on the low end of a “star” and the high end. A 5.7 3-star is generally recruited way differently than a 5.4 3-star. A 5.7 will usually have offers from numerous P5 schools, while a 5.4 may only have a couple. That discrepancy shows up in the points metric, but not average stars.

 

[InsideLB]

26-35 because we wont have as much attrition as other schools, and therefore smaller classes.

Relative to historical, which is more like 35-50 or so it's a big step up. But I dont see factory like 'crootin at GT.

 

[lv20gt]

26-35 because we wont have as much attrition as other schools, and therefore smaller classes.

I believe most recruiting services only use the top 20 recruits in each class to judge.

 

[swampsting]

We were ranked 24-28th last year depending on which scouting website you used. If Collins figures things out on the field, he better average a recruiting class ranking of 16-25.

at this point, that is a mighty big if.

 

[takethepoints]

Points are how the rankings are determined... so yeah that’s the better metric. Average stars isn’t really useful because 1 player can heavily skew the average.

The average star metric includes every player in a signing class past the 20 that the points metric counts. So if a team signs 20 3-star players a 1 2-star player, their average stars drop to 2.95. Does that mean a team who signs 20 3-stars and 1 2-star had a worse recruiting class than a team who just signs 20 3-stars? No.

The average star system also doesn’t account for the discrepancies between a player on the low end of a “star” and the high end. A 5.7 3-star is generally recruited way differently than a 5.4 3-star. A 5.7 will usually have offers from numerous P5 schools, while a 5.4 may only have a couple. That discrepancy shows up in the points metric, but not average stars.

The problem with this analysis is the one with every evaluation system I've ever looked at and I've looked at plenty of them. It assumes that there is a discernible gradation between gross rankings.

Is it easy to discern who is a "5 star" player? Sure; they stick out like sore thumbs. It is less easy, but not an outrageous step to pick out some "4 star" players, provided you have the staff to do so and a well validated scale to use. But … that simply isn't the case with the ratings services. Their "staff' is made up of fans of various sorts who volunteer or are paid minimal amounts to rate players, usually for teams they are interested in. ( In the survey biz this is called rater bias.) Then they proceed to rank them on scales that use decimal points down to the 10K place to "determine" differences. This is done because the ratings sites demand it and they demand it because it is click bait, pure and simple. It should be possible to rate football players out of high school more effectively, though it would be much harder then for, say, baseball due to the smaller sample sizes. Do the rating sites try to do this? Why, no, they don't. It isn't in their interests to try.

This is why I've always thought that using average stars is a better way to approach comparing recruiting classes. Not much better, I'll admit, but better. Doing it that way gets rid of the noise introduced by the decimal ratings and cuts to the chase. There are problems with even this approach, as Jacket points out, but it is less likely to deceive the consumers about what the actual status of the recruiting efforts is. And the les likely a school is to bring in a top 10 class, the more accurate an approach based on gross rankings is likely to be.

 

[sgreer]

To ever get to a Top 20 class Tech will have to get some of the Top 10-20 rated players in the state of Ga. This means we will have to steal one from dogs, Auburn, Clemson, etc.

 

[JacketOff]

The problem with this analysis is the one with every evaluation system I've ever looked at and I've looked at plenty of them. It assumes that there is a discernible gradation between gross rankings.

Is it easy to discern who is a "5 star" player? Sure; they stick out like sore thumbs. It is less easy, but not an outrageous step to pick out some "4 star" players, provided you have the staff to do so and a well validated scale to use. But … that simply isn't the case with the ratings services. Their "staff' is made up of fans of various sorts who volunteer or are paid minimal amounts to rate players, usually for teams they are interested in. ( In the survey biz this is called rater bias.) Then they proceed to rank them on scales that use decimal points down to the 10K place to "determine" differences. This is done because the ratings sites demand it and they demand it because it is click bait, pure and simple. It should be possible to rate football players out of high school more effectively, though it would be much harder then for, say, baseball due to the smaller sample sizes. Do the rating sites try to do this? Why, no, they don't. It isn't in their interests to try.

This is why I've always thought that using average stars is a better way to approach comparing recruiting classes. Not much better, I'll admit, but better. Doing it that way gets rid of the noise introduced by the decimal ratings and cuts to the chase. There are problems with even this approach, as Jacket points out, but it is less likely to deceive the consumers about what the actual status of the recruiting efforts is.

Which is why I believe Rival’s ratings are much better than 247. Like I said, there’s generally some pretty discernible differences in a player’s recruiting status as a 5.7 rated player than a 5.4. Does that mean a 5.4 can’t be as good as that 5.7? Not necessarily, but it means that the 5.7 is probably a better fit at more schools and will probably be in better position to get early playing time than the 5.4.

I think it’s kind of ridiculous to rank prospects numerically (as in top 50, 100, 250, etc.) The main reason is because they’re ranking players in different positions, who play in different states, against different competition. Just as an example, the 17th, 18th, and 19th rated players on Rivals are a QB from Arizona, a S from Florida, and an OT from Indiana respectively. There’s really no way you can logically compare those players enough to rank one ahead of the others.

But, as a whole, the recruiting services do a very good job of giving a really good idea about teams stand in recruiting, but they don’t (nor can they) give the full picture. That’s because teams lose points for not signing a “full” class of at least 20 players, and there’s no way to account for teams filling/failing to fill holes in their depth chart. Just as an example, Tech desperately needs to build depth at LB and DT. Tech has the #25 class on rivals for 2020, but didn’t sign a single DT, and only 2 LBs. So while some of the 5 DEs we signed will almost definitely have to change positions to LB or DT at some point in the future, they’re rated as DE’s and their impending positional changes don’t affect their rating.

I think as far as team rankings go, the services do a really good job. Player ratings are hit or miss sometimes, but are generally pretty accurate. I think what most people don’t realize is that players are rated based on what the evaluators view as their floor, not their ceiling. Every player has virtually the same ceiling, that would be an NFL Hall of Famer. Every player doesn’t have the same floor. You’d be hard pressed to find a 5 star player who isn’t a starter, or at the very least one of the biggest platoon guys. Not all of them turn into superstars, but almost none of them are complete busts. Just about every 4 star will at least see decent playing time throughout their careers. Most 3 stars are average players, and there aren’t many players rated below 3 stars who get more than occasional playing time.

But going back to the metrics that are used to rank teams. I still think average stars is a very poor barometer because of how much 1 player can skew that average. Especially since players past the cutoff of 20 are still averaged in. If Tech had simply not signed Ryan Spiers or Emmanuel Johnson last year, the average star total would’ve went from 3.08 to 3.18. I don’t think signing more players should hurt your recruiting rankings.
You can also look at a team like TCU who signed a 5-star player, but only signed 6 4-stars and 10 3-stars, and they have a better average star score than Washington who signed no 5-stars but 10 4-stars and 9 3-stars. Is that one 5-star player that TCU signed worth more than the 4 extra 4-stars that Washington signed?

 

[RonJohn]

Which is why I believe Rival’s ratings are much better than 247. Like I said, there’s generally some pretty discernible differences in a player’s recruiting status as a 5.7 rated player than a 5.4. Does that mean a 5.4 can’t be as good as that 5.7? Not necessarily, but it means that the 5.7 is probably a better fit at more schools and will probably be in better position to get early playing time than the 5.4.

I think it’s kind of ridiculous to rank prospects numerically (as in top 50, 100, 250, etc.) The main reason is because they’re ranking players in different positions, who play in different states, against different competition. Just as an example, the 17th, 18th, and 19th rated players on Rivals are a QB from Arizona, a S from Florida, and an OT from Indiana respectively. There’s really no way you can logically compare those players enough to rank one ahead of the others.

But, as a whole, the recruiting services do a very good job of giving a really good idea about teams stand in recruiting, but they don’t (nor can they) give the full picture. That’s because teams lose points for not signing a “full” class of at least 20 players, and there’s no way to account for teams filling/failing to fill holes in their depth chart. Just as an example, Tech desperately needs to build depth at LB and DT. Tech has the #25 class on rivals for 2020, but didn’t sign a single DT, and only 2 LBs. So while some of the 5 DEs we signed will almost definitely have to change positions to LB or DT at some point in the future, they’re rated as DE’s and their impending positional changes don’t affect their rating.

I think as far as team rankings go, the services do a really good job. Player ratings are hit or miss sometimes, but are generally pretty accurate. I think what most people don’t realize is that players are rated based on what the evaluators view as their floor, not their ceiling. Every player has virtually the same ceiling, that would be an NFL Hall of Famer. Every player doesn’t have the same floor. You’d be hard pressed to find a 5 star player who isn’t a starter, or at the very least one of the biggest platoon guys. Not all of them turn into superstars, but almost none of them are complete busts. Just about every 4 star will at least see decent playing time throughout their careers. Most 3 stars are average players, and there aren’t many players rated below 3 stars who get more than occasional playing time.

But going back to the metrics that are used to rank teams. I still think average stars is a very poor barometer because of how much 1 player can skew that average. Especially since players past the cutoff of 20 are still averaged in. If Tech had simply not signed Ryan Spiers or Emmanuel Johnson last year, the average star total would’ve went from 3.08 to 3.18. I don’t think signing more players should hurt your recruiting rankings.
You can also look at a team like TCU who signed a 5-star player, but only signed 6 4-stars and 10 3-stars, and they have a better average star score than Washington who signed no 5-stars but 10 4-stars and 9 3-stars. Is that one 5-star player that TCU signed worth more than the 4 extra 4-stars that Washington signed?

I have said before that I don't trust team rankings as a direct measurement. As a measurement of order of magnitude, it is fine. In 2020, Alabama's class was better than GT and ranked higher. But was Florida State's class definitely better than GT's? FSU was 20th and GT was 25th. Was GT's class definitely better than Purdue's? GT was 25th and Purdue was 30th. People like to pay attention to and argue about differences in hundredths or maybe even thousandths when the measurements used are not even accurate to whole numbers.

You can also look at a team like TCU who signed a 5-star player, but only signed 6 4-stars and 10 3-stars, and they have a better average star score than Washington who signed no 5-stars but 10 4-stars and 9 3-stars. Is that one 5-star player that TCU signed worth more than the 4 extra 4-stars that Washington signed?

Going by that example, Washington would have been higher based on number of stars. 59/17 = 3.47 (TCU) 67/19 = 3.53 I take it from your question that you would think the extra 4 stars would be worth more than the 5 star. Average stars would say that is the case. The Rivals ranking said that is the case. However, I would ask the same question as above. Are you definitely, 100% sure that Washington's class was better and will end up performing better on the field than TCU's? I don't think you are, and I don't think it is possible to be. It is similar to attempting to determine which Dasani bottle has the most water in it just by looking at the bottles. The 1.5L bottle (Alabama) definitely has more than a 20 ounce bottle (TCU, Washington). But you can't tell just by looking at the bottle which 20 ounce bottle actually has more water in it.

 

[JacketOff]

Going by that example, Washington would have been higher based on number of stars. 59/17 = 3.47 (TCU) 67/19 = 3.53 I take it from your question that you would think the extra 4 stars would be worth more than the 5 star. Average stars would say that is the case. The Rivals ranking said that is the case. However, I would ask the same question as above. Are you definitely, 100% sure that Washington's class was better and will end up performing better on the field than TCU's? I don't think you are, and I don't think it is possible to be. It is similar to attempting to determine which Dasani bottle has the most water in it just by looking at the bottles. The 1.5L bottle (Alabama) definitely has more than a 20 ounce bottle (TCU, Washington). But you can't tell just by looking at the bottle which 20 ounce bottle actually has more water in it.

While the math you did to come up with those averages is correct, it’s not how they’re ranked. Washington signed 23 players; 10 4-stars, 9 3-stars, and 4 2-stars for a star average of 3.26. TCU signed 19 total players; 1 5-star, 6 4-stars, 10 3-stars, and 2 2-stars for a star average of 3.32. Washington would’ve had a better star average if they just hadn’t signed those 2-star players, which is the biggest reason why I think using average stars is a flawed metric.

As far as the rest of your comment is concerned, it doesn’t really have anything to do with the rankings themselves. Recruiting rankings aren’t a metric for a team’s success in the future because they aren’t supposed to be. They’re supposed to give a gauge on how much talent a team has. If anything they’re only useful to determine which teams are underperforming, exceeding expectations, or playing to their potential. Do I think the 20th ranked class is absolutely and wholly better than the 30th ranked class? No. And I’ve said as much in 2 different ways. The rankings don’t take a team’s current depth or needs into account, and there’s not a ton of difference between the 20th and 30th ranks anyway. It’s generally only a difference of 1 or 2 players between 30 and 20. That’s why in a previous comment on this very thread I said I though the choices for the poll should’ve been changed. I think recruiting, just like the final rankings, generally line up in tiers. The top 5 have their own tier, 6-12 are all pretty close, 13-20 close again, 21-30, 31-40, etc. All of those tiers generally have a difference of 1 or 2 players between the last team in a tier and the first team.

But regardless of where a team finishes in recruiting, it’s up to them where they’ll end up in the final rankings. Having the recruiting rankings just lets us know where they stand relative to their talent.

 

[RonJohn]

While the math you did to come up with those averages is correct, it’s not how they’re ranked. Washington signed 23 players; 10 4-stars, 9 3-stars, and 4 2-stars for a star average of 3.26. TCU signed 19 total players; 1 5-star, 6 4-stars, 10 3-stars, and 2 2-stars for a star average of 3.32. Washington would’ve had a better star average if they just hadn’t signed those 2-star players, which is the biggest reason why I think using average stars is a flawed metric.

That would depend also on how many you count. I think the recruiting services only count the top 20. That actually hurts a team like TCU who only had 19 recruits because they lose points for not signing an extra person. If they only had 19 spots available because they were loaded with top "talent" and had no more scholarship spots available, does only signing 19 hurt the overall "talent" on their team?

As far as the rest of your comment is concerned, it doesn’t really have anything to do with the rankings themselves. Recruiting rankings aren’t a metric for a team’s success in the future because they aren’t supposed to be. They’re supposed to give a gauge on how much talent a team has. If anything they’re only useful to determine which teams are underperforming, exceeding expectations, or playing to their potential. Do I think the 20th ranked class is absolutely and wholly better than the 30th ranked class? No. And I’ve said as much in 2 different ways. The rankings don’t take a team’s current depth or needs into account, and there’s not a ton of difference between the 20th and 30th ranks anyway. It’s generally only a difference of 1 or 2 players between 30 and 20. That’s why in a previous comment on this very thread I said I though the choices for the poll should’ve been changed. I think recruiting, just like the final rankings, generally line up in tiers. The top 5 have their own tier, 6-12 are all pretty close, 13-20 close again, 21-30, 31-40, etc. All of those tiers generally have a difference of 1 or 2 players between the last team in a tier and the first team.

But regardless of where a team finishes in recruiting, it’s up to them where they’ll end up in the final rankings. Having the recruiting rankings just lets us know where they stand relative to their talent.

So the non-accurate measurements are actually valid because instead of measuring something that can be checked, such as future performance, they only measure a vague notion of "talent"? Is there any way to validate the "talent" measurement, or is it just something that can be argued about among fans? The recruiting sites are simply entertainment, nothing more.

You are saying about the same thing I was saying with respect to the tiers. However, I wouldn't even group them that close together. You can tell if a team is closer to Alabama or closer to North Texas. But you can do that without some points system that ranges from 700 to 3000 points. People can look at the height/weight and stats of the Alabama recruits and understand that they are better players than the North Texas players without some made up numbers to rank them.

I don't believe that the recruiting services and the rankings actually show where teams stand against each other in "talent". They are just entertainment trying to make money off of ads and subscriptions. There is no measurement that tells you how much talent a player has. There is no formula to tell you which player is better than another player. Even when they play in college, there is disagreement about who the better player is. Is there a formula or measurement to tell us with certainty who the best running back in the NCAA in a typical year? No. And that is with actual numbers in competition against teams who can be compared to each other using connected competitions. Sometimes there will be a running back in high school who stands out, such as Herschell or Marcus Dupree. Most years there is absolutely no way to know which running back is the best, or even to say that the number 1 ranked running back is definitely better than the number 15 ranked running back. The recruiting sites want people to believe their formulas and subscribe to their sites. But just putting a number on something doesn't make it an accurate measurement.

 

[JacketOff]

That would depend also on how many you count. I think the recruiting services only count the top 20. That actually hurts a team like TCU who only had 19 recruits because they lose points for not signing an extra person. If they only had 19 spots available because they were loaded with top "talent" and had no more scholarship spots available, does only signing 19 hurt the overall "talent" on their team?

This argument is in response to average stars being the preferred metric for measuring classes. Every player is counted to create the average stars metric, regardless of whether or not a team exceeds 20 signees. I literally said that, and that’s how the teams are listed in the rankings. It’s why I said points are the better metric, but points aren’t perfect either, because like you said TCU is at a disadvantage they didn’t sign 20 players. While that’s true, it’s irrelevant to my point about average stars, and if anything proves my point about how team rankings cannot be 100% accurate because depth/needs aren’t taken into account.

So the non-accurate measurements are actually valid because instead of measuring something that can be checked, such as future performance, they only measure a vague notion of "talent"? Is there any way to validate the "talent" measurement, or is it just something that can be argued about among fans? The recruiting sites are simply entertainment, nothing more.

You are saying about the same thing I was saying with respect to the tiers. However, I wouldn't even group them that close together. You can tell if a team is closer to Alabama or closer to North Texas. But you can do that without some points system that ranges from 700 to 3000 points. People can look at the height/weight and stats of the Alabama recruits and understand that they are better players than the North Texas players without some made up numbers to rank them.

I don't believe that the recruiting services and the rankings actually show where teams stand against each other in "talent". They are just entertainment trying to make money off of ads and subscriptions. There is no measurement that tells you how much talent a player has. There is no formula to tell you which player is better than another player. Even when they play in college, there is disagreement about who the better player is. Is there a formula or measurement to tell us with certainty who the best running back in the NCAA in a typical year? No. And that is with actual numbers in competition against teams who can be compared to each other using connected competitions. Sometimes there will be a running back in high school who stands out, such as Herschell or Marcus Dupree. Most years there is absolutely no way to know which running back is the best, or even to say that the number 1 ranked running back is definitely better than the number 15 ranked running back. The recruiting sites want people to believe their formulas and subscribe to their sites. But just putting a number on something doesn't make it an accurate measurement.

You lost me on this. I’m honestly confused about what your point is. The recruiting services provide an idea about where teams stand in talent. It’s not perfect, once again I’ve already said as much. I think we’re on the same page with a lot of what you’re saying, but you are trying to discredit the services because they don’t provide an actual measurable statistic. While I can see the value in the services because of the ideas they provide the general public about which teams are the most talented. Once again, they’re not perfect but they’re pretty good at what they do. The final AP rankings aren’t accurate just because they put a number next to a team’s name, just like the recruiting services aren’t accurate. But, once again, both polls/rankings give a general idea about where teams stand/perform. I don’t subscribe to any of the sites. I don’t care to. But I can see that the vast majority of the time their ratings and rankings are pretty accurate and provide some value for comparing teams.

 

[Pointer]

To me elite recruiting is top 15 or better. Multiple coaches have claimed that is what the goal is from CGC to CBK.

More importantly, the coaches have made it a goal of acc and national championships. Taking a quick look at the previous link with average class ranking over a 5 year period, the only teams making it or in serious contention for the playoffs are in the top 15 or better (Oregon being the exception with an average recruiting class rank of 17 over that period).

So far the polls here seem to indicate that many here don't have faith in being able to recruit high enough consistently to reach said goals. Or is it that we think Tech can somehow achieve those goals by beating the odds? Is my understanding of CGC's definition of elite recruiting wrong?

Am I missing something?

My goal with this thread is to get a good understanding of what metrics we can use to judge the program and coaches after they gave been given the proper amount of time to get established because it doesn't seem to be well defined.

 

[Oakland]

Looking at the results of the poll, I have to give credit to many Tech fans for being realistic. As we all know, Tech is a difficult school and does not have multiple majors. Plus, Tech has not been known as a football powerhouse that produces a lot of NFL players. Ratings of 20 to 29 every year would be super.

My opinion, all of the good work done by the current football staff will be lost by performance results. High school coaches and parents of recruits want their kid to get elite coaching during their college experience and that's not evident. The coaching staff has been disappointing. Recruits, parents and high school coaches are not going to get excited by 73-7 embarrassments, losses to teams like the Citadel and offensive shutouts (Temple and VPI). It's up to Collins to see that.

 

[takethepoints]

Almost everything I would have said has been said already by Ron. I would only add this: rankings of high school football players are inherently noisy. The people ranking them are (usually) fans attached to particular programs. When the kid expresses an interest in that school, the rater takes a look. They don't use uniform measures or scales when they do. Also, the raw data usually consists of highlight films, an inherently less useful source. If the kid goes to one of the camps, you can get something approaching actual physical data, but the comparisons of talents there are so unlike game conditions that it is of dubious use. I do agree that there is probably less noise of this sort in the Rivals ratings, but not enough to convince me.

Like I said, the 5 star kids stand out. I didn't have doubt one about Shamire; he'd start and soon. Calvin was similar. (I still remember what Chan said when he signed, "Coach Nix just became a much better football coach.") They aren't the problem. The difficulties come with the 4 star and even more 3 star players. Trying to get comparative rankings on them is, imho, almost impossible, given the way it is done now. It could be improved, but I doubt the rating sites will ever do that.

Does this mean I don't pay attention to the ratings sites? Of course not; that's where I gate most of the information about who Tech has signed and other pertinent data. But I really don't pay much attention to where Tech ends up in the recruiting rankings. The real question is whether the coaches think a player fits their needs and that he will stay in school long enough to contribute regularly. It is their ratngs that matter.

 

[boger2337]

My best bet is we average just outside the top 25. Say 26-30th. We will land a couple classes in the top 20, but for the most part with the 2 ranking sources I don't know if we ever land a concensus top 20 class. 247 seems to be tougher to crack this year.

I don't have faith in the staff to win the big games (uga, clemson, VT, Miami) consistently enough to get us into the top 15 range.

I will also go on a limb and say CGC never beats uga in his tenure. Cold take, but been thinking about that lately. I'm glad we moved on from the TO, but it was the only thing that gave us a shot. We won't beat them by brute force. Have to gimmick to beat them.

 

[boger2337]

To me elite recruiting is top 15 or better. Multiple coaches have claimed that is what the goal is from CGC to CBK.

More importantly, the coaches have made it a goal of acc and national championships. Taking a quick look at the previous link with average class ranking over a 5 year period, the only teams making it or in serious contention for the playoffs are in the top 15 or better (Oregon being the exception with an average recruiting class rank of 17 over that period).

So far the polls here seem to indicate that many here don't have faith in being able to recruit high enough consistently to reach said goals. Or is it that we think Tech can somehow achieve those goals by beating the odds? Is my understanding of CGC's definition of elite recruiting wrong?

Am I missing something?

My goal with this thread is to get a good understanding of what metrics we can use to judge the program and coaches after they gave been given the proper amount of time to get established because it doesn't seem to be well defined.

Collins thinks he can land top 10 classes.
I'm not sure he understands he has to win games to do so. Good coaching has to come before good recruiting. Otherwise good recruiting and bad losses will look extremely bad, basically showing we wasted their talent.

If we ruin this for Gibbs, Sims,, and Brooks we won't ever touch a top 15 class. Gibbs, Sims, and Brooks have to have a 9+ win season for recruits to have faith in the program.

 

[Pointer]

Almost everything I would have said has been said already by Ron. I would only add this: rankings of high school football players are inherently noisy. The people ranking them are (usually) fans attached to particular programs. When the kid expresses an interest in that school, the rater takes a look. They don't use uniform measures or scales when they do. Also, the raw data usually consists of highlight films, an inherently less useful source. If the kid goes to one of the camps, you can get something approaching actual physical data, but the comparisons of talents there are so unlike game conditions that it is of dubious use. I do agree that there is probably less noise of this sort in the Rivals ratings, but not enough to convince me.

Like I said, the 5 star kids stand out. I didn't have doubt one about Shamire; he'd start and soon. Calvin was similar. (I still remember what Chan said when he signed, "Coach Nix just became a much better football coach.") They aren't the problem. The difficulties come with the 4 star and even more 3 star players. Trying to get comparative rankings on them is, imho, almost impossible, given the way it is done now. It could be improved, but I doubt the rating sites will ever do that.

Does this mean I don't pay attention to the ratings sites? Of course not; that's where I gate most of the information about who Tech has signed and other pertinent data. But I really don't pay much attention to where Tech ends up in the recruiting rankings. The real question is whether the coaches think a player fits their needs and that he will stay in school long enough to contribute regularly. It is their ratngs that matter.

It's the chicken and egg situation. Are your class rankings good because the recruits are good, or are the recruits rated higher because the team they commit to is good? (Again like you, this is not directed at the obvious 5 star elite players)