Our Class Ranking?

[robsker]

looking at the data in this thread re: where numerically NU has finished the season in the rankings vs. the recruiting rankings one can only give a loose assessment --- nothing overly detailed can be concluded. The loose assessment is this... of late (past 6 years or so) generally NU recruits outside the nations top 20 and generally NU finishes the season ranked below #20 as well. NU is not a top 20 program any longer --- if say a 10 year window of assessment is in view. Of course, historically NU is a top 5 program (if the window of assessment extends to the last 50 years). But currently, NU is outside the top 20 looking in --- both in recruiting and in on-field performance.

You need to show that the correlation holds when NU is in the top 20. As in NU has top 20 recruiting classes when NU is ranked in the top 20 and vice versa. Otherwise you're not making a point but just pointing out stuff.

Actually that was precisely my point. The data is not sufficient to make concrete defensible correlations that can lead to a "point" that is anywhere near definitive. One is left only with broad impressions --- no correlation between recruiting class and final ranking was attempted. All one can do is say, generally, that NU's recruiting rankings in the broadest sense over the recent years is relatively speaking about where season ending team ranking resides.

Interestingly, when one looks at NU's glory years --- say 1970 - 1999 --- the season ending rankings tended to be higher than the recruiting ratings (though make no mistake, NU recruited well). Seemingly... and again only a broad generalism can be made --- NU performed on the field better than (and at times quite a bit better than) recruiting rankings predicted. That seemingly is not the case now.

 

[robsker]

Nebraska was not historically a top five recruiting team either, so that throws a pretty good size wrench in that argument.

No argument was posited at all. I did not attempt a quantitative correlation between recruiting ranking and seasons ending ranking. I simply said that NU is not a top 20 program at present --- at least as measured by season's ending rankings --- when measured over the past 10 year window. I also said, generally recent recruiting ratings and season's ending rankings are fairly comparable. WRT the glory years of 1970-1999 NU was a top 5 overall program (based on season rankings) and recruited well too... but not a top 5 recruiting program... so I agree. NU played in that period above what recruiting rankings would have predicted. It should be noted however that NU did recruit well in that window (albeit not top 5)

 

[dylan]

So then you agree that it's possible to out perform recruiting rankings and that low rankings in no way dooms us to low results on the field, where it actually matters.

 

[robsker]

So then you agree that it's possible to out perform recruiting rankings and that low rankings in no way dooms us to low results on the field, where it actually matters.

I do in fact agree with your premise. It is certainly possible to out perform recruiting ratings. Many programs do. Also, it is possible to under perform relative to recruiting ranking --- I would hold recent North Carolina, Texas, USC, Michigan and Florida as program examples.

Like you Dylan, I concur that on field performance is the metric of value. Recruiting rankings are fun though because it gives us something to talk about after the season ends. And... such rankings are so speculative that ample grounds for all sorts of interesting interchange exists.

In this class, the film on a few of the guys stands out to me and I am excited about --- of course, no telling how they will do at the next level, but McClain, Wills, Newell, and a guy not too many have spoken about Pierson-Els all look like hopefuls (others too... but these guys on film looked particularly good).

 

[dylan]

Pierson-Els is the receiver I expect the most from, unless Harrison actually makes it to campus.

 

[robsker]

Pierson-Els is the receiver I expect the most from, unless Harrison actually makes it to campus.

Oh yes. Harrison does look great... I just did not comment on him because many assert he will not actually arrive here. But if he does... and who knows, he might, then I agree he is the most promising WR on film. The Peirson-El film has not generated many posts worth of comment... but he is shifty and quick... could be a great returned (if he has hands).

 

[Danny Bateman]

Agreed on Pierson-El. Even I sometimes forget he's a part of this class. I feel he's strongly underappreciated and even almost taken for granted.

Amidst the news Stewart may not make it, I think a lot of us forgot we already have Pierson-El committed and he's likely even more versatile and able to create mismatches. Very glad to have such a fluid athlete for the skill positions.

 

[It'sNotAFakeID]

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.

 

[dylan]

Correlation vs. causation? I would say the greater portion is indeed causative, but you can't deny that teams that have won historically end up with their recruits ranked higher because of their prestige. In other words, their winning percentage artificially inflates their recruit rankings to some extent. We've seen that an offer from Nebraska has this effect on kids with relatively low offers. Teams with even more prestige presumably tend to inflate rankings even more.

I also think the rankings skew towards populated areas and the south, artificially inflating rankings, but that's a bit off topic.

 

[It'sNotAFakeID]

Correlation vs. causation? I would say the greater portion is indeed causative, but you can't deny that teams that have won historically end up with their recruits ranked higher because of their prestige. In other words, their winning percentage artificially inflates their recruit rankings to some extent. We've seen that an offer from Nebraska has this effect on kids with relatively low offers. Teams with even more prestige presumably tend to inflate rankings even more.

I also think the rankings skew towards populated areas and the south, artificially inflating rankings, but that's a bit off topic.

Just correlation. As for the prestige factor [recruits go to teams who win more], I also looked at the correlation between wins in the 2012-2013 season and class ranking in 2012 and controlled that for wins in the 2011-2012 season; it looks at the correlation between the two if everyone had won the same amount of games.

The correlation is still pretty strong [r = .28], but it is significantly lower than the .44 value given above.

 

[RedDenver]

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?

 

[It'sNotAFakeID]

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.

 

[It'sNotAFakeID]

Here's what happens when you control each the relationship between a season's recruiting rank with the wins of that season (i.e. 2009 recruiting ranking and wins in 2009-2010) for the wins in the previous season.

2009 RRank & 2009 wins controlling for 2008 wins: r = .214

2010 RRank & 2010 wins controlling for 2009 wins: r = .24

2011 RRank & 2011 wins controlling for 2010 wins: r = .265

2012 RRank & 2012 wins controlling for 2011 wins: r = .287

All of those correlations are still significant, despite removing last season's win from the equation. Last season's success is still important; all the values decreased significantly when last year's win total was controlled. So I think it's safe to say, and again this isn't eye opening information, that teams who have a good season, then follow it up with a good recruiting class, will have a good season the next season as well.

Recency seems to play more of a factor than primacy in shaping a recruit's perception of a school. How a school does when a player first starts to care about a school doesn't matter as much as how a school performs when that player has to make his decision. The correlations between recruiting rank and wins the same season increase (with the exception between 2009 and 2010 which stayed the same) as the season controlled for gets further away from the actual year.

2012 RRank and 2012 wins controlling for 2011 wins: r = .287

2012 RRank and 2012 wins controlling for 2010 wins: r = .340

2012 RRank and 2012 wins controlling for 2009 wins: r = .339

2012 RRank and 2012 wins controlling for 2008 wins: r = .381

Yet, even when the relationship is controlled for wins in the past 4 seasons, the correlation is still significant [r = .23]

 

[Mavric]

But if you're looking at success in the same year as recruiting, isn't that saying the opposite (or, perhaps not opposite, but at least distinctly different)? The original premise was that recruiting leads to good teams. Those results seem to say that good teams leads to good recruiting. To show a better correlation as to recruiting leading to on-field success, wouldn't you have to compare recruiting to wins 2-4 years down the road?

And all that is still presupposing that the recruiting rankings are pristine and unbiased. Five stars are the best, two stars at the bottom but the vast majority of BCS signees (I'd speculate high-90%) are three and four star recruits which is very hard to tell the difference between a lot of them. There are myriad reasons why a mid-four-star in Texas could be a mid-three-star in Wisconsin.

 

[It'sNotAFakeID]

But if you're looking at success in the same year as recruiting, isn't that saying the opposite (or, perhaps not opposite, but at least distinctly different)? The original premise was that recruiting leads to good teams. Those results seem to say that good teams leads to good recruiting. To show a better correlation as to recruiting leading to on-field success, wouldn't you have to compare recruiting to wins 2-4 years down the road?

And all that is still presupposing that the recruiting rankings are pristine and unbiased. Five stars are the best, two stars at the bottom but the vast majority of BCS signees (I'd speculate high-90%) are three and four star recruits which is very hard to tell the difference between a lot of them. There are myriad reasons why a mid-four-star in Texas could be a mid-three-star in Wisconsin.

The pattern of causation will never be clear. Did the chicken or the egg come first kind of thing. But, here are the correlations for recruiting ranks of past seasons and wins in 2012:

2008: r = .388

2009: r = .365

2010: r = .349

2011: r = .423

2012: r = .440