Why recruiting rankings matter

Donate To GopherHole


Didn't read anything after this:

"more lower-rated prospects are on the All-America teams than former five-stars."

See? Proof positive that stars mean nothing. In fact, it is irrefutable evidence that lower ranked players ALWAYS outperform higher ranked prospects! There can be nothing written, said, or illustrated to disprove this fact.
 

Unregistered User said:
Didn't read anything after this:

"more lower-rated prospects are on the All-America teams than former five-stars."

See? Proof positive that stars mean nothing. In fact, it is irrefutable evidence that lower ranked players ALWAYS outperform higher ranked prospects! There can be nothing written, said, or illustrated to disprove this fact.
Click to expand...
Sarcasm?
 

The Unanswered Question

One of the more interesting question is how does Boise State do so well with three star or lower recruits? For 2010, 2011, and 2012 they have had only one four star recruit and that was in 2010.

I would love to see an analysis of why some schools do better than their recruiting star rankings would predict. I believe Coach Kill amd Nothern Illinois would be one of them. If you check their classes they would rank behind Minnesota.
 


Killjoy said:
One of the more interesting question is how does Boise State do so well with three star or lower recruits? For 2010, 2011, and 2012 they have had only one four star recruit and that was in 2010.

I would love to see an analysis of why some schools do better than their recruiting star rankings would predict. I believe Coach Kill amd Nothern Illinois would be one of them. If you check of their classes they would rank behind Minneota.
Click to expand...
Part of it for Boise is they pick the players that aren't as wanted in California, where the lower-rated players can still be pretty good and someone on the recruiting services just didn't do as good of a job evaluating them as the Boise coaching staff.
 

Killjoy said:
One of the more interesting question is how does Boise State do so well with three star or lower recruits? For 2010, 2011, and 2012 they have had only one four star recruit and that was in 2010.

I would love to see an analysis of why some schools do better than their recruiting star rankings would predict. I believe Coach Kill amd Nothern Illinois would be one of them. If you check of their classes they would rank behind Minneota.
Click to expand...
Isn't the obvious answer better coaching/evaluation/player development? That said, they are still outliers. For every Boise State there are 20 other small FBS programs that stink. Your odds of winning will still increase exponentially when you're pulling in higher-rated talent. It's not an absolute, but few things in life are.
 

This is the endless chicken and the egg argument. Yes, having more stars correlates to having a better football team. How does a player get more stars? By getting offers from good football teams...

So what is it? The good players go to the good teams? Or the good teams identify good players?
 

John Galt said:
This is the endless chicken and the egg argument. Yes, having more stars correlates to having a better football team. How does a player get more stars? By getting offers from good football teams...

So what is it? The good players go to the good teams? Or the good teams identify good players?
Click to expand...

Did you even read the article?
 



dpodoll68 said:
Did you even read the article?
Click to expand...
I don't think reading the article changes his question at alll. I very much agree that while "stars" can be predictive of the quality of player, recruiting services get far too much credit for the identification of said talent. It is, at its heart, still a chicken and an egg argument.

If a kid is from Ohio and has offers from OSU, Notre Dame, Michigan, and Michigan State, I don't need the recruiting services to tell me he's a "4 or 5 star player".

Likewise, if a kid is from Ohio and only has MAC offers, I don't need recruiting services to tell me he's a "2 star player".

Until there is a strong deviation from that basic formula, I'll still hold that the star system is still largely a reflection of a kid's offer list. The only time when this doesn't seem to be the case is when a kid commits so early that other schools don't have the time to offer.
 

If you can't get the 5-star athletes, there are other strategies you can attempt. You can scout better, and look for the hidden gems. Another is to adapt your game to what you can get.
 

We should just forfeit. Way to buy the Rivals BS!
Drop the five stars and see how they do.
 

corcoran1 said:
We should just forfeit. Way to buy the Rivals BS!
Drop the five stars and see how they do.
Click to expand...

Coaches would rather have higher rated athletes than lower rated athletes. Teams that have higher recruit rankings generally do better than those with lower recruit rankings.
 



Gopherprof said:
I don't think reading the article changes his question at alll. I very much agree that while "stars" can be predictive of the quality of player, recruiting services get far too much credit for the identification of said talent. It is, at its heart, still a chicken and an egg argument.

If a kid is from Ohio and has offers from OSU, Notre Dame, Michigan, and Michigan State, I don't need the recruiting services to tell me he's a "4 or 5 star player".

Likewise, if a kid is from Ohio and only has MAC offers, I don't need recruiting services to tell me he's a "2 star player".

Until there is a strong deviation from that basic formula, I'll still hold that the star system is still largely a reflection of a kid's offer list. The only time when this doesn't seem to be the case is when a kid commits so early that other schools don't have the time to offer.
Click to expand...

Gopherprof is right here - these recruiting services largely just follow who schools are recruiting. Do they do their own evaluations and tweak their rankings accordingly? I take them at their word that they do. But MV has done his own analysis and determined that recruiting rankings follow on-field rankings, not the other way around.

Anyway, I think one thing we all should have learned from the Brewster era is that recruiting is just one piece - and probably not the most important piece - of developing a quality football program. Someone pointed out Boise St, but there are many other examples. I'll just point out one: from 2008-2010, there was a team from the MAC that went 2-3 against Big 10 teams and all 3 of the losses were close (actually the total points from all game was exactly even). You could argue that maybe the B1G teams weren't as "up" to play against a MAC team, but I'd also point out that all 5 games were at the B1G team's home stadium. Now how did this team manage to hang with these teams when players on MAC teams rarely - if ever - have any scholarship offers from BCS level programs?

Bottom line, I'm much less concerned about what star ratings our players have than I used to be. Recruiting is interesting, and I enjoy following the ups and downs, but I don't take it so seriously anymore. I'd rather have higher-ranked guys than lower-ranked guys, but in reality we won't know how good these guys are until we see them on the field.
 

http://collegefootballmatrix.wordpress.com/articles-2/the-before-during-and-after-hogs-profiled/

This is an article about how Bobby Petrino added more value in terms of wins/losses above the talent level of his team than any other SEC coach during his tenure at Arkansas.

In regards to this never-ending argument, this particular line jumped off the page at me:

"In the SEC in 2011, when adjusted for the field the game is played on, [it] was 81% correct in predicting every SEC game 6 months before the season started."

The implied "it" there (missing in the original text) is referring to the 4-year moving average of each team's composite recruiting ranking. In other words, if you looked at the team's 4-year moving average after NSD 2011, and used nothing more than that (and the standard homefield advantage/disadvantage) to predict the outcomes of the 2011 SEC conference games, you would correctly predict the winner 81% of the time. Anyone who gambles knows that an ability to predict the winner 81% of the time is practically a license to print money.

But, again - recruiting rankings are meaningless.
 


Gopherprof said:
I don't think reading the article changes his question at alll. I very much agree that while "stars" can be predictive of the quality of player, recruiting services get far too much credit for the identification of said talent. It is, at its heart, still a chicken and an egg argument.

If a kid is from Ohio and has offers from OSU, Notre Dame, Michigan, and Michigan State, I don't need the recruiting services to tell me he's a "4 or 5 star player".

Likewise, if a kid is from Ohio and only has MAC offers, I don't need recruiting services to tell me he's a "2 star player".

Until there is a strong deviation from that basic formula, I'll still hold that the star system is still largely a reflection of a kid's offer list. The only time when this doesn't seem to be the case is when a kid commits so early that other schools don't have the time to offer.
Click to expand...
Well said, with the caveate that sometimes kids that develop relatively late, but commit early don't get that many offers (Bielima's point, and I think he is right).

It also appears that the recruiting services are much better at telling the differences between recruits at the high school level than judging their ultimate potential. Basically my take is that the recruiting rankings tell one more about the readiness of recruits to play early as it does their ultimate potential.
 

dpodoll68 said:
In other words, if you looked at the team's 4-year moving average after NSD 2011, and used nothing more than that (and the standard homefield advantage/disadvantage) to predict the outcomes of the 2011 SEC conference games, you would correctly predict the winner 81% of the time.
Click to expand...
Is this statistically significant compared to other factors? That's the sort of detail that would make this fact more impressive. Recruiting rankings are not meaningless (and argument that is just silly and asinine), but this stat doesn't say how meaningful they are either (the more interesting/important topic of discussion).
 

WOW!

dpodoll68 said:
http://collegefootballmatrix.wordpress.com/articles-2/the-before-during-and-after-hogs-profiled/

This is an article about how Bobby Petrino added more value in terms of wins/losses above the talent level of his team than any other SEC coach during his tenure at Arkansas.

In regards to this never-ending argument, this particular line jumped off the page at me:

"In the SEC in 2011, when adjusted for the field the game is played on, [it] was 81% correct in predicting every SEC game 6 months before the season started."

The implied "it" there (missing in the original text) is referring to the 4-year moving average of each team's composite recruiting ranking. In other words, if you looked at the team's 4-year moving average after NSD 2011, and used nothing more than that (and the standard homefield advantage/disadvantage) to predict the outcomes of the 2011 SEC conference games, you would correctly predict the winner 81% of the time. Anyone who gambles knows that an ability to predict the winner 81% of the time is practically a license to print money.

But, again - recruiting rankings are meaningless.
Click to expand...

Wow, that is pretty powerful stuff. Very interesting read. One thought of many: So, Arkansas got better, won more games, had a "great" coach but they couldn't recruit any better. Kinda scarey.

The majority of us are very optimistic about Gopher football. But for this to work like we think it is going to...it is going to be an aberation. Or unlike Boise and Arkansas we do recruit better going forward. Or Jerry Kill is a great coach?
 

Badger2010 said:
Agree with what you are saying about recruiting rankings, but predicting the winner of a game 81% of the time is actually not amazingly difficult when it comes to college football games. For gambling, it's more about whether a team will cover the spread or not.

Brian Bennett and Adam Rittenberger, the ESPN Big Ten bloggers, correctly predicted the winner in Big Ten games 76% of the time and 73% of the time, respectively.
Click to expand...

It's not amazingly difficult if you know something about the teams. If all you knew was team A vs team B, you could expect to be right 50% of the time. AR and BB were able to do better than that because they know info about the teams (probably most notably the talent levels) and not blindly guessing. The recruiting site study shows that they are a reliable source of info.
 

FreakyDeke said:
It's not amazingly difficult if you know something about the teams. If all you knew was team A vs team B, you could expect to be right 50% of the time. AR and BB were able to do better than that because they know info about the teams (probably most notably the talent levels) and not blindly guessing. The recruiting site study shows that they are a reliable source of info.
Click to expand...

Oh I agree completely. The only thing I was trying to say was that correctly predicting the outcome of a college football game 81% of the time is not actually a license to print money. Now if you can correctly predict whether or not a team is going to cover the spread 81% of the time....then you a set for life.
 

Gopherprof said:
Isn't the obvious answer better coaching/evaluation/player development? That said, they are still outliers. For every Boise State there are 20 other small FBS programs that stink. Your odds of winning will still increase exponentially when you're pulling in higher-rated talent. It's not an absolute, but few things in life are.
Click to expand...

About as good a response as it gets for a 10K view. Boise has been nothing short of amazing playing AQ schools over the last 5+ years. But also consider, outside TCU, they have the biggest gap in 4 year recruiting ranking over the rest of their conference. It was even bigger in the WAC. They were supposed to go 10-2 or better every year with their talent and conference. Give the Gophers the Boise or TCU schedule and they are 10-2 or better and in the top 25 as well every year.

Fortunately the Matrix system is not more accurate in its pre-season game picks. FWIW - That 81% was from all games picked in April last year not each week like the ESPN bloggers.

Love the thread and conversation. Nice to get fans perspective on recruiting and winning games.

- Dave
 


I was looking at it through the TCU schedule and average coaching at Minnesota. Their 3-9 record last year was as predicted. Straight recruiting put them at 5-7. Home field adjustment dropped it to 4-8 and a new coach averages -1 game lost to the prediction. So the 3-9 was a 'best odds'. One team or one year doesn't make for a good sample size but given a average zero effect game coach with a field adjusted recruiting ranking of #52, Gopher fans should expect a 10-2 season with the 2011 TCU schedule. It is just trends and odds. Any team can still catch lightning in a bottle in any given season. However, it is usually the teams with positive effect game coaches (as aforemention Vandy and Wake) that make unlike runs above their talent levels.
 

cfbmatrix said:
I was looking at it through the TCU schedule and average coaching at Minnesota. Their 3-9 record last year was as predicted. Straight recruiting put them at 5-7. Home field adjustment dropped it to 4-8 and a new coach averages -1 game lost to the prediction. So the 3-9 was a 'best odds'. One team or one year doesn't make for a good sample size but given a average zero effect game coach with a field adjusted recruiting ranking of #52, Gopher fans should expect a 10-2 season with the 2011 TCU schedule. It is just trends and odds. Any team can still catch lightning in a bottle in any given season. However, it is usually the teams with positive effect game coaches (as aforemention Vandy and Wake) that make unlike runs above their talent levels.
Click to expand...

Ah gotcha, I thought you meant if Minnesota had their schedule the past few years they would've been 10-2 and in the top 25 each year. I was going to say, I'm sure the Gopher's record would have been significantly better, but 10-2 and top 25 each year is definitely a bit much.
 

dpodoll68 said:
http://collegefootballmatrix.wordpress.com/articles-2/the-before-during-and-after-hogs-profiled/

This is an article about how Bobby Petrino added more value in terms of wins/losses above the talent level of his team than any other SEC coach during his tenure at Arkansas.

In regards to this never-ending argument, this particular line jumped off the page at me:

"In the SEC in 2011, when adjusted for the field the game is played on, [it] was 81% correct in predicting every SEC game 6 months before the season started."

The implied "it" there (missing in the original text) is referring to the 4-year moving average of each team's composite recruiting ranking. In other words, if you looked at the team's 4-year moving average after NSD 2011, and used nothing more than that (and the standard homefield advantage/disadvantage) to predict the outcomes of the 2011 SEC conference games, you would correctly predict the winner 81% of the time. Anyone who gambles knows that an ability to predict the winner 81% of the time is practically a license to print money.

But, again - recruiting rankings are meaningless.
Click to expand...

So are you predicting the quick demise of Jerry Kill unless his Rivals recruiting rankings drastically improve the next 2 years? They finished 12th in 2012 and 9th in 2011. If the current trend holds up, their average recruiting class will be 10.5 in the B1G, meaning they should win 1-2 conference games at most per year.

College football isn't like the NFL, most posters on here could easily pick 3/4 games right. It doesn't take a ton of knowledge to predict ahead of time that USC will finish 10-2 (+/- 1 game) next year.
 

John Galt said:
So are you predicting the quick demise of Jerry Kill unless his Rivals recruiting rankings drastically improve the next 2 years? They finished 12th in 2012 and 9th in 2011. If the current trend holds up, their average recruiting class will be 10.5 in the B1G, meaning they should win 1-2 conference games at most per year.
Click to expand...

It's not a death sentence, but yes, it will make it much more difficult to be a consistent contender if the recruiting rankings do not pick up. Teams can and do outperform their recruiting rankings (Wisconsin, for example), but it is extremely difficult to do so for an extended period of time. Coaching matters (more than most people want to admit), but even the best coaches on Earth can't consistently overcome vast disparities in talent. Put Tim Brewster and staff in place of Brady Hoke and staff on Michigan last year and they still beat Minnesota 95 times out of 100.

John Galt said:
College football isn't like the NFL, most posters on here could easily pick 3/4 games right.
Click to expand...

All games? Possibly. Intra-conference games only? Very doubtful.

John Galt said:
It doesn't take a ton of knowledge to predict ahead of time that USC will finish 10-2 (+/- 1 game) next year.
Click to expand...

Exactly. All you need to know is that USC is a top 5 recruiting team virtually every season, and the reasonable expectation on that basis alone is that they will win a ton of games every year. That's the whole point.
 

You must log in or register to reply here.



Top Bottom