Class of 2016 Interesting read for those anti-star ppl

[anomander]

Footballstudyhall always does a great job, and I stumbled across this read from after the Championship game. This article is pretty bulletproof, but fire away.

http://www.footballstudyhall.com/20...-matters-why-the-sites-get-the-rankings-right

 

[Libero2]

What he's saying makes perfect sense. At the individual level many mistakes are made, and are very easy to point out. But on the whole it's very easy to see they matter. I believe Bam was talking about this fact, that anybody can see how ridiculously far above most kids the true 4 and 5 star kids are. They are very easy to pick out for anybody, and it's in separating the next pack of kids that is much more difficult and much more difficult to project a kids ability to play successfully at the next level.

Also it was sad to see us in the absolute bottom tier in their rankings, and it shouldn't surprise anyone why we've been essentially terrible as a football team during that time frame.

 

[HOFCeluck]

Stars mean nothing because we're a program that does not bring in a lot of four and five star players.

 

[djorange1989]

If the folks who rank the kids are ranking them BEFORE Nick Saban and Urban Meyer come calling, then, sure, their rankings mean something. If those coaches come calling first and then the kid gets ranked 4 or 5 stars, big deal. No sh** that Saban and Meyer will be recruiting the best athletes. Pretty obvious there. It's simple, we need to recruit better and we need to start winning so that we can recruit better. It has improved for sure but it needs to keep improving, especially on the offensive side of the ball.

 

[phil77]

Stars only mean nothing when you aren't getting them.

 

[007]

Footballstudyhall always does a great job, and I stumbled across this read from after the Championship game. This article is pretty bulletproof, but fire away.

http://www.footballstudyhall.com/20...-matters-why-the-sites-get-the-rankings-right

Not only this article anomander, but every statistical analysis I have ever read on the predicative validity of star/class ranking systems have reached the same conclusions.

 

[blizz]

Thanks for this ano. Eye opening for sure. Wondering if anyone can explain the .500's at same star matchups, i know this is the result of a statistical control but I'm not following entirely.

 

[007]

Thanks for this ano. Eye opening for sure. Wondering if anyone can explain the .500's at same star matchups, i know this is the result of a statistical control but I'm not following entirely.

Not sure specifically from the linked article, but I would assume that it implies when a 3 star class played a 3 star class, or any same star class played each other, there is a 50% probability success rate. Same probability as calling heads or tails. Statistically, that means that there is no difference between the 2 variables. So, if star rankings are meaningful, and 2 similar ranked teams played each other, the chances of winning are 50/50.

 

[Bambrewer]

so they are saying when you put a bunch of good players together they win? crazy! Of course the top players being put together win games. the problem is sites are seeing who offers kids and giving them a rating after they see what schools are in on them. The problem is they aren't identifying talent they are giving you a result of someone elses work. or better yet they ARE NOT identifying the under the radar kid. They are weather men. The article even said win some lose some. Take a 3 star kid and plug him into 5 star spot on the ohio team last year do they still play well? Surround a 2 star kid with great players will he play better? JJ Watt 2 stars Tom Brady unrated Arron Rodgers unrated Jordy Nelson 2 stars Leveon Bell 2 stars. The list goes on and on. Stars are meaningless

 

[tep624]

Stars mean nothing because we're a program that does not bring in a lot of four and five star players.

That's basically the case though. Star rankings for 4 and 5 star blue chip type recruits are fairly accurate. We've played in the 2-3 star range for the better part of 12 years now and those rankings are highly subjective.

 

[LeMoyneCuse]

What he's saying makes perfect sense. At the individual level many mistakes are made, and are very easy to point out. But on the whole it's very easy to see they matter. I believe Bam was talking about this fact, that anybody can see how ridiculously far above most kids the true 4 and 5 star kids are. They are very easy to pick out for anybody, and it's in separating the next pack of kids that is much more difficult and much more difficult to project a kids ability to play successfully at the next level.

I think you hit the nail on the head. The Most Talented Players are so freaking obviously talented that the guys who rate them for websites can pick them out. These are the kinds of guys coaches can build systems around.

It gets to be a crap shoot when there's less separation and less obvious skill. These are the guys who fit into systems and flourish later.

 

[HOFCeluck]

That's basically the case though. Star rankings for 4 and 5 star blue chip type recruits are fairly accurate. We've played in the 2-3 star range for the better part of 12 years now and those rankings are highly subjective.

I'd say it's hard to get a read on them based on the sheer number of three star- caliber players alone.

It is easy to get it right when you're only labeling 20-25 kids a year as five star talents.

 

[Sudano]

so they are saying when you put a bunch of good players together they win? crazy! Of course the top players being put together win games. the problem is sites are seeing who offers kids and giving them a rating after they see what schools are in on them. The problem is they aren't identifying talent they are giving you a result of someone elses work. or better yet they ARE NOT identifying the under the radar kid. They are weather men. The article even said win some lose some. Take a 3 star kid and plug him into 5 star spot on the ohio team last year do they still play well? Surround a 2 star kid with great players will he play better? JJ Watt 2 stars Tom Brady unrated Arron Rodgers unrated Jordy Nelson 2 stars Leveon Bell 2 stars. The list goes on and on. Stars are meaningless

Stars are not meaningless you are just confusing what they do mean. Incorrectly using an individual player's ranking has nothing to do with the overall strength of overall recruiting classes. By highlighting the many lower rated players who become great ignore the many who do not. Its called statistical significance. 2-star recruits become great at such a low rate they have little significance on the field when their teammates aren't talented. One player per class who significantly outperforms his star rating has little affect on the field unless his teammates also perform at that level.
By the way Tom Brady was a top recruit:
'People often forget that Tom Brady was actually a highly rated prospect coming out of high school. While the star system didn’t exist at the time, and there were fafewer publications producing rankings, where he was rated was consistent with a modern day solid four-star prospect. After his redshirt year, Michigan signed top 100 prospect Jason Kapsner, however. This was then closely followed by the verbal pledge from local superstar Drew Henson and most fans had already written off Brady before he had a chance to throw a single pass. But at the end of the day, yes, Tom Brady was a highly rated prospect in high school and he certainly lived up to that billing'

 

[anomander]

so they are saying when you put a bunch of good players together they win? crazy! Of course the top players being put together win games. the problem is sites are seeing who offers kids and giving them a rating after they see what schools are in on them. The problem is they aren't identifying talent they are giving you a result of someone elses work. or better yet they ARE NOT identifying the under the radar kid. They are weather men. The article even said win some lose some. Take a 3 star kid and plug him into 5 star spot on the ohio team last year do they still play well? Surround a 2 star kid with great players will he play better? JJ Watt 2 stars Tom Brady unrated Arron Rodgers unrated Jordy Nelson 2 stars Leveon Bell 2 stars. The list goes on and on. Stars are meaningless

You listed the outliers, just like they said there are. But not really sure how you can say stars are meaningless? Maybe individually they don't mean much, but when you put them all together they mean a whole lot, and there is even proof to back that up. Where is your proof?

And I don't think anybody really cares where the recruiting site guys came up with the ratings? I surely don't, and I highly doubt many others do either. What I do care about is how many of those 4 star kids we get, and where our national recruiting ranking is. And just like you said the recruiting site guys see Meyer, Saban, and Fisher offering kids and give them 4 and 5 stars. Well guess what? I want those kids. So fine. I'll even give you the recruiting site guys may not know too much, but the coaches they "allegedly" follow sure do. So therefore give me as many as those kids as we can get, and give me a recruiting class in the top 40 and I guarantee you we have a much better chance at winning more.

 

[TheCusian]

To me - it's always us fans who think our coaches are somehow gaming the system (ie. better talent evaluators than everyone else) - but I'm confident that they would evaluate the 4/5 stars as such. Then it becomes id'ing the right 3 star guys that fit the system and outplay their ranking through hard work/development. That's where I think our coaches are pretty good.

We won't be nationally relevant until that momentum builds to where we are grabbing 3/4/5 star kids that fit, regularly. But you have to build to that point and momentum is a funny thing.

(Note: I do think if you hit on a 3-star QB, the job gets much easier - as the affect the game the most.)

 

[HOFCeluck]

so they are saying when you put a bunch of good players together they win? crazy! Of course the top players being put together win games. the problem is sites are seeing who offers kids and giving them a rating after they see what schools are in on them. The problem is they aren't identifying talent they are giving you a result of someone elses work. or better yet they ARE NOT identifying the under the radar kid. They are weather men. The article even said win some lose some. Take a 3 star kid and plug him into 5 star spot on the ohio team last year do they still play well? Surround a 2 star kid with great players will he play better? JJ Watt 2 stars Tom Brady unrated Arron Rodgers unrated Jordy Nelson 2 stars Leveon Bell 2 stars. The list goes on and on. Stars are meaningless

I don't think listing five guys over a fifteen year period is a great indictment of the star system.

 

[anomander]

To me - it's always us fans who think our coaches are somehow gaming the system (ie. better talent evaluators than everyone else) - but I'm confident that they would evaluate the 4/5 stars as such. Then it becomes id'ing the right 3 star guys that fit the system and outplay their ranking through hard work/development. That's where I think our coaches are pretty good.

We won't be nationally relevant until that momentum builds to where we are grabbing 3/4/5 star kids that fit, regularly. But you have to build to that point and momentum is a funny thing.

(Note: I do think if you hit on a 3-star QB, the job gets much easier - as the affect the game the most.)

Hey i'd love to be the outlier. Hopefully Shafer is the next Billy Snyder, and we are the next KST.

 

[OttoinGrotto]

But not really sure how you can say stars are meaningless? Maybe individually they don't mean much, but when you put them all together they mean a whole lot, and there is even proof to back that up.

Well, I feel vindicated.

 

[HOFCeluck]

You're chances for success are directly proportional to the number of highly regarded prospects you accumulate.

I don't see that being too controversial.

 

[Bambrewer]

I would say the 2 sides will never agree on this subject plain and simple.

 

[HOFCeluck]

That's cool, but one side is wrong plain and simple

 

[ImperialOrange]

It's really easy to be correct with star ratings in revisionist history. See the final recruiting class, look at the overall performance of schools over the last 10 year cycle, and assign stars accordingly. You'll be right more time than you'd be wrong even though you'd get outliers like Texas, Miami (fl), and Boise St. Even with them, it's easy to believe they will revert back to their median within the next 10 year cycle and you can say "See I told you!". Suffice to say it's easy to establish a transitive property if you can change the values later in the process.

The value's needed, that none of these site's keep (though ours now is), is the star ratings of players when they commit to school x, y, z vs what their final ratings are on signing day which is what they keep for records and what these articles quote.

 

[tep624]

I'd say it's hard to get a read on them based on the sheer number of three star- caliber players alone.

It is easy to get it right when you're only labeling 20-25 kids a year as five star talents.

Agreed. And that's why I feel that there is more than one argument here. The 4/5 rankings and the 2/3 rankings.

 

[baggerbob]

Hey i'd love to be the outlier. Hopefully Shafer is the next Billy Snyder, and we are the next KST.

Going by the article, it shows we still are in the rebuilding era, as far as the players recruited during that time. You can't say Shafer didn't get the job with a lot of work to do, and then tout this article as being correct. I agree with the ratings in the article, and have said we weren't turned around when Doug left.

 

[tep624]

Footballstudyhall always does a great job, and I stumbled across this read from after the Championship game. This article is pretty bulletproof, but fire away.

http://www.footballstudyhall.com/20...-matters-why-the-sites-get-the-rankings-right

All this proves is the predictive value of 4/5 star recruits. It still does not address the obvious bias in the system for the 2/3 star recruits, which is where we've played over the past 12 years and has been the source of about 95% of the complaints made about the star system. Jake Pickard and Steven Clark are 2 great examples from last year - did their Michigan and Florida offers, respectively, which suddenly made them jump from 2 star to 3 star recruits overnight make them any better of a prospective recruit?

There's 250 4 or 5 star recruits on an annual basis, that's enough for the top 25 recruiting teams to get 10 each. You do that over 4 years, and your 2 deep is full of 4 and 5 star recruits for the most part. It's not rocket science that you should have a good team with those recruiting results.