Jump to content


Our Class Ranking?


Recommended Posts


So, if I'm reading that right, there is more correlation to recruiting in the same year than in previous years. Thus, would tend to show that success affects recruiting (more correlation in the same year, whose recruits have nothing to do with that year's on-field success) more than recruiting affects success (less correlation in previous years who's recruits are directly responsible for on-field success). Would that be correct (condeding your first two sentenses)?

Missed this question.

 

I don't think success affects recruiting more than recruiting affects success, because the correlations for recruiting rank in 2009 and 2010 (with a majority of those players seeing the field) are still pretty strong [depending on your discipline], and the difference between those correlations aren't quite statistically significant. Rather, I think that it's a virtuous cycle: for the most part, if you recruit well, you play well. If you play well, you recruit well.

I was doing some research for a slightly different question and found some interesting tidbits. I was looking at how the recruiting classes compared in the B1G since 2008 - including Nebraska over that time even though we obviously weren't in the conference the entire time. I used Rivals team rankings.

 

- Michigan had been #2 each year except for #3 in 2011. They fell to #4 this year.

- Nebraska had been #3-2-3-3 the previous four years until falling to #5 this year.

- Michigan State had been #5 for the previous four years before jumping to #2 this year.

- Minnesota has been #12 the previous two years but moved up to #8 this year.

- The other schools were either similar or all over the board.

 

There is nothing definitive about that and it is a small sample size but that is a noticeable (if unscientific) correlation indicating that recruiting improves when you do well on the field and declines when you don't. There was really no indication in recruiting that Michigan St. would suddenly be a Top 5 (in the country) team, definitely no indication that Wisconsin should have gone to the previous three Rose Bowls (they were #6-8-11-7-8 in recruiting in the conference from 08-12) or that Ohio State should have fallen on their face a couple years ago (they are basically #1 every year).

 

I'm not trying to make the case that recruiting doesn't matter - it definitely does. I just think that it tends to follow success rather than lead it. You happen to find a few gems in recruiting, get a couple breaks, win a few extra games and you start to get some better recruits and the process snowballs.

Link to comment

 

Recruiting3.jpg

 

 

You lost me at this projection. That red line is more or less pulled out of thin air.

The red line is a linear extrapolation of the blue line.

 

In general, this thread should really stop trying to discuss correlation vs. causation and statistics in general. It's painfully obvious that the hard hitting analysis required to answer that question isn't offered until the graduate level pyschology courses.

Link to comment

A pretty strong correlation exists, there's really no debating that.

 

Just to add further evidence to KJ's point:

 

For the 2012-2013 season, I took a look at the correlation between wins that season and recruiting rankings (total points according to Rivals) for 2008, 2009, 2010, 2011, and 2012 [the five classes of players who were a part of the 2012-2013 season].

 

For 2008: r = .39

For 2009: r = .36

For 2010: r = .35

For 2011: r = .42

For 2012: r = .44

 

All of those are strong correlations: teams who recruit well also win more games and are ranked higher as a result. Furthermore, teams who recruited well in the past tended to recruit well in the future too; all the intercorrelations between the recruiting ranks were extraordinarily high. Not eye opening information.

Isn't r>0.6 usually considered strong correlation?

 

Depends on your discipline. To me, anything > .30 is strong.

 

BBB12, overall I agree with the overall point you're trying to make because unless someone is completely oblivious to how the world works, they can't argue against it.

 

But please tell the the discipline in which you consider an R^2 as "strong." I'm literally dying to hear what industry or company is using that in their analysis so I can drop my net worth on shorting that stock.

Link to comment

A pretty strong correlation exists, there's really no debating that.

 

Just to add further evidence to KJ's point:

 

For the 2012-2013 season, I took a look at the correlation between wins that season and recruiting rankings (total points according to Rivals) for 2008, 2009, 2010, 2011, and 2012 [the five classes of players who were a part of the 2012-2013 season].

 

For 2008: r = .39

For 2009: r = .36

For 2010: r = .35

For 2011: r = .42

For 2012: r = .44

 

All of those are strong correlations: teams who recruit well also win more games and are ranked higher as a result. Furthermore, teams who recruited well in the past tended to recruit well in the future too; all the intercorrelations between the recruiting ranks were extraordinarily high. Not eye opening information.

Isn't r>0.6 usually considered strong correlation?

 

Depends on your discipline. To me, anything > .30 is strong.

 

BBB12, overall I agree with the overall point you're trying to make because unless someone is completely oblivious to how the world works, they can't argue against it.

 

But please tell the the discipline in which you consider an R^2 as "strong." I'm literally dying to hear what industry or company is using that in their analysis so I can drop my net worth on shorting that stock.

 

I'm confused by your statement. Are you talking about the whole r = .40 being considered strong, or are you insinuating that I'm implying that all of these numbers actually prove anything? If it's the first, a lot of the social sciences do consider r = .40 to be a pretty strong effect, especially if it has enough power (a large enough sample). If it's the latter, then your insinuation is incorrect. I have not used causal language in any of these posts. I have not said that good recruiting causes more wins, nor have I said that winning causes good recruiting. In fact, you can't really imbue cause to anything in college football; it's impossible to randomly assign and there's no manipulation. You can't put Alabama in the crappy recruiting condition, and you can't put Wyoming in the a lot of wins condition. All you have to go on is associative relationships, which are "better than nothing." You can say that good recruiting leads to wins. You can say that a lot of wins leads to good recruiting; and I've said that I think that it is a bidirectional relationship.

 

I know that I've said nothing in this thread that should come as a surprise. I was just looking at the relationship between the two variables. A lot of people say that our recruiting rank doesn't matter, and all I'm doing (which was just adding on to KJ's point) is showing that there's a positive linear relationship between those two variables; that our recruiting rank means more than some of us would like to think.

 

That is all. Nothing more. Nothing less.

Link to comment

By the way, here's the correlation between 2012 wins and the average of the average recruit rank from Rivals for the past 5 classes: r = .42.

 

Interestingly enough, when you throw that composite into (what is probably an oversimplified) a regression model with pythagorean wins, the composite is still a significant predictor; which means [and please do correct me if I'm wrong] that there's an aspect of the composite rank that uniquely contributes to a team's success. The individual rankings didn't contribute whatsoever.

 

Recruiting services, no matter how biased we may claim them to be, actually do a good job in predicting future success.

Link to comment

So, if I'm reading that right, there is more correlation to recruiting in the same year than in previous years. Thus, would tend to show that success affects recruiting (more correlation in the same year, whose recruits have nothing to do with that year's on-field success) more than recruiting affects success (less correlation in previous years who's recruits are directly responsible for on-field success). Would that be correct (condeding your first two sentenses)?

Missed this question.

 

I don't think success affects recruiting more than recruiting affects success, because the correlations for recruiting rank in 2009 and 2010 (with a majority of those players seeing the field) are still pretty strong [depending on your discipline], and the difference between those correlations aren't quite statistically significant. Rather, I think that it's a virtuous cycle: for the most part, if you recruit well, you play well. If you play well, you recruit well.

I was doing some research for a slightly different question and found some interesting tidbits. I was looking at how the recruiting classes compared in the B1G since 2008 - including Nebraska over that time even though we obviously weren't in the conference the entire time. I used Rivals team rankings.

 

- Michigan had been #2 each year except for #3 in 2011. They fell to #4 this year.

- Nebraska had been #3-2-3-3 the previous four years until falling to #5 this year.

- Michigan State had been #5 for the previous four years before jumping to #2 this year.

- Minnesota has been #12 the previous two years but moved up to #8 this year.

- The other schools were either similar or all over the board.

 

There is nothing definitive about that and it is a small sample size but that is a noticeable (if unscientific) correlation indicating that recruiting improves when you do well on the field and declines when you don't. There was really no indication in recruiting that Michigan St. would suddenly be a Top 5 (in the country) team, definitely no indication that Wisconsin should have gone to the previous three Rose Bowls (they were #6-8-11-7-8 in recruiting in the conference from 08-12) or that Ohio State should have fallen on their face a couple years ago (they are basically #1 every year).

 

I'm not trying to make the case that recruiting doesn't matter - it definitely does. I just think that it tends to follow success rather than lead it. You happen to find a few gems in recruiting, get a couple breaks, win a few extra games and you start to get some better recruits and the process snowballs.

 

I tend to agree with that, but it would also help if this staff would do more recruiting during the season. This is no doubt that NU is at a disadvantage due to it's location, (which is vastly overused as a crutch by the pro Pelini crowd IMO) but it can and has been overcome before when anyone not named Solich or Pelini were head coaches. This staff has proven that they can recruit before and after the season. The problem lies when you know as a coaching staff that you have to outwork other top tier programs for the top recruits, and you just don't try, that is where the problems lay with this staff on this subject. They also have a history of not keeping in touch with the recruits that commit before the season and allow other coaches to change thier mind.

 

It all comes down to effort and if Bo wants to stay as the HC of NU, that effort on the recruiting trail needs to be raised year round by the entire staff. If they don't like it, then they are in the wrong profession. I guess what I am saying is that this staff doesn't give themselves the chance to put themselves in the position that you stated in your last sentence to get better recruits due to the lack of effort.

Link to comment

 

 

I'm confused by your statement. Are you talking about the whole r = .40 being considered strong, or are you insinuating that I'm implying that all of these numbers actually prove anything? If it's the first, a lot of the social sciences do consider r = .40 to be a pretty strong effect, especially if it has enough power (a large enough sample). If it's the latter, then your insinuation is incorrect. I have not used causal language in any of these posts. I have not said that good recruiting causes more wins, nor have I said that winning causes good recruiting. In fact, you can't really imbue cause to anything in college football; it's impossible to randomly assign and there's no manipulation. You can't put Alabama in the crappy recruiting condition, and you can't put Wyoming in the a lot of wins condition. All you have to go on is associative relationships, which are "better than nothing." You can say that good recruiting leads to wins. You can say that a lot of wins leads to good recruiting; and I've said that I think that it is a bidirectional relationship.

 

I know that I've said nothing in this thread that should come as a surprise. I was just looking at the relationship between the two variables. A lot of people say that our recruiting rank doesn't matter, and all I'm doing (which was just adding on to KJ's point) is showing that there's a positive linear relationship between those two variables; that our recruiting rank means more than some of us would like to think.

 

That is all. Nothing more. Nothing less.

 

I was not referring to your second statement and my point is closer to the first statement. Again, I agree with what you're trying to point out, but I was honestly just curious what discipline considered r = .3 as "strong" as you stated.

 

If all those social scientists out there consider that strong, well, ok. I will take your word for it.

Link to comment

They also have a history of not keeping in touch with the recruits that commit before the season and allow other coaches to change thier mind.

Feel free to post the statistics you have to prove this happens more to us than anyone else.

I'll even make it easier for you (since I know you have no such statistics): Simply go through the decommits we had this year and explain why they decommitted.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...