"Rodrick Williams Jr.'s 10-month old, 2-foot-long savannah monitor named "Kill" gets the RB some strange looks when they go for walks together."
Two years ago I piled onto the idea that coaches don’t maximize on fourth down. Or at least they don’t maximize to scoring the most points. More likely, they minimize the chance of getting fired for making a risky decision.
Within the post I noted that coaches are too conservative and presented an initial treatment to what the decision matrix should look like. I also identified several of the key objections to the “Stop Punting” movement. Over the last two years I have accumulated more data, thought about the topic further and think there is more to add to the conversation. The general conclusion is the same but I have added some new tools that should shed some new light on the subject and bring some power to the people.
A quick review of the main objections:
4th Down stops don’t account for momentum change
I still haven’t found that this has a quantitative impact to anything. Last month I found that teams go for big passing plays more often but that they don’t actually score any more points.
Assumes all offenses and defenses are average
This is a big one. Two years ago I gave this a surface review but will go into more detail below.
Doesn’t account for in-game situations
It’s a valid critique but one that is rarely applicable and easily approximated. With more than 2 possessions left and a competitive game means expected points are directly correlated with win likelihood. A close game late or a blowout will usually point you where you need to go with common sense. In game situations matter, just less than you think they do.
Men Offenses, Defenses and Kickers are Not Created Equal
This is the big adjustment. Rather than looking at the very best and how their expected value differ from average as I did last time, I made the tool dynamic. I can now look at how each situation changes when you have a dominant offense, or if your defense is a sieve. It can also account for kicker quality. A really good or really bad kicker definitely changes the mindset. Announcers generally frame the decision as “you have to take the three points” even though the odds of success are significantly less than 100%. A great kicker will push the decision closer to that thinking for a quarter of the field. Likewise a terrible kicker takes the field goal out of play for a large portion of the situations.
So to see how different strengths and weaknesses affect the outcomes lets start with the updated baseline chart:
Maximum “go” territory is around the Opponent’s 32 yard line where even on 4th and 10 going for it might be the best play. Once you hit the 20 yard line the field goal is the best option unless you only have one yard to go. This stays true until you get inside the 10, then the opportunity for 3 yards or less becomes more optimal.
So what happens when you get a large variance in teams’ offense and defense. A great example would be the infamous Michigan 2010 team. The offense was as great as the defense and kicking game were terrible. Plugging in extreme values for all three dramatically changes the decision matrix.
Michigan 2010 should have gone for it early and often. Field goals should not have been attempted of 40 yards or more and if the yards to go were 6 or less the offense should have always stayed on the field.
Michigan 2011 was a dramatically different team from 2010. The offense was still good, but not as great as 2010. The defense stopped being a tire fire and the kicking game was much more reliable. I adjusted the offense down a bit and the defense up considerably and moving kicking to average.
The end result is a chart that is somewhere between the baseline and the Michigan 2010. Field goals become the best option when the distance to go is greater but aggressiveness when crossing the 40 is still warranted.
Life, Liberty and the Pursuit of Proper 4th Down Strategy
As part of this work, I have built a public Google Spreadsheet that you at home can use to overanalyze your favorite 4th down situation, complete with dynamic skill ranges.
Each situation adjusts the 4th down conversion chances and upside and downside of any decision. It outputs what the best decision would be and how many expected points you would be leaving on the table with the “wrong” choice.
The offensive and defensive ratings go from +5 to -5 with 0 being average. They are a measure of the difference between the offense and the defense. Michigan’s offense should be a +5 against Minnesota but probably a -2 against Alabama. The rating affects both the odds of conversion and the value of the resulting 1st down if successful. Defense is the same in that it is a measure of the difference between your defense and the opponent offense. Adjusting the defensive metric affects the value of a resulting possession change whether by missed FG, punt or missed fourth down conversion.
I hope you enjoy looking at this and I included some special Easter Eggs for the curious.
Close followers of this site are well aware that for all his faults, Rich Rodriguez was not handed a classic, loaded Michigan roster when he arrived. If you followed his recruiting at all you know he didn’t do the Wolverines any favors with his recruiting during his three years in Ann Arbor. Based on 11 years of recruiting and roster history, I have compiled a look at how Michigan’s talent and experience have ebbed and flowed since 2006 and how many elite caliber classes like the currently assembled 2013 class it will take for Michigan’s talent base to catch up with the nation’s best recruiting programs.
For those interested in how the roster values were calculated, I’ve added a section to the end to explain the methodology.
The Lloyd Peak
How long before he retired he knew when the end was is hard to know, but Lloyd Carr certainly had the roster set up for a great 2006 and 2007. In 2006 Michigan entered the season with the highest rated roster in the Big Ten and trailed only LSU and Miami (YTM) for top marks nationally. Both sides of the ball were well represented, ranked third nationally on both offense and defense. The stacked lineup led Michigan to the brink of their first BCS National Championship game before losing to top ten talent teams in the final two games of the season.
After a disappointing finish to 2006, the roster was even more setup for a big send off for Lloyd Carr in 2007. The roster trailed only USC in terms of overall talent and the offense had a sizeable lead as #1 nationally. Unfortunately, the losses on the defensive side dropped the defense out of the top 10 and were exposed in the first two games, ending Michigan’s national title hopes before the calendar turned to October.
The RichRod Hangover
The 2008 version of Michigan was not a team lacking talent (except for quarterback), but it was a ways from the talent level in Ann Arbor the previous two seasons. The defense rose a bit to sneak back into the top 10, where the loss of the Henne/Hart/Long dropped the offense from #2 in the nation to #16. By the 2010 season things hadn’t gotten better but they hadn’t gotten worse, either. The offense and the defense consistently ranked in the top 15 nationally but the results, especially the GERG effect on the defense never matched up with the talent-level on campus. The roster left for Rodriguez didn’t do him any favors but it was also far above the record and defensive performances that came from it.
The High of Hoke…
…will have to wait. As great as the Hoke recruiting era has begun, it will likely be a few seasons before the dividends begin to payout. The next two seasons will be the two where the roster effects of the Three and Out era will come due. This season the talent level will drop down into the 20s overall and likely stay there for the 2013 season as well before a big jump two years from now when the majority of the Hoke classes move from observers to contributors.
Michigan Vs The Rivals
Michigan dominated the end of the Lloyd era versus all the big three rivals but by 2008 Ohio and Notre Dame both moved into a better roster position. Despite the head to head losses, Michigan maintained an overall roster advantage over Michigan State. That nearly closed over the next two seasons, before the Hoke recruiting pushed Michigan out front once again. By 2015 Michigan should move close to par with Ohio and Notre Dame for the first time since 2008.
Michigan Vs The Best
Despite a lull coming up over the next two seasons, Michigan is poised to move into the upper echelon of national programs in terms of overall program talent. If Michigan can maintain top 5 level classes, by 2015 Michigan will move into the elite group of rosters in the country. Michigan’s projections begin to look a lot like that opening day opponent, with a five year delay. Nick Saban had similar rosters his first two seasons that Michigan is facing in the coming years, before taking off into elite status. Hopefully Michigan doesn’t need to go through the roster manipulation to get there, but with the first two Hoke classes, Michigan is making the first step to achieve a similar roster look as the team that has won 2 of the last 3 national championships.
What it All Means
Even the most die-hard numbers guy like myself knows that at the end of the day these are just numbers on a computer screen and the game is played on the field. The numbers are far from everything. Coaching, luck, player development and other factors are all major contributors to team success. With that said, the composition of the roster does mean a lot. As far back as I have reliable data for roster composition (2005) only Auburn has won a National Championship and not been ranked in the Top 10 for roster talent, and it took one of the greatest individual single-season performances ever for that to happen. Conference championships can be won with middle of the road talent for the conference. Last year Wisconsin won the Big Ten with the 7th highest rated roster in the conference. But over the long haul talent will win out.
Michigan will have more to overcome in the talent department than at any other time in the internet era. 2012 and 2013 will be lean years by Wolverine standards. Thankfully, we have the kind of coaches that have a high likelihood of mitigating that drop through their other talents of player development, in game coaching and for the future with recruiting.
Unlike Lindy’s I’ll do my best to give you a rundown of how I arrived at the valuations for each roster. Each player when recruited is ranked by each service. If they fall out of the Top X for any site, I do my best to approximate a ranking based on stars, position rank and grade where available. That rank is then translated into points. The formula I use is -4.5*ln([Rank]+11)+36. The specific formula was generated to more heavily weight the top players and produce an output where a consensus #1 rated prospect is worth a total of 99 points (summed across the four sites) and bottom of the barrel anonymous 2 stars are worth only a couple points.
Using this system, Michigan’s highest rated recruit was Prescott Burgess in 2003 at 90 points, Kyle Kalis was tops for the incoming class at 69 points and Shane Morris is currently at 79 points. Players are then weighted based on years in the system. Freshman only add 25% of their points to the roster total. 2nd year players (red-shirt freshmen or true sophomores) add 75% of their recruiting value to the overall total and all players in at least their third year in college football get 160% of their recruiting value. These numbers were derived based on actual usage of running backs (split of carries), quarterbacks (split of attempts), receivers (split of catches) and defenders (split of tackles) and then normalized so that the total roster number would approximately equal the sum of the unweighted recruiting points.
The total only counts players still on the roster at the beginning of the season. On the recruiting side a thousand points is usually at least a top ten or better class (The 2013 class is sitting at 1034 points). From the charts above you can see that the best rosters reflect 4-5 great classes and are 4-5,000 points.
If you’ve read enough of my pieces you know I don’t put a lot of stock in most of what football announcers talk about. Whether it is harmless and/or mindless cliches or things that are flat out wrong, much of football announcer conventional wisdom is more conventional than wisdom.
One of those cliches I wanted to look at was the concept that coaches love to go for the kill shot after a big momentum change like a turnover or a stop on downs. After a big defensive play, are coaches really trying to seize the opportunity and turn an expected possession into a quick score, and if so, is it working.
One thing to note is that a quick change does not significantly impact the offense’s ability to score. After adjusting for field position, there is virtually no difference in the offense’s expected points whether the drive was obtained by turnover, fourth down stop, punt or kickoff. Of course the field position is a win, but intercepting a ball 40 yards downfield on third down yields no existential benefit over knocking it down and fair catching a 40 yard punt on the next play. Punts and stops on fourth down have very slight positive impacts and turnovers actually decrease a team’s likelihood of scoring, but the effect is so small it’s typically not even worth a field goal over the course of an entire season.
So there is no special advantage of the big defensive play for the offensive side, but how often does the offense attempt to capitalize in a big way right off the bat. To evaluate this you have to figure out what you can measure. There aren’t really running plays that are designed to be big plays, except maybe a reverse, but that’s hard to identify over a large set of data. Same with trick plays. You can’t even tell where a pass is thrown. Was it a screen or a deep ball. Both could be incomplete or big plays.
To try and and answer the question the best approximation I could find was how often the first play of a drive was a pass and went for at least 20 yards. More big passing plays than in other situations would be a good indicator that teams were gunning for a quick strike. For the baseline I looked at drives obtained via kickoff, either after a made field goal, touchdown or start of a half. For these drives, only 8.9% of first plays were passes for at least 20 yards.
Drives resulting from a turnover on downs generated the highest deviation in successful attempts down field, producing big pass plays nearly 35% more often than the baseline scenario. Interceptions and punts were right behind with a 32% increase in big play generation. Coaches were still aggressive but to a lesser extent after a recovered fumble or an opponent’s missed field goals, producing big completions just over 20% more often. Despite safeties being a big momentum swing, that was the most conservative scenario for coaches long passes being less likely than any other scenario, even after adjusting for less field to work with. Even with large deviations, we still aren’t talking about this being a regular occurrence. The baseline is just under 9% and no situation generated more than 12% big pass plays.
So even though the question was answered that yes big plays do happen more frequently in this situation, there are still many things that I am unsure about. Is the increase in big plays because the defense is unprepared for the cliche move or because the opportunity for a big pass play is always there but the offense just doesn’t go for it as much in normal course of play. On top of that, is the increase success coming at a risk for offenses. As noted above, despite getting more big plays on the first play of the new drive, offenses aren’t actually scoring any more points on these drives. Maybe for each big momentum piling pass there are two incompletions that put the offense in an unnecessarily risky situation.
Having exhausted the stretches of my data my personal conjecture would be that yes offenses are going for the jugular at a higher rate immediately after a big defensive play. However, this strategy is probably a high deviation, zero average change result for both offense and defense. The evidence points to teams taking more and completing more big passes, but the untold story are the misses which are the likely culprit to the big defensive play not translating into any measurable offensive boom.
Based on the data and my inferences off of it, if I were advising coaches I would not recommend introducing the new risk to the offense and play it straight unless I was in a trailing or underdog situation. Defensivley, I would make sure the team is prepared to cover a big attempt. This should help reduce the likelihood of the big play for the opposing offense and hopefully increase the likelihood of a wasted down, putting the quick start defense in a good position right off of the bat.
Last month I took a look at how the recruiting services stacked up when it came to predicting future college success. With the latest NFL Draft in the books I wanted to review it to see if there were major differences between services predicting the draft versus college success.
Even though NFL draft position doesn’t always equate to success at college, it is an indicator of talent and there are very few instances were great college players don’t ultimately end up in the NFL, mainly highly efficient noodle armed quarterbacks. Even rushing quarterbacks typically find their way onto draft boards where the Kellen Moore’s and Case Keenum’s of the world sign on as undrafted free agents.
Biggest Hits & Misses
Here are the players from the first three rounds that one service rated significantly different than the other two.
ESPN had significantly more outliers than other two services. Luckily, the hits outweighed the misses. ESPN was the sole advocate for five first rounders and had more hits (9) than the other two services combined (7). The Worldwide Leader also racked up the misses at a record pace, missing on six top picks.
Morris Claiborne, Ryan Tannehill, Luke Kuechly, Kendal Wright, Whitney Mercilus, Courtney Upshaw, TJ Graham, Mohamed Sanu, Sean Spence
Matt Kalil, Mark Barron, Nick Perry, Mitchell Schwartz, Cordy Glenn, Isaiah Pead
Scout had only four players from the first three rounds that they significantly diverged from the other services on. Their divergence was centered on receivers as three of the four players they deviated on were wide outs.
Justin Blackmon, Stephen Hill, Olivier Vernon
The players were different but Rivals hit and miss counts were virtually identical to Scout.
Quinton Coples, Brandon Taylor, Jamell Fleming, Jayron Hosley
Round by Round Scorecard
How each round went, points given for whichever service had a draft pick rated highest (must be in the top 1000 ranked). Numbers listed as Rivals/Scout/ESPN.
Round 1: Push 8/9/9
Round 2: Rivals 11/4/6
Round 3: Rivals/ESPN 8/3/8
Round 4: Scout/ESPN 6/12/10
Round 5: Scout 5/9/3
Round 6: Rivals/ESPN 5/3/6
Round 7: Scout 8/12/6
Overall the wins were pretty evenly balanced across the services. Rivals taking at least a share of the first three rounds would probably give them the overall edge. They also had the lowest overall average ranking of draft picks.
Scorecard by Class
Class of 2007: Push 18/18/15
Class of 2008: Scout 22/29/26
Class of 2009: Rivals 10/6/5
Despite their overall strength, Rivals struggled in the main class of the draft, 2008.
The look at the NFL draft results versus the recruiting rankings mostly reinforce each other. Based on the data from both looks, here is how I would describe each service’s strengths and weaknesses.
Rivals: The best and most consistent performer. Year after year producing output at or above the level of any other service.
Scout: A very good second to Rivals. Similar consistency to Rivals and an overall level that is close but still a notch below. A quality second voice.
ESPN: The all or nothing service. Demonstrated that overallocation to the SEC that isn’t necessarily supported by the NFL draft. Blatant neglect of west coast recruits but willing to deviate from the other services, with mixed results.
[Ed-Ace: Brian (knee) is day-to-day, though he did prepare some content that will be posted this afternoon. Post-Burke-return hoops stuff and a Spring Game primer will appear later this week. In the meantime, enjoy some Mike Hart.]
In honor of Michigan’s all-time leading rusher’s birthday yesterday, a look at one of the unique careers in college football.
Since the 2011 season completed, I have been re-loading 9 seasons worth of games (6,063 to be exact) to update my database to include 2011’s new feature of Win Percent Added. In doing so, something immediately popped out at me. No running back added more wins to their team than Mike Hart did for Michigan.
Sometimes when you are looking at advanced stats you are surprised by how counter-intuitive results can be and sometimes you are surprised how well the data fits the existing narrative. Mike was the back who wouldn’t go down, always got the extra yard, killed the clock and never fumbled. Those are all the things that factor highly in Win Percent Added, especially the 4th quarter capabilities. Burning the clock in the fourth quarter is a key requirement of a successful running back. Especially a Michigan running back. No one did it better than Mike.
For his career, Mike Hart was responsible for 4.4 Wins running the ball. Reggie Bush edges him out if you count receiving WPA, as well, but those are tainted wins. It’s not just longevity and playing time that pushed him to the top. His per game average of 0.11 is fifth, behind two players with only a single season in the database and two more with two seasons at non-BCS level schools.
At this point, writing about Mike Hart is a daunting task. What is left to write that hasn’t been written? He joined the team in the 2004 class as a 3 star recruit. He nearly set the national high school rushing record but wasn’t even the highest ranked running back in Michigan’s class. He would have been the fifth highest rated running back in Miami’s (YTM’s) recruiting class. He saw his first quality action in his second game of his career against Notre Dame in the second week of the season. By week three he was over 100 yards and posting a +5 EV+ and a crucial .36 WPA as Michigan held on for a 24-21 win over San Diego St.
Hart would go on to string together three straight 200 yards games in Big Ten play, including a 0.26 WPA in the Braylon Edwards game. His EV+ was always strong for a running back but where his EV+ was strong, his WPA was Herculean. Mike Hart made all the plays to win the game but none of them to lose them. By the end of the 2004 season true freshman Mike Hart had gone from anonymous three star to posting a per game WPA of 0.15, still my best recorded number in the Big Ten.
Injuries killed a large portion of the 2005 season. Kevin Grady, Max Martin, Antonio Bass and Jerome Jackson all took carries but none could come close to the production from Mike Hart. Kevin Grady was the only one to surpass a +1 EV+ in his absence, and that was mostly unnecessary against Indiana. Jerome Jackson did have a solid 0.14 WPA on 11 carries in an overtime win against Iowa, but that was limit of the success when Hart was out. In five full games of action Hart averaged 0.23 WPA which if replicated across an entire season would have given him the second highest (Reggie Bush, 2005) WPA average in a season for any running back since 2003.
It’s hard to think about what could have been with a healthy Mike Hart. Three carries in a seven point loss to Notre Dame, a DNP in a three point loss to Wisconsin eight ineffective carries in a four point loss to Ohio. There’s a very real chance he swings those three games and Michigan shares a Big Ten title with Penn State and spends its holiday taking on Florida State in the Orange Bowl rather than getting screwed over by the refs in the Alamo Bowl.
With fewer games coming down to key fourth quarter possessions in 2006, Michigan didn’t need the fourth quarter machine Mike Hart. He finished the season with a profile almost exactly like Chris Perry’s 2003 season. With not much in the way of close games, he didn’t have any massive, WPA pushing games like he had in his first two years, but 10 of 13 games would finish at .07 or better. For the year Hart ended at .09 WPA/game, his third top 20 Big Ten WPA year in as many tries. John Clay is the only player to have even 2 top 20 finishes.
For the second time in his career, injuries would derail an outstanding Mike Hart season. After surviving The Horror and somehow managing a strong WPA in the follow-up beating by Oregon, Hart was on track for a season to along side his junior year. An ankle injury in mid-season cost him a couple games of action and a couple more of effectiveness. 2007 would be his lowest rated season but still crack the Big Ten top 50. He would finish the year with enough quality carries to become Michigan’s all-time leading rusher and set the then non-existent WPA record.
When I talk to people about how much more valuable quarterbacks are than running backs they usually point to running out the clock in the fourth as the unquantifiable equalizer between the two. When I first developed the Win Percent Added I was anxious to see how true it was. If you properly value the ability for a running back to keep the clock running and close out a game, what happens to the value relationship between quarterback and running back. After I crunched the numbers I found that the fourth quarter benefit was largely overstated. Until I looked at Mike Hart. There are very few running backs whose value is truly magnified by the little things like the narrative claims.
Mike Hart is the narrative.
Mike Hart, Seasons
Mike Hart, Games
|Year||Week||Vs||EV+||WPA||Rush EV+||Rush Att||Yards|
|2004||3||San Diego St||5||0.36||5||25||121|
As Michigan enters year #3 with Denard starting the season at the helm and year #2 under OC Al Borges, I did a deep dive into teams that have made the leap into offensive greatness. A handy guide to going from a good offense to an elite offense in one year.
In making predictions you are always safer predicting things to regress towards the mean. I wanted to look at the teams that have gone from good but not great offensive teams to truly elite teams in one season. There were some surprisingly strong correlations within this group.
1. A quarterback with experience returning.
2. The same offensive coordinator as the previous year
3. Continuity in your receiving core
These three three aren’t sufficient conditions for making the leap, but they are necessary conditions.
From 2004 to 2011 there were 20 teams, including 2010 Michigan, that increased their offensive EV+ rating by at least 4 points per game and ended the year above 10. I wanted to take a look at these 11 teams to figure out what was necessary to make a jump like this, what might seem necessary but wasn’t and how many teams fit this same criteria but couldn’t make the leap.
It’s not rocket science to say that having a returner at quarterback is a good thing, but the exceptions to this rule prove it as much as anything. Of the 20 teams to make the leap, five didn’t technically have a returning starting quarterback. Two, Auburn 2010 and Stanford 2009, did it with future first overall draft picks. Wisconsin 2011 technically counts but Russell Wilson was a three year starter for NC St before arriving in Madison. Tim Tebow got quality field time on a national championship team despite Chris Leak’s position as starter. The final exception was Oregon 2010 when Darron Thomas stepped in to great success.
The returners themselves take all forms. There are runners like Josh Nesbitt from Georgia Tech in 2009, Pat White at West Virginia in 2006. There are pass-first athletes like Robert Griffin on Baylor 2011 and Vince Young on Texas 2005. There are future first round picks like Sam Bradford at Oklahoma in 2008 and Matt Leinart at USC in 2005.
Three times it was done with multiple quarterbacks seeing action. Michigan in 2010 saw returning starter Tate Forcier make regular appearances during the Denard Robinson injury hour every Saturday. Cincinnati didn’t miss a beat in 2009 when Zach Collaros stepped in for injured starter Tony Pike. Arizona State had the same story in 2005 when Rudy Carpenter replaced the injured Sam Keller.
How many years at the helm wasn’t a significant issue. There were as many teams who made the leap with second year starters as third and fourth starters combined. The only team to do it with a non-transfer four year starter was Northwestern in 2005 under Brett Basenez. Zac Robinson did it in his second year at Oklahoma St in 2007, Chase Daniel did it at Missouri in his third year of four running the offense at Missouri.
Not all the QB’s developed within the program, either. Cam Newton, Russell Wilson, Aaron Rodgers and Ryan Mallett all accomplished the task after transferring from a previous school or Junior College.
On average, about 35% of BCS programs turn over their offensive coordinators in a given year. For teams making the offensive leap it drops to 10%. Charlie Weis found his famed “Decided Schematic Advantage” in 2005 (only to lose it thereafter) with Brady Quinn. Arkansas and Ryan Mallett where the only others to accomplish the feat, but the presence of head coach and offensive guru Bobby Petrino indicates the continuity was likely high.
Teams making the leap returned 75% of their value from their receiving core, nearly fifty percent higher than the average team (53%). Only four teams returned less than 60% of their receiving value from the prior, and those teams all managed to make the leap due to ultra efficient passers, Tim Tebow, Russell Wilson, Aaron Rodgers and Vince Young.
Missing The Cut
Returning Running backs
My crusade to bury the position of the running back continues. Teams making the leap where all over the map on returning RBs. Texas and Northwestern both made the leap in 2005 with virtually no non-QB carries returning from the 2004 season. Baylor, Michigan, Florida and Cal all did with only marginally used carriers returning. On the flip side USC returned Reggie Bush and LenDale White in 2005. Georgia Tech, Oregon, West Virginia and Missouri all returned over 80% of the previous years carries.
It’s not that having a key ball carriers returning is a bad thing, it’s just not necessary to make the leap.
Just like the running back question, the recruiting rankings are mixed bag. Cincinnati, West Virginia, Cal, Northwestern and Baylor all made the leap without the benefit of a roster full of 4 and 5 star recruits. USC, Auburn, Florida, Texas and Oklahoma always enter a season stacked. Better recruits = better results is true but Better Results = Better Recruits is often incorrect.
Odds of Making It
Of teams that meet criteria 1-3 above (with 3 being defined at 60%) and are within 1 standard deviation of +10 EV+ (4.4 or higher) 25% make the leap to +10 or higher. 40% of this group improves but not to an elite level. That leaves 35% of teams to regress under this environment. The worst offender is Michigan State in 2006. Poised to become a potentially great offense under Drew Stanton, Sparty fell apart and dropped from a +7.2 in 2005 to a –2.6 in 2006.
Can Michigan be that team
With the loss of Junior Hemingway, Kevin Koger and Martavious Odoms Michigan falls below the 60% threshold (38%). All of the other criteria fit nicely for Michigan and when it has been done without the aid of returning receivers, its come on the back of a Heisman level quarterback performance. If year 2 of Al Borges Denard Fusion Cuisine comes together, that doesn’t seem out of the realm of possibility. I won’t be on record predicting a leap like year for the 2012 Michigan offense, but they are one of a handful of schools that I would even consider.