# Returning Offense of 2012 Opponents-Part 2

Submitted by hart20 on July 26th, 2012 at 2:05 AM

As promised, albeit a little late, here are the pretty, pretty charts, along with some analysis.

For reference, here’s the first part of the diary: http://mgoblog.com/diaries/returning-offense-2012-opponents-part-1

I make extensive use of Z-scores in this diary, which are simply standardized scores, representative of the number of standard deviations above or below the mean the raw scores are. These scores are helpful in this scenario because they allow us to quantify the difference between teams in each category of returning offense.

Here’s a good resource for more on Z-scores: http://statistics-help-for-students.com/What_are_Z_scores.htm

Now for the fun stuff:

I. Returning Contributors

 Returning Contributors Z Score Ohio St. 17 1.777897506 Air Force 15 0.711159002 Purdue 15 0.711159002 Illinois 15 0.711159002 Northwestern 15 0.711159002 Minnesota 14 0.177789751 Nebraska 13 -0.355579501 Iowa 13 -0.355579501 Michigan 13 -0.355579501 Alabama 12 -0.888948753 Notre Dame 12 -0.888948753 MSU 10 -1.955687256

 Returning Contributors Mean 13.66667 Median 13.5 Mode 15 Standard Deviation 1.874874 Range 7 Minimum 10 Maximum 17

Top 3 Teams with Most Returning Contributors

1. Ohio St.

2a. Air Force

2b. Purdue

2c. Illinois

2d. Northwestern

3 Teams with the Least Amount of Returning Contributors

1. MSU

2a. Notre Dame

2b. Alabama

Teams on the schedule return an average of 14 contributors on offense, σ=1.87. The minimum number of returning contributors is 10, while the maximum number of returning contributors is 17.

Ohio St. returns the most contributors on offense, returning 17 contributors, z=1.78. Air Force, Purdue, Illinois, and Northwestern all return 15 contributors, z=0.71.

MSU returns the least number of contributors, returning only 10, z=-1.96. Notre Dame and Alabama both return 12 contributors on offense, z=-0.89.

Michigan returns 13 contributors on offense, z=-0.36.

II. Rushing Offense

The formula that I used to calculate the total returning percent for each team’s rushing offense was to give a 33.3% weight to each of returning carries, returning rushing yards, and returning rushing TDs.

 Returning Weighted Percent Z Score Nebraska 93.83372971 1.337053 Michigan 92.79665306 1.291064 Purdue 88.97601017 1.121639 Ohio St. 75.16063082 0.508999 Minnesota 70.10965805 0.285015 Notre Dame 61.59637443 -0.0925 MSU 60.10200379 -0.15877 Northwestern 59.73306701 -0.17513 Illinois 56.30634269 -0.32709 Air Force 47.60320235 -0.71303 Alabama 40.62151205 -1.02263 Iowa 17.34953531 -2.05462

 Weighted Percent Mean 63.68239 Median 60.84919 Standard Deviation 22.55059 Range 76.48419 Minimum 17.34954 Maximum 93.83373

Top 3 Returning Rushing Offenses

2. Michigan

3. Purdue

Worst 3 Returning Rushing Offenses

1. Iowa

2. Alabama

3. Air Force

Teams on the schedule return an average of 63.68% of their rushing offense from their last year, σ=22.55. The minimum returning rushing offense returns 17.34%, while the maximum returning rushing offense returns 93.83%.

Nebraska returns 93.83% of their rushing offense from last year, z=1.34, and Michigan closely follows by returning 92.80% of the rushing offense from last year, z=1.29.  Purdue follows closely, returning 88.98% of their rushing offense, z=1.12.

The effect of the AIRBHG can’t be understated, as well, as Iowa returns just a meager 17.35%, z=-2.05, of their rushing offense from last year. Also of note is the loss of Trent Richardson for Alabama. Just 40.62%, z=-1.02, of their rushing offense returns from last year. Air Force follows behind Alabama, returning 47.60%, z=-0.71, of their rushing offense.

III. Returning Receiving Production

The formula that I used to calculate the total returning percent for each team’s receiving production was to give a 33.3% weight to each of returning receptions, returning receiving yards, and returning receiving TDs.

 Returning Weighted Percent Z Score Ohio St. 86.10358522 1.723013 Nebraska 77.64401481 1.321687 Purdue 75.40875274 1.215645 Notre Dame 61.71283973 0.565904 Michigan 49.66700893 -0.00556 Iowa 45.0441055 -0.22487 Illinois 43.30445644 -0.3074 Northwestern 40.21911191 -0.45377 Minnesota 35.92596899 -0.65744 Alabama 35.02397172 -0.70023 Air Force 29.3543474 -0.9692 MSU 18.00131419 -1.50779

 Weighted Percent Mean 49.78412 Median 44.17428 Standard Deviation 21.07904 Range 68.10227 Minimum 18.00131 Maximum 86.10359

Top 3 Teams With Returning Receiving Production

1. Ohio St.

3. Purdue

The 3 Teams with the Worst Returning Receiving Production

1. MSU

2. Air Force

3. Alabama

Teams on the schedule return an average of 49.78% of their receiving production from last year, σ=21.08. The minimum returning receiving production returns 18.00%, while the maximum returning receiving production returns 86.10%.

Ohio St. is the clear cut leader in returning receiving production on the schedule, returning 86.10% of their receiving production from last year, z=1.72.  Nebraska follows by returning 77.64%, of their receiving production,  z=1.32. And in 3rd place, Purdue closely follows Nebraska, returning 75.41% of their receiving production, z=1.22.

MSU returns the least amount of receiving production, returning just 18.00%, z=-1.51. Air Force returns 29.35% of their receiving production, good for second-to-last place, z=-0.97. Alabama returns 35.02% of their receiving production, z=-0.70.

IV. Returning Passing Production

The formula that I used to calculate the total returning percent for each team’s passing production was to give a 25% weight to each of returning passing attempts, returning completions, returning passing yards and returning passing TDs.

 Returning Weighted Percent Z Score Illinois 100.00% 0.650396 Nebraska 100.00% 0.650396 Minnesota 100.00% 0.650396 Michigan 100.00% 0.650396 Purdue 99.83% 0.646041 Iowa 99.02% 0.625394 Notre Dame 95.84% 0.544583 Alabama 94.61% 0.513397 Ohio St. 68.83% -0.14238 Northwestern 28.34% -1.1722 MSU 5.44% -1.7546 Air Force 1.23% -1.86182

 Weighted Percent Mean 74.43% Median 97.43% Mode 100.00% Standard Deviation 39.32% Range 98.77% Minimum 1.23% Maximum 100.00%

Top 3 Returning Passing Offenses

1a. Illinois

1c. Minnesota

1d. Michigan

Worst 3 Returning Passing Offenses

1. Air Force

2. MSU

3. Northwestern

Teams on the schedule return an average of 74.43% of their passing production from last year, σ=39.32, although the median returning passing production is 97.43%. The minimum returning passing production is 1.23%, while the maximum returning passing production is 100%.

Illinois, Nebraska, Minnesota, and Michigan return 100% of their passing production from last year, z=0.65.

Air Force returns a meager 1.23% of their passing production from last year, z=-1.86, while MSU is close behind, returning only 5.44% of their passing production from last year, z=-1.75. Northwestern follows behind both Air Force and MSU, returning 28.34% of their passing production from last year, z=-1.17.

V. Conclusion: Total Returning Offense

The formula that I used to calculate the total percent of returning offense was to give a 33.3% weight to each of returning rushing offense, returning receiving production, and returning passing production.

 Returning Weighted Percent Z Score Nebraska 90.48351741 1.285521374 Purdue 88.06234289 1.173796166 Michigan 80.81313455 0.839281119 Ohio St. 76.69039019 0.649036894 Notre Dame 73.04231837 0.480696455 Minnesota 68.67166881 0.279012633 Illinois 66.53014632 0.180191981 Alabama 56.74612482 -0.271292134 Iowa 53.79815298 -0.407326428 Northwestern 42.75978026 -0.916692637 MSU 27.84560183 -1.604908076 Air Force 26.0597271 -1.687317346

 Weighted Percent Mean 62.62524 Median 67.60091 Standard Deviation 21.6708 Range 64.42379 Minimum 26.05973 Maximum 90.48352

Top 3 Returning Offenses

2. Purdue

3. Michigan

Worst 3 Returning Offenses

1. Air Force

2. MSU

3.  Northwestern

Teams on the schedule return an average of 62.63% of their offense from last year, σ=21.67. The minimum returning offense is 26.06%, while the maximum returning offense is 90.48%.

Nebraska returns the most offense from last year, returning 90.48% of their offense, z=1.29. Purdue closely follows Nebraska, returning 88.06% of their offense from last year, z=1.17. Michigan comes in 3rd place, returning 80.81% of the offense from last year, z=0.84.

Air Force returns the least amount of offense from last year, returning just 26.06% of their offense, z=-1.69. MSU is right behind, returning only 27.85% of their offense from last year, z=-1.60. Northwestern returns 42.76% of their offense from last year, z=-0.92.

Any comments, questions, or constructive criticisms are appreciated. I hope to have the defensive version out in the next couple of weeks.

All stats are courtesy of http://www.cfbstats.com/

Font confusion is confusing.

What's the correlation between returning offensive contributors and wins?  I realize this a dumb question but it seems like the driver for your analysis, cause we care about wins.

Time to retroactively analyze past years' data, but I probably won't find the time until the season starts. I do plan on examining the relationship though. Honestly, I just included the data on returning contributors because it was available as a byproduct of putting together the rest of the data and I found it interesting. I believe the correlation between returning starters and increase in wins starts at 17 returning starters though. My primary motivation behind this all was more of a focus on returning offense than the number of returning contributors, although I suspect that they're closely correlated. Putting this together actually raised a lot of questions that I'd like to examine in the future.

Thank you for your efforts as there's quite a bit of data here.  It would also be interesting to see that data retrospectively.  Then you could see if it was a worthwhile predictor and maybe look at correlations for championship seasons, bad seasons, et al.

With all of the relationship and correlation math I've read on this blog, you would think that some software guy would combine them in a program and have it spit out results every week.  It could be a thing, like a feature thing.

Hey Brian, if you're reading this,

Y U NO HAVE ALREADY ON BLOG??

is in the "three worst" grouping in every category except returning rushing O.

I think Boise State will probably put 8 in the box and force a rookie qb with inexperienced recievers to beat them.

He's better recruit than cousins and is a RS Soph. Between that and the WR transfer from Tennessee plus a solid Running Game....MSU offense will be fine. I bet we're hard pressed to see an objective difference between Maxwell and Cousins. State's offense may even be better than they were last year (not saying much).

Lots of work on that original post.

Mark Twain: There are three kinds of lies...lies, damn lies, and statistics.

Maxwell is a RS JR, and his staff and players have a ton of confidence in his ability and decision making.  The WR's are young, but he's been practicing with guys like Lippett, Mumphrey, and Fowler for a couple of years, and Arnett should be good to go in no time.

Baker was a really good RB, but he left to the NFL, in part because he lost the starting job to Bell.  More carries for Bell is not a bad thing.  And not sure if Caper was considered in the count of returning contributors, but he should be.  And both are really good for short passes.   That's a strong 1,2.  Probably the best in the big ten east of Wisconsin.  Maybe better.

The o-line has 4 returning starters, and able back ups.  3 returners started for the first time last year.  Gotta imagine they've improved.  Actually, they were pretty good by the end of last year.

Sims is the biggest x-factor, IMO. 280 and still an athlete.  Could cause all sorts of mis-match situations.

MSU's O could be better that last year, and MSU's O only needs to be average. That D is Saban era good.

And what about Conroy?  He's one of the top kickers in the league.

A poster mentioned BSU with an 8-man box.  They play a 3-3-5.  Remember that D?  Bell is 244, and a bear for even a LB.  A DB will get pummelled.

Lots of work on that original post.

Mark Twain: There are three kinds of lies...lies, damn lies, and statistics.

Maxwell is a RS JR, and his staff and players have a ton of confidence in his ability and decision making.  The WR's are young, but he's been practicing with guys like Lippett, Mumphrey, and Fowler for a couple of years, and Arnett should be good to go in no time.

Baker was a really good RB, but he left to the NFL, in part because he lost the starting job to Bell.  More carries for Bell is not a bad thing.  And not sure if Caper was considered in the count of returning contributors, but he should be.  And both are really good for short passes.   That's a strong 1,2.  Probably the best in the big ten east of Wisconsin.  Maybe better.

The o-line has 4 returning starters, and able back ups.  3 returners started for the first time last year.  Gotta imagine they've improved.  Actually, they were pretty good by the end of last year.

Sims is the biggest x-factor, IMO. 280 and still an athlete.  Could cause all sorts of mis-match situations.

MSU's O could be better that last year, and MSU's O only needs to be average. That D is Saban era good.

And what about Conroy?  He's one of the top kickers in the league.

A poster mentioned BSU with an 8-man box.  They play a 3-3-5.  Remember that D?  Bell is 244, and a bear for even a LB.  A DB will get pummelled.

Good stuff, thanks!

You went all out on the sig figs, eh?

haha. I didn't even realize that I forgot to limit the sig figs. I did everything in Excel, so I just copied and pasted the tables. Something to fix for the defensive diary though.

ohio lost most of Devier Posey's production last year, except against Michigan. So, for our interests, they will lose more than it looks like. On the other hand, we could be losing more of our rushing production, depending on how things fall out.

without Fitz?

Here's to hoping he's out for only 1-2 games, but a blood-alcohol of 0.12?  Yesh...

Isn't a dismissable offense at many universities. Hell, Michael Floyd had 2 at ND and he didn't miss a game. Frank Clark, on the other hand, seems to have been dismissed so I'll discount him. Once, we find out how long Fitz is suspended for, I can put out a small update comparing our returning production to our opponents'.

Ohio losing the Walrus has to count for something doesn't it?

Nicely done, but I think you'd need to include something about the O-line to get a more accurate representation.

Good call-out re: OL.  Maybe something along the lines of number of starts per position at OL.

Also, a pet peeve of mine is the whole "least amount" thing.  Since you're counting things here (like a discrete number of teams), you should compare Most Returning Contributors with Fewest Returning Contributors, not Least Amount of Returning Contributors.  I actually had an argument with a high school girlfriend over this (yeah, that was like 20 years ago, and I admit to being a bit of a nerd).

I knew that it didn't sound right, but I couldn't think of what did. I'll change it when I get to a computer, which unfortunately won't be for a few days. As to the point about taking into account returning O linemen that you and the poster above raised, it's a good point and I'll try to go back before the season starts and account for that in a new formula.

you double count passing game production, since you give 33% weight to both passing and receiving.  In essence, a one yard completion from Denard to Roundtree gets credited as two yards.  I realize that the data's all there and I could do it myself, but it seems like a better formula would be Total Returning Offense = 0.5*Rushing Offense + 0.25*Passing Offense + 0.25*Receiving Offense.  Does that change the results significantly, or even at all?

Good diary--thanks.

That gives equal weight to rushing offense and the passing offense, splitting passing offense into equal parts returning passing production and receiving production, and that also takes into account returning O linemen. Unfortunately, I won't be able to access my computer for a few days and I'm working on a few other diaries that I'd like to post before the season starts. I'll definitely try to put out an update before the season starts though.

Old Cork was the best QB in the history of the MSU program, its kind of a reach to assume that Maxwell will seemlessly take over. I certainly expect some growing pains, Cousins pulled out 2-3 games the last two years where MSU was all but dead. How many days left until Dallas....too many!