Miller: Recruiting Rankings vs Big Ten Wins

JonDMiller

Publisher/Founder
With the 2012 edition of National Signing Day now in our rear view mirror, fans from all across the country are discussing their newest players. Other debates are also beginning to pick up, depending on how fans perceive their classes and recent fortunes on the gridiron.

I was reading a few comments from one of the items I wrote on HawkeyeNation.com yesterday, some fans voicing their displeasure with Iowa's recruiting class and recent on field efforts. I will come back around to that at the bottom of this piece, but I want to first take a look at Big Ten recruiting in general and juxtapose historical recruiting class rankings against actual on field performance.

THE BIG TEN

I ran two sets of numbers for this 'study'. One raw set of data took Scout.com's Big Ten recruiting class rankings dating back to 2002. I assigned a numerical value to each school for each year. For instance, in 2002 Ohio State had the top rated class among Big Ten schools and Indiana's class ranked 11th among 11 Big Ten schools. Ohio State therefore received a numerical value of 1 and Indiana received an 11, with the other schools falling where they ranked. I did this for each recruiting class 2002-2012.

I then juxtaposed those numbers with the actual on field conference winning percentage numbers for the same time period. These are conference games only and I did not add in this year's Big Ten title game victory for Wisconsin, as I used regular season results only.

One flaw with this 11 year aggregate is the fact that 2002 on field results were influenced by recruiting classes from 1999-2001. However, the numbers are still interesting to look at.

The second set of data took the same recruiting class ranking analysis but from the 2007-2010 and then juxtaposed those results against on field conference wins for 2008-2011. Most Big Ten football players redshirt their freshman year, so the class of 2007 would have been true freshmen in 2007. This is why I began the on field comparison with 2008 season through 2011, because several 2007 recruiting class alums would have been fifth year seniors in 2011.

This second data set is a more realistic reflection of how the programs have been evaluating and developing talent these paste five years and illustrates the correlation between recruiting class rankings and on field production, or lack there of.

I will display the graphic now, and you can also access a larger version of the data with more detailed data field descriptions at this shared google doc I have created by clicking on this link.

B1GRankUse.jpg



In the raw 11 year raw data set, you will not be surprised to see that Ohio State, Michigan and Penn State are 1-2-3 in the combined recruiting class rankings. Iowa sits at #4 over that time span, which may be a bit of a surprise to some of you. Wisconsin sits at #7, a little lower than I would have pegged.

Wisconsin's on field conference ranking is well ahead of their recruiting class combined ranking; they are 2nd in Big Ten wins during this time frame while their recruiting class ranking was 7th. Iowa and Northwestern are two positions higher than you would have predicted based upon recruiting class rankings with Purdue one position better. Ohio State and Michigan are right on target, first and second in both wins and recruiting class ranking. Unfortunately for Indiana and Minnesota, their win totals are also on target with their recruiting performance.

Some may also argue that if a team is Plus One or Minus One, you could say they are on target for meeting recruiting class ranking expectations on the field.

Penn State finished Minus two in wins compared to where their talent might suggest they should have finished, while Illinois was the worst in the league at Minus Four.

The second data set is the more refined set that I spoke of earlier. It saw Ohio State and Indiana on target from where the recruiting class ranking data would have predicted they would be on the field. Michigan State (+5) and Northwestern (+4) lead the way in the overachiever department. Wisconsin was just +1 even after receiving a nice little bump this past season due to the fifth year transfer of quarterback Russell Wilson. The Badgers were probably an eight-win team at best in 2011 before Wilson fell into their lap, mostly due to their favorable schedule.

Illinois is once again on the wrong side of the on field performance at Minus Five. It's not surprising they have made a coaching change just this past season. The same can be said for Michigan's disappointing Minus Four departure and they fired Rich Rodriguez after the 2010 season.

The biggest surprise in this four year data set was that Michigan State and Ohio State have the same number of conference wins over the course of the last four seasons.

IOWA-CENTRIC

It's going to be interesting to see the opinions people form from reviewing this data. Here is my take away; Iowa is doing just fine.

Sure, the 2010 7-5 regular season is going to be a wound that likely won't heal for some time as it was a big time missed opportunity. I still believe that much of the fan base angst in 2011 was a hangover from 2010, as the 2011 Hawkeyes were the least experienced team in the Big Ten to start the season.

I hate to say this, but expect more of the same in 2012. I will write more on that in a few minutes.

When you consider Iowa's low population and the low number of BCS caliber football players the state produces each year, I think Iowa has probably outperformed expectations in the aggregate recruiting class rankings in both the 12 and four year snapshots. Their conference on field winning numbers are higher than the average ranking of their recruiting classes. When Iowa goes 6-2 in league play, like they did in 2009, they are outperforming their mean expectations in a big way. The 4-4 record in 2010 was on par with a 7th ranking in the conference recruiting standings, but below the recruiting average expectation as well as the eyeball expectation.

So on the whole, or the mean, the Iowa football program is performing to what you would expect in Big Ten conference play on an 11 year look and slightly above what the recruiting rankings suggest on the four year look.

WHAT DOES THE FUTURE HOLD?

In looking at the first data set from above, you see Iowa's 2008-2009 recruiting classes ranked 7th and 11th in the Big Ten. Only Northwestern and Indiana had a lower two year aggregate ranking during that span.

To make matters worse, 19 of the players Iowa signed in those two classes are no long in the program. So not only were these two classes lightly regarded by the recruiting services, the 2008-2009 classes will comprise most of the leadership in Big Ten teams this year in the form of fourth and fifth years players. In other words, Iowa gets the double whammy.

I think Iowa is in for a 2012 season that will be similar to what we saw in 2011; inconsistent due to a high number of inexperienced players being on the field. This will be most glaring at defensive line, running back and wide receiver.

However, I do think there is hope for better on field production down the line.

Take a look at the first data set one more time, and focus on the recruiting class rankings for 2010 through 2012.

As it compares to their conference foes, Iowa's last three classes will average out to fourth, without factoring in Nebraska. This is three slots better than their 2007-2011 four year average.

That being said, when you look at the 12 year data set, you will not find any better three-year consecutive recruiting stretch than what Iowa has just accomplished. Again, I do not include Nebraska in this numbers because they have not been in the Big Ten for the entire time frame.

I think even pessimistic Iowa fans would admit the Hawkeye coaching staff has an excellent track record of developing players. Given the amount of talent they have pulled in these past three years, as long as they can do a better job of player retention than we have seen in the past two years, the Hawkeye football program has a bright future.

Perhaps there is something to be said of the 'Ferentz Oscilation', something a HawkeyeNation.com message board user brought up a while back.

Here is the Ferentz Era in three-year windows, from a total wins perspective:

1999-2001: 11
2002-2004: 31
2005-2007: 19
2008-2010: 28
2011-2013?

With seven wins in 2011 and I think around that number in 2012, the 'dip' in this three year set won't be as 'deep' as the two that came before it. The first 'dip' was when Ferentz was taking over a program that had sort of run off the rails.

The quarterback question will loom for the 2013 campaign, but there are a lot of young playmakers on this team, players with a great deal of speed on defense and some highly touted defensive line prospects who will be on the verge of coming into their own by that time.

At least the data thinks that will be the case.
 
Last edited:
OSU and Michigan both had outstanding classes that are head and shoulders better than the rest of the league and are off to great starts with their '13 classes. The talent gap may widen to the point where its the Big 2 and the rest of the B10 again. The rest of the B10 had a pretty rough recruiting year. Iowa had a good class, Nebraska and MSU did ok and Penn State is heading straight toward mediocrity.
 
One very simple measure you didn't include in your article - the correlation coefficient between recruiting ranking and wins in your data set is 0.8. That's pretty strong.

For those unfamiliar with the concept, correlation coefficient is a range between -1 to 1. A coefficient close to zero means the relationship is random. Amounts going from 0 to -1 or 1 mean that the relationship is 100% inversely correlated (for negative numbers) or 100% correlated for positive numbers.

0.8 is a moderately strong indicator. Correlation never proves causation, but it indicates that there is at a bare minimum a strong indicator that it's predictive.
 
One very simple measure you didn't include in your article - the correlation coefficient between recruiting ranking and wins in your data set is 0.8. That's pretty strong.

For those unfamiliar with the concept, correlation coefficient is a range between -1 to 1. A coefficient close to zero means the relationship is random. Amounts going from 0 to -1 or 1 mean that the relationship is 100% inversely correlated (for negative numbers) or 100% correlated for positive numbers.

0.8 is a moderately strong indicator. Correlation never proves causation, but it indicates that there is at a bare minimum a strong indicator that it's predictive.

So...does that mean I done good?
 
One very simple measure you didn't include in your article - the correlation coefficient between recruiting ranking and wins in your data set is 0.8. That's pretty strong.

For those unfamiliar with the concept, correlation coefficient is a range between -1 to 1. A coefficient close to zero means the relationship is random. Amounts going from 0 to -1 or 1 mean that the relationship is 100% inversely correlated (for negative numbers) or 100% correlated for positive numbers.

0.8 is a moderately strong indicator. Correlation never proves causation, but it indicates that there is at a bare minimum a strong indicator that it's predictive.

Plz advz re r-squared. thx.
 
OSU and Michigan both had outstanding classes that are head and shoulders better than the rest of the league and are off to great starts with their '13 classes. The talent gap may widen to the point where its the Big 2 and the rest of the B10 again. The rest of the B10 had a pretty rough recruiting year. Iowa had a good class, Nebraska and MSU did ok and Penn State is heading straight toward mediocrity.

can't believe you'd share your big 1 little 10 title with another team, especially the hated. I wouldn't be so eager to share your ill-gotten spoils quite yet.
 
You don't need statistics to analyze how the B10 stacks up. It's pretty simple to see.
 
r^2 is the coefficient of determination. Assuming the correlation coefficient of r=0.8 is correct, this means that r^2 is 0.64. This means that approximately 64% of the variation in win totals can be attributed to an approximate linear relationship between recruiting rankings and win totals. :D
 
You don't need statistics to analyze how the B10 stacks up. It's pretty simple to see.

This wasn't about arriving at a 'see, Ohio State should be the best' conclusion...I didn't know where it would go and was surprised to see how closely Iowa's recruiting efforts matched their on field performance, and how solid that performance has been through the years.
 
Nice piece but need some help.........why do the recruiting ranking numbers on the bottom chart for each individual team and year not match up (in all cases) with the entries on the top chart?

e.g. ......Iowa is ranked 4, 3, 4, 8 on the bottom but 5, 7, 11, 6 on the top for the 2007-2011 period
 
can't believe you'd share your big 1 little 10 title with another team, especially the hated. I wouldn't be so eager to share your ill-gotten spoils quite yet.

Hoke and his style fit the B10 well and he is recruiting well on defense unlike RRod. Michigan will be a threat for some time. I mean they wont be OSU good but good none the less.
 
One very simple measure you didn't include in your article - the correlation coefficient between recruiting ranking and wins in your data set is 0.8. That's pretty strong.

For those unfamiliar with the concept, correlation coefficient is a range between -1 to 1. A coefficient close to zero means the relationship is random. Amounts going from 0 to -1 or 1 mean that the relationship is 100% inversely correlated (for negative numbers) or 100% correlated for positive numbers.

0.8 is a moderately strong indicator. Correlation never proves causation, but it indicates that there is at a bare minimum a strong indicator that it's predictive.

I wish I had the ability to break it down like you just did. Nice job.
 
Data seems to invalidate the "Little ol' Iowa" argument. On average, Hawks are the 4th most talented team in the B1G.
This means 1 of 2 things:
1) The coaches are very good recruiters, able to overcome the "challenges" of recruiting to Iowa, or,
2) Those challenges are not nearly as great as many want to make them out to be.

It also seems to establish a reasonable, expectation of where Iowa should finish each year; I calculate the 4th best team in the conference = 8-4 / 5-3.
(Use the inverse ranking of the 12 teams in your data pool, best = 12, 2nd = 11, etc means Iowa is 9th of 12, or a value of .75. .75 x 12 games = 9 wins; .75 x 8 conference games = 6 wins. I lower this to allow for immeasurable influences, such as injury / attrition, home field, weather, etc.)

Still, year-in, year-out, Iowa is talented enough to win 8 and 5, which is exactly what I've maintained as my expectation for a few years, now.

One other inference I take from the data, there seems to be a direct correlation between coaching and scheme and over / under achievement. Even Jon acknowledged this link in his comments regarding Illinois / Zook and Michigan / Rich-Rod -- 2 glaring examples of poor coaching and / or incompatible scheme / philosophy. The positive examples being Fitzgerald / NwU and Purdue -- probably the most unique schemes to prepare for in the B1G.

I think there is some validation to those who argue there is some scheme / strategy disparities, not to mention some coaching issues (from Norm's health to lack of special teams preparedness to general stagnation) that explain some of Iowa's recent underachievement.
 
This wasn't about arriving at a 'see, Ohio State should be the best' conclusion...I didn't know where it would go and was surprised to see how closely Iowa's recruiting efforts matched their on field performance, and how solid that performance has been through the years.

I was more responding to Einstein and his theory of relativity response than anything.
 
Hoke and his style fit the B10 well and he is recruiting well on defense unlike RRod. Michigan will be a threat for some time. I mean they wont be OSU good but good none the less.

there you go my friend. i thought for a minute you slid into the abyss.

you might want to coin a phrase like big 1, mediocre 1 and the little 10, or something like that. i'm sure you'd be able to come up with something catchier than i....i'm a checkers player....
 
Data seems to invalidate the "Little ol' Iowa" argument. On average, Hawks are the 4th most talented team in the B1G.
This means 1 of 2 things:
1) The coaches are very good recruiters, able to overcome the "challenges" of recruiting to Iowa, or,
2) Those challenges are not nearly as great as many want to make them out to be.

By using straight rankings 1-2-3-4, Jon's approach obscures the significant gap most years between the Big 2 and the rest in overall recruiting rankings.

2012: UM 4, OSU 3, IA 39
2011: UM 29, OSU 6, IA 25
2010: UM 12, OSU 20, IA 45
2009: UM 14, OSU 1, IA 75
2008: UM 6, OSU 4, IA 44
2007: UM 10, OSU 16, IA 37
2006: UM 9, OSU 13, IA 40
2005: UM 2, OSU 7, IA 8
2004: UM 5, OSU 11, IA 41

PSU was also usually much higher, with several Top 10 rankings.

Using weighted recruiting scores you'll find Iowa and Northwestern overachieved even more than Jon states, and PSU significantly underachieved.
 

Latest posts

Top