The debate on stars is as old as the existence of recruiting rankings and message boards. There has always been the "stars don't matter - trust the coaches!" crowd banging heads with the "star gazers".
A few things that have been clear to anyone who has rolled up their sleeves and gotten into the numbers on this...
-Saying "Recruit X > Recruit Y because x=4 and y=3" is dumb. Star ratings are way too blunt an instrument to measure an individual recruit. 3* Mike Hart was better than his top 100 classmate, Max Martin. But recruit 10 Mike Harts and 10 Max Martins over a period of 20 years and 6-7 times out of 10, Max Martin will end up better.
-Teams with more 4* and 5* recruits will, over time, outperform teams with fewer. There's a reason Iowa and Wisconsin never play for the national championship while Bama, Georgia, Clemson, LSU and Ohio State often do and Pete Carroll's USC, Jimmy Johnson's Miami, Bobby Bowden's FSU often did. This happens because, on average, 5* recruits become college starters, college stars, and NFL players at a measurably higher rate than 4* recruits do, and the same is true for 4* recruits compared to 3*.
-While programs like Wisconsin and Oregon have marginally outperformed the recruiting services over time, and a team like Dantonio's MSU or last year's TCU can occasionally slip by the doorman and get into the Playoff club for a year... they're interlopers. One-offs. Outliers. Not a formula that can be replicated. And they basically never win the national championship.
-Loading up on 4* and 5* players does not guarantee you'll make the Playoff or win a national championship. However, NOT loading up on 4* and 5* talent all but guarantees that you won't.
So let's put some numbers to what it takes to win a national championship, recruiting-wise. Are there clear trends or thresholds that must be crossed in order to get into the national championship discussion?
The short answer: abso****inglutely.
The 2nd-3rd-4th year players are the core of any college team. So I looked at the composite star average for the 2nd-3rd-4th year players on every national championship team in the era of the recruiting services.
2022 Georgia 3.91
2021 Georgia 3.97
2020 Bama 3.99
2019 LSU 3.67
2018 Clemson 3.72
2017 Bama 4.03
2016 Clemson 3.53
2015 Bama 3.94
2014 OSU 3.76
2013 FSU 3.73
2012 Bama 3.74
2011 Bama 3.85
2010 Auburn 3.22
2009 Bama 3.51
2008 Florida 3.83
2007 LSU 3.69
2006 Florida 3.75
The mean here is a star average of 3.76 across all the 2nd, 3rd and 4th year players on these rosters.
A few notes on the above list...
Clemson: Their classes have been notably small, averaging about 20 per class leading into their better seasons, because their attrition rate has been extremely low. Almost no one transfers out of Clemson. They either make the NFL or stay 4-5 years. This low churn has been integral to their success. They have been the model of recruiting efficiency: very few misses, guys max out and see out their eligibility there.
FSU: The class that was in their 4th year on their 2013 Champions had 7 five-stars.
Auburn: The lone serious outlier here. Because of some unusual stuff that went on down there, they signed a MASSIVE number of players in the 2007 and 2008 classes, who were in their 3rd and 4th years in 2010. While they literally had no five-stars on the roster - the only national champion in the age of the services that didn't have a handful of them - they had an unusually veteran roster. Nearly 3/4 of their National Championship roster was 3rd-4th-5th year. On most teams, it’s close to half.
Bama: 2008 was when they became a recruiting behemoth. Not coincidentally, 2008 was Saban’s first recruiting class. It was a quantum leap from their prior classes that looked very much like ours or even a touch worse, and they’ve never looked back.
So how is Michigan doing on this measure?
Our 2nd through 4th year classes on recent/upcoming rosters...
2021 - 3.58
2022 - 3.65
2023 - 3.59
And this makes perfect sense! These numbers are very good, so it tracks that we'd be in the Playoff mix. But these numbers also would only win a national championship about every 8-9 years or so, and they're way off the top teams.
I know 3.6 to 3.75 doesn't seem like much, but let me lay it out so you can picture it on a roster:
.15 stars per player, across 75 players, is 11.25 stars.
That means: take the 11 sleepiest of sleeper recruits on your roster and replace them with 3 more 5* and 5 more 4*.
Or 11 more 4*.
The difference between 3.6 and 4.0 is 30 more stars. That means replacing the bottom quarter of your roster with a half dozen more 5* studs and 15 more 4*.
This is why you saw Georgia step on our necks in Miami: we brought a .22 (3.58) to a fight with an AR-15 (3.97). Where we had Dom Giudice and Nikhai Hill-Green, they had Jalen Carter and Nakobe Dean.
And this is why what happened in Glendale was tantamount to us shooting ourselves in the foot.
This is why I have been tracking the roster and noting the number of "bold type" (4* and 5*) players for the last decade.
This is why the weak 2023 class was a big deal... and why the 2024 rebound we're seeing is massive.
And I think we're going to be the best team in the country this season, with recruiting numbers in the core of our roster that would be the 4th worst for a national champ in the last 18 years... but they're at least in the range that we wouldn't be a massive outlier. Just beating the odds slightly.
A few things that have been clear to anyone who has rolled up their sleeves and gotten into the numbers on this...
-Saying "Recruit X > Recruit Y because x=4 and y=3" is dumb. Star ratings are way too blunt an instrument to measure an individual recruit. 3* Mike Hart was better than his top 100 classmate, Max Martin. But recruit 10 Mike Harts and 10 Max Martins over a period of 20 years and 6-7 times out of 10, Max Martin will end up better.
-Teams with more 4* and 5* recruits will, over time, outperform teams with fewer. There's a reason Iowa and Wisconsin never play for the national championship while Bama, Georgia, Clemson, LSU and Ohio State often do and Pete Carroll's USC, Jimmy Johnson's Miami, Bobby Bowden's FSU often did. This happens because, on average, 5* recruits become college starters, college stars, and NFL players at a measurably higher rate than 4* recruits do, and the same is true for 4* recruits compared to 3*.
-While programs like Wisconsin and Oregon have marginally outperformed the recruiting services over time, and a team like Dantonio's MSU or last year's TCU can occasionally slip by the doorman and get into the Playoff club for a year... they're interlopers. One-offs. Outliers. Not a formula that can be replicated. And they basically never win the national championship.
-Loading up on 4* and 5* players does not guarantee you'll make the Playoff or win a national championship. However, NOT loading up on 4* and 5* talent all but guarantees that you won't.
So let's put some numbers to what it takes to win a national championship, recruiting-wise. Are there clear trends or thresholds that must be crossed in order to get into the national championship discussion?
The short answer: abso****inglutely.
The 2nd-3rd-4th year players are the core of any college team. So I looked at the composite star average for the 2nd-3rd-4th year players on every national championship team in the era of the recruiting services.
2022 Georgia 3.91
2021 Georgia 3.97
2020 Bama 3.99
2019 LSU 3.67
2018 Clemson 3.72
2017 Bama 4.03
2016 Clemson 3.53
2015 Bama 3.94
2014 OSU 3.76
2013 FSU 3.73
2012 Bama 3.74
2011 Bama 3.85
2010 Auburn 3.22
2009 Bama 3.51
2008 Florida 3.83
2007 LSU 3.69
2006 Florida 3.75
The mean here is a star average of 3.76 across all the 2nd, 3rd and 4th year players on these rosters.
A few notes on the above list...
Clemson: Their classes have been notably small, averaging about 20 per class leading into their better seasons, because their attrition rate has been extremely low. Almost no one transfers out of Clemson. They either make the NFL or stay 4-5 years. This low churn has been integral to their success. They have been the model of recruiting efficiency: very few misses, guys max out and see out their eligibility there.
FSU: The class that was in their 4th year on their 2013 Champions had 7 five-stars.
Auburn: The lone serious outlier here. Because of some unusual stuff that went on down there, they signed a MASSIVE number of players in the 2007 and 2008 classes, who were in their 3rd and 4th years in 2010. While they literally had no five-stars on the roster - the only national champion in the age of the services that didn't have a handful of them - they had an unusually veteran roster. Nearly 3/4 of their National Championship roster was 3rd-4th-5th year. On most teams, it’s close to half.
Bama: 2008 was when they became a recruiting behemoth. Not coincidentally, 2008 was Saban’s first recruiting class. It was a quantum leap from their prior classes that looked very much like ours or even a touch worse, and they’ve never looked back.
So how is Michigan doing on this measure?
Our 2nd through 4th year classes on recent/upcoming rosters...
2021 - 3.58
2022 - 3.65
2023 - 3.59
And this makes perfect sense! These numbers are very good, so it tracks that we'd be in the Playoff mix. But these numbers also would only win a national championship about every 8-9 years or so, and they're way off the top teams.
I know 3.6 to 3.75 doesn't seem like much, but let me lay it out so you can picture it on a roster:
.15 stars per player, across 75 players, is 11.25 stars.
That means: take the 11 sleepiest of sleeper recruits on your roster and replace them with 3 more 5* and 5 more 4*.
Or 11 more 4*.
The difference between 3.6 and 4.0 is 30 more stars. That means replacing the bottom quarter of your roster with a half dozen more 5* studs and 15 more 4*.
This is why you saw Georgia step on our necks in Miami: we brought a .22 (3.58) to a fight with an AR-15 (3.97). Where we had Dom Giudice and Nikhai Hill-Green, they had Jalen Carter and Nakobe Dean.
And this is why what happened in Glendale was tantamount to us shooting ourselves in the foot.
This is why I have been tracking the roster and noting the number of "bold type" (4* and 5*) players for the last decade.
This is why the weak 2023 class was a big deal... and why the 2024 rebound we're seeing is massive.
And I think we're going to be the best team in the country this season, with recruiting numbers in the core of our roster that would be the 4th worst for a national champ in the last 18 years... but they're at least in the range that we wouldn't be a massive outlier. Just beating the odds slightly.
Last edited:
