Has Michigan Faced a Good Offense? Also including one man's take on B1G SOS.

Submitted by justingoblue on

Has Michigan faced a good offense? ND has big time turnover issues, WMU and EMU are both MAC teams (although WMU seems like a very good MAC team), SDSU didn't manage very much against us and Minnesota is possibly the worst BCS conference team this year.

I took a look at each opponent on our schedule and looked at their opponents total and scoring defenses, then took a look in each box score for total yards gained by our opponent as well as final score. I added total offense and scoring offense into a year to date chart and divided actual yardage/score by their expected performance based on opponents NCAA statistics. Here is an example of part one:

 

WMU TD Yardage Yards SD PPG Points Record
Michigan 325.5 462 10.25 10 5W 0L
Nicholls State NA NA NA NA NA
CMU 398.2 442 32.6 44 1W 3L
Illinois 295.6 341 17.4 20 4W 0L
Uconn 315.6 490 18.4 27 1W 3L
Total 1334.9 1735 78.65 101 0.647
% of Normal 130   128  

FCS teams are excluded, and I'll get to the opponents win percentage later in the diary. You can see that WMU is outperforming their total offense expectation by about 130%, and their offense is scoring 128% above par. I did this for each of our twelve opponents, and have ranked each offense by performance percentage in both scoring and total, then I try to put a finger on exactly how well each has performed to date by including an average of the two with opponents record factored in. As I write this, I have not made those calculations and do not know how accurate the list will come out. Nevertheless, I will include it for discussion's sake.

Total offense vs. Expectation:

  1. ND 136%
  2. WMU 130%
  3. Nebraska 115%
  4. MSU 112%
  5. SDSU 110%
  6. Illinois 108%
  7. Iowa 107%
  8. Northwestern 99%
  9. OSU 93%
  10. EMU 88%
  11. Minnesota 87%
  12. Purdue 79%

Scoring offense vs. Expectation:

  1. ND 157%
  2. Nebraska 150%
  3. Iowa 142%
  4. WMU 128%
  5. SDSU 115%
  6. Northwestern 106%
  7. MSU 105%
  8. OSU 95%
  9. Illinois 92%
  10. EMU 79%
  11. Minnesota 72%
  12. Purdue 62%

Average offense vs. Expectation (list 1* list 2/2):

  1. ND 146.5%
  2. Nebraska 132.5%
  3. WMU 129%
  4. Iowa 124.5%
  5. SDSU 112.5%
  6. MSU 108.5%
  7. Northwestern 102.5%
  8. Illinois 100%
  9. OSU 94%
  10. EMU 83.5%
  11. Minnesota 79.5%
  12. Purdue 70.5%

Taking out our opponent's opponent's FCS games, here is my take on SOS for Michigan's opponents 2011, based on opponent win percentage:

  1. SDSU .692 (Army, Washington State, Michigan)
  2. ND .684 (USF, Michigan, MSU, Pitt, Purdue)
  3. WMU .647 (Michigan, CMU, Illinois, UConn)
  4. Nebraska .643 (Fresno State, Washington, Wyoming, Wisconsin)
  5. EMU .615 (Michigan, PSU, Akron)
  6. Minnesota .579 (USC, NMSU, Miami (NTM), Michigan)
  7. Iowa .571 (Iowa State, Pitt)
  8. Illinois .533 (Arkansas State, ASU, WMU, Northwestern)
  9. Northwestern .462 (BC, Army, Illinois)
  10. Purdue .429 (MTSU, Rice, ND)
  11. MSU .389 (FAU, ND, CMU, OSU)
  12. OSU .3 (Akron, Toledo, Miami, Colorado, MSU)

Putting both of these lists together to try and get a clear picture as to who shows up against better opposition, here we have:

  1. ND
  2. Nebraska
  3. WMU
  4. SDSU
  5. Iowa
  6. Illinois
  7. EMU
  8. Northwestern
  9. Minnesota
  10. MSU
  11. Purdue
  12. OSU

This last list very well might be useless since it's double counting data (a good defense allows less yards, a good defense wins more games as a result), but I do think it best reflects my observations so far, all things considered.

Caveats:

  1. I'm neither a statistician nor a mathematician, I fully expect this to have some issues, and I'll do my best to correct them within the framework of this diary.
  2. Each team is counted towards the defensive statistics. I realize that with a sample size as small as three that can get problematic, but I didn't feel like I had too much of a choice due to data and time constraints.

Comments

justingoblue

October 4th, 2011 at 3:36 PM ^

If you were to follow this through the season (which maybe I will if it gets a halfway decent reaction) ND would begin to drop as M and MSU play each other, Nebraska and MSU plays Wisconsin. They're being bolstered big time by showing up in those two games (and the huge yards against USF).

I do think ND's offense is good, however. No denying our offense was at least "good" last year and those issues might end up being bigger than yours.

ND TD Yardage Yards SD PPG Points Record
USF 355.2 508 22.4 20 3W 1L
Michigan 325.5 513 10.25 31 5W 0L
MSU 173.4 275 10.2 31 2W 1L
Pitt 414.4 398 21.6 15 2W 2L
Purdue 379 551 21.5 38 1W 2L
Total 1648.5 2245 85.95 135 0.684
% of Normal 136   157  

 

Irish

October 4th, 2011 at 3:48 PM ^

Eh I can't really go with that.  ND's offensive grade at the end of the season isn't what UM played agaisnt, its the offensive grade they received at the end of week 2.  I expect a week 6 MSU to grade out higher than a week 2 ND.

At the end of week 2 this is what ND's offense looked like:

Passing O: 10th

Rush O: 57th

Total O: 13h

Scoring O: 75th

Red Zone O: 117th

If you can't put points on the board you do not have a good offense

justingoblue

October 4th, 2011 at 3:51 PM ^

Fair point, but this doesn't only relate to Michigan. If ND were to go and rip off 50 point games against LSU, Alabama and Wisconsin, shouldn't they move up as a season to date offense? If this was written the day after UTL, ND would be much lower in scoring offense. Their strong showing against MSU helped, as did Michigan showing to be a good defense against EMU, SDSU and Minnesota.

Irish

October 4th, 2011 at 4:05 PM ^

No not at all imo.  Teams get weaker and stronger all over the field as seasons go on.  Injuries, adding experience are all defining traits of a football team at whatever point in the season.

Its fair to average the teams performance up to that week of the game.  Last week ND played a 2 QB purdue team.  Offensively they could look much better as the season goes on if they have 1 QB emerge.  If UM can find someone other than Denard to carry the ball consistantly, UM's offense will become stronger.  You may have the same names on the field but its a new team every week that is put on the field.

justingoblue

October 4th, 2011 at 4:26 PM ^

While I agree that teams change, they do so within the context of a season. We'll never have any idea how good ND's offense or M's defense was UTL without looking at what happened later in the year. If ND finished the year 0-12, or if M lost and did the same, that would be highly relevent to a discussion on how good either team was in 2011, IMO.

Moleskyn

October 4th, 2011 at 3:43 PM ^

A few questions:

  • Does the TD Yardage column refer to the number of yards accumulated on scoring plays? If so, and I mean this honestly, is that relevant a stat?
  • What does the SD PPG column stand for?
  • What did you use as your "Normal" for comparison?

Overall it's interesting research, thanks for taking the time to put this together.

Edit: Also, to go off of what Irish said, do you think turnovers should be included in this?

justingoblue

October 4th, 2011 at 3:48 PM ^

TD= Total defense. Yards allowed per game.

SD PPG= Scoring defense points per game (as opposed to rank or something).

Normal is the average on the season.

So Michigan would look like this:

TD Yardage: 325.5

SD PPG: 10.25

Normal would be either of those numbers, depending on whether it's a scoring or yardage comparison.

Edit to respond to your edit: Maybe, but what you end up with is an integer not really relating to anything. I have no idea how you would factor in turnovers without going to a drive by drive analysis like FEI or Mathlete. I suppose you could subtract some kind of EV points, but coming up with something standard is tough. ND's last turnover in UTL really didn't affect the outcome of the game (the kickoff fumble), while Denard's INT's against MSU killed us. Both look the same when looking at NCAA turnover stats (which is what I use for scoring/yardage).

Moleskyn

October 4th, 2011 at 3:58 PM ^

Ahh, I see now. Makes sense.

Edit to your edit to my edit: I can see how it would be difficult to logically factor that in, and I'm not even sure how you would incorporate turnovers, but when evaluating the overall level of an offense, aren't you missing a piece of the puzzle if you leave them out? Maybe include each team's number of turnovers compared to the national average? Even that would have to include caveats though, and you don't need look further than the ND game for them. As you said, some TOs have no influence on the game, and some aren't a result of a lack of skill (referring to Rees' woopsie fumble), but I think, generally speaking, good offenses will have fewer turnovers, and bad offenses will have more.

justingoblue

October 4th, 2011 at 4:13 PM ^

Good points, but the same question arises: what do you dock a team for turnovers when the original analysis was average defensive performance compared to performance against a particular offense? If you take a look at the WMU game, their turnovers (most likely) already penalized them by not getting their offense on the board, and thus decreasing their scoring and their place in scoring rank.

Also, I don't pretend to have a comprehensive solution to ranking offenses after having a free morning. The only thing I really did was see which offenses are scoring/racking up yards at a rate greater than the same defense is giving up in other weeks. So while it very well might be lacking from a turnover analysis, it isn't meant to be all inclusive, just one measure.

If I were to do this again, maybe have a TO ranking in the diary listing cumulative turnovers with the NCAA median included? That would give a look similar to the SOS list, and readers could take it for whatever they feel it's worth.

justingoblue

October 4th, 2011 at 5:18 PM ^

Seventy yards and seven points (with 10.2 as a baseline, so 160%) more than Wisconsin is used to. Wisconsin's defense probably won't face a better offense until bowl season or the BTCG, so it's natural that they gave up more to Nebraska than average.

I really don't think Nebraska is that bad, it's Wisconsin being that good, IMO.

FreddieMercuryHayes

October 4th, 2011 at 5:25 PM ^

I have a question about the Western total yardage number.  Specifically where are you getting 462 yards gained on us, if I am in fact reading the chart correctly.  You have have come across something similiar to what I found earlier today.  Specifically, ESPN lists WMU total yardage in the box score as 462, but I don't think that is correct (unless they did some extrapolation to a full game or something).  SI in the box score lists WMU total yardage as 279.  I noticed that in the ESPN box score, it lists Carder as the only passer with a total of 182 yards, but in the team total for passing, double his yardage, attempts and yardage to 366.  Maybe I'm missing something.  Care to help me out? 

Baldbill

October 7th, 2011 at 11:30 AM ^

Just me, if I were to use the stats from the Western game, I would have to fudge them a bit to invent a 4th quarter. Since Western is credited with 279yds at the time the game ended. Lets say that is all they got through 3 quarters, simply multipy by 4 and divide by 3 to give you the hypothetical 4th quarter total of 372yds. I would use that number not the other one which I can't figure out where it comes from.

XtremeUMich

October 5th, 2011 at 2:39 PM ^

Short answer: No

Long answer: NOOOOOOoooooo........

Western and ND were good warm up games, but just like the last year, beating the cupcakes doesn't make for long term success