frequency VS Pts scored in a game. so, hopefully the integral [0 ---> +infinity] of those distributions each equates to unity.
Apparently, this was the worst showing by U-M in the draft since 1994 when Derrick Alexander was the program's only player selected that year. People are using this as further evidence that the cupboard was bare when Rich Rod arrived on campus (as if anyone paying attention needed more evidence). But one draft doesn’t tell you much about the talent level of a particular team. For example, that 1993/94 team still finished 7-4 and 23rd in the Coaches’/AP. Why? Well, because that team also had three players who would be selected in the first round of the 1995 draft, and five players overall. If we want to know how bare the cupboard was when Rich Rod arrived, we also have to look at the 2010 draft. So, of the current players eligible for the draft next year, who other than Graham is likely to get drafted? What’s the fewest number of players drafted from a major program over a two-year period? Does this tell us anything about Rich’s cupboard that we didn’t already know? Obviously, it was bare but was it far worse than people realize compared with other major programs?
Yikes. A quick combing of Michigan's roster comes up with the following potential 2010 draftees outside of Graham:
So, yeah, it's Brandon Graham, a couple running backs, and the space punter. I don't know what the fewest number of players drafted from a power program over a two-year period is, but that's probably not the right question. The right question is "how many teams with like one high NFL draft pick and three or four mid-round picks are any good?" and the answer is "none, but there are plenty that didn't go 3-9."
This following one concerns variance, as discussed in the earlier post on Gladwell and basketball and Carr and the non-scoring offense. It's long, so I've chosen to respond after each paragraph. Though this looks fisk-y, it's not intended to be confrontational.
Your recent blog entry, detailing variance, risk versus reward, defense, offense and modern versus older systems, beginning with a basketball analogy, seems correct, but I have some issues. Your presumption seems to be that solid defense allows for a brute strength, low variance offensive strategy, in the style of Bo, and likewise with Carr. At the same time, however, you insinuate that a slow, grinding offense that keeps the other teams’ offenses off the field is of a critical nature towards that end.
I was not entirely clear about my thinking here. I do think that a really, really good defense allows for that sort of offensive strategy, and more specifically makes the run-run-probably-run-punt style of closing up a game make sense. In that sort of situation you're playing towards your strength.
However, when your defense is mediocre and you have a future NFL player at quarterback, shutting up shop and hoping your mediocre defense comes through is playing to your weakness. Carr did this a lot, if we're expansive about the word "mediocre".
As far as what sort of offense you want at the end of the game, yes, the sort of offense that can grind out a first down is nice to have, but if you don't have that offense—and not many do when they opposition is selling out like mad—you're doing yourself a disservice. There are specific situations where grinding it and punting makes sense, but none of them come with more than two minutes on the clock.
This is reasonable, as you can’t rely on a small lead and a low scoring game if you can’t keep the other team from scoring. The problem, however, comes in making the assumption that defense can’t be, or at least wasn’t, considered a weapon. Absolutely, using a prevent defense, clogging the running lanes, and keeping opposing offenses to short, clock eating runs between the tackles works towards that end. But what of the Michigan defenses through the years, especially in the early Carr era, that actually produced more variance, not less? Sacks, fumbles, and interceptions all increase variance in a game. Sudden turn-overs and backward yards are not supposed to happen on an offensive possession. I would say that in as much as a thundering, slow moving, ground based offense is designed to reduce variance, keep games simple and allow dominant talent to win out; the same strategy of good fundamentals (tackling, stripping the ball, pass coverage) has the exact opposite effect, creating lots of variance and unexpected.
Your definition of "good fundamentals" on defense varies from mine. When I think of good fundamentals, I think of a two-deep shell, minimal blitzing, and conservative strategies. Bend but don't break sort of things. A defense heavy on the blitzing and light on deep safeties is more prone to wild swings. And many of the things you cite as good fundamentals are zero-cost activities from a strategic standpoint: tackling, forcing fumbles, etc.
It seems that you’re positing that the more an offense scores, the more variable and therefore less predictable a game becomes. I think that’s the exact opposite of the truth. Offenses are supposed to score. To assume they will do ANYTHING but that is fallacy. I think the variance comes in when they fail to. Therefore, I don’t think that Bo’s and Lloyd’s game plans were low variance at all. I believe they simply tried to keep the variance, the sudden swinging changes, to one side of the ball. After all, if your defense FAILS to produce variance, the worst that will happen is the other team will score. That can be recovered. If your offense does produce variance, then the worst that will happen is you will lose your chance to score back. You can’t get that back.
This wasn't what I was getting at, but it wasn't the opposite of it either. What I was trying to say was this: all other things being equal, I'd rather Michigan play a game where both teams have sixteen possessions than eight. (Assuming that they don't suck, of course.) Michigan's more likely to come out on top in that situation. The way Michigan played under Lloyd, however, seems like it lent itself to a lot of long drives on both sides of the ball and generally depressed the number of possessions.
Simplified – You’re saying that offenses produce variance by moving quickly, scoring. As talent entropy occurs, this is harder and harder to stop, and so Bo and Lloyd saw their wins weaken, because their goals were to reduce variance. I believe that defenses produce variance by preventing scoring, and scoring on defense. We saw less success against higher level and middling teams in the last few years because talent entropy, and the coinciding spread of more complex, harder to stop offenses, has leveled the playing field, reducing defensive variance.
Different song, same title.
Okay, to properly address this we need to bring in variance's buddy: expectation. In layman's terms, expectation is the average of all expected outcomes. When you roll a die the expectation is 3.5. When you kick an extra point the expectation is 0.98. Variance is a measure of the average difference between trials. I could kick up the variance of the dice roll by turning 1 into –101 and 6 into 106 without affecting the expectation. I could kick it down by weighting it so that 3 and 4 came up twice as often as other rolls.
If you expect to win a game, variance is your enemy. I'm going to borrow some graphs from the excellent Advanced NFL Stats to demonstrate:
So here we've got two teams with the same variance in their play, one of which is a touchdown favorite. The underdog has about a 31% chance of winning.
Now the underdog has gone mad, probably going for it on fourth down a lot, inventing and deploying something called HELICOPTER PUNTING, and trying to block every extra point. They get blown out a lot more but also win more: 35% of the time.
This effect is powerful enough to overcome reductions in expectation:
But this time the underdog’s average is reduced from 17 to 16. The increase in variance still results in a slightly better chance of winning despite its overall reduction in average points scored. In this case, it's 33.2% for the underdog.
And it's the same for the favorite and reducing their variance: sometimes it's worth reducing expectation to get it, but only in certain situations and when you're a considerable favorite. In Bo's time, Michigan was a considerable favorite much more often and the game lent itself to low-variance moves: a 40-yard punt is much more valuable in an era when ten points is a potentially game-winning number.
Anyway, to the assertion above: modern offenses have more variance to them* because they score more. Don't lose sight of expectation here: Missouri had a lot of variance in their scores but that was because they averaged 42 points a game. Michigan had far less but they were averaging 20.
Offenses that do this quickly are actually more predictable because they get in more trials. Moving fast without sacrificing expectation is advantageous to the better team, which is why Oklahoma was in zero even halfway close games against the Big 12 rabble. (Texas is not rabble, obviously.)
Defenses reduce variance by, you know, having safeties that can tackle. The very best defenses are low variance because all of the outcomes have the same result for the opposition: shame and humiliation. In that situation, punting your ass off makes sense, because you're a big favorite, you're not giving the opponent much of an opportunity and you're reducing variance in a way that helps your overall chances of winning. The main problem with Michigan's defense over the last few years has been their suckiness, which by the way increases variance as your defense falls to a point where opponents can drive the field on them regularly.
I always go back here to the end of the 2005 Ohio State game: Michigan has a two point lead and drives down to the Ohio State 40. Facing third and ten, they run a wide receiver screen for six yards, and then punt on fourth and four from the 34, gaining 15 yards. Ohio State promptly drives the field for the winning touchdown. This came after a Henne-demanded fourth-and-short conversion on Michigan's 40 that led to an apparently-clinching field goal, and was interpreted by yrs truly as a panicked reversion to base instincts from another time.
*(The variance of something that's always zero is zero and it's not much higher for something that's almost always zero. As offenses move towards 50/50 efficiency the variance increases, but in a world like the 54-51 game against Northwestern the variance is low because everyone's always scoring touchdowns. An even distribution of probabilities is always more unpredictable than a set where most of the events are drawn to one or two outcomes.)
frequency VS Pts scored in a game. so, hopefully the integral [0 ---> +infinity] of those distributions each equates to unity.
*punch to the face*
A couple of points:
(1) Don't forget Donovan Warren, if he blows up this year and goes pro; not sure offhand of any other Juniors who might be flight risks; and
(2) Remember, the first day of the NFL Draft is now only 2 rounds. I like Minor and Brown a lot, but I don't have much confidence that either will go in the first two rounds. Shit, Beanie Wells was *barely* a first-round pick this year.
Color me skeptical that Minor or Brown go in the first two rounds of the draft, barring explosive years. So, hopefully they do, but it's hard to envision.
What color is skepticism?
it's a reddish hue. if you have the 600 crayon box it's original spot is in between "broken dreams blue" and "cynical orange".
So if you were colored skeptical, you might look slightly sunburned? (As opposed to being colored cynical, where you'd look fake-tanned.)
I dont know about future drafts, but I have never seen so many Michigan defensive players get taken in a single draft like we saw in 2007. Follow that up with the Henne/Hart/Long and a small group of quality defenders in the 2008 draft, and I think its the highest number of UM players taken over a 2-year span, especially in the smaller, 7-round drafts that we see now.
Thats a lot of talent. None of it seemed to be replaced during the 2007 seasons and then last year happened.
Call me crazy, but I'm glad we can have a 2-year stretch where we dont see a lot of kids drafted because it means our better players are still on campus.
oh god, that game hurts worse than any other loss in the past 5. we HAD that one
Agreed we had that game won and then just simply gave it away.
just had to bring it up and make us cry
Humble me. I knew you put a lot of research into your stuff, but dang!
Phil Steele, in his preseason rag, touts a draft formula, where teams get points for players picked and what round they get picked in.
If a team reaches a certain threshold of points (I cant recall exactly as I am going from memory here), then historically speaking they have like an 80 percent chance of having a worse record the following year.
I believe Michigan has reached that threshold each of the last seasons....and last year, they also dipped into another theory of his as it pertains to a low number of returning letterman and no returning starting QB, but thats another story.
The one interesting catch with Steele's NFL Draft formula is that if you're in that "group" with a high number of points, your record is likely to decrease. However, something like 7 of the last 8 national champs have been on the list as well. So, the exceptions to his rule do really well.
Replacing talent with talent thats the name of the game and why people follow recruitng so closely.
I like Phil Steele. No sugarcoat, the stats are there.
Maybe I am Caveman Football Fan and I dont know the ways of your fancy mathematical world.
I just look at rushing stats during a game.
An underdog who outrushes the favorite covers 2/3rds of the time.....so, when, sitting at the Big House, I just glance at the scoreboard and if Michigan (usually favored) is winning the rushing game, I'm chilling, loving a fall football saturday. When they're not, I get nervous, knowing that our foe is likely to do better than the experts think, putting them in more than close enough range to win.
I'll start counting drives, and bringing a notebook to calculate variance beginning on 9/5, I guess.
"The very best defenses are low variance because all of the outcomes have the same result for the opposition: shame and humiliation."
The very worst defenses are also low variance, as if you give up a touchdown each time your variance is also 0.
Brian - this stuff is the reason I read this blog...I think a slightly simpler way of saying some of the stuff you say would be to say that, if you have the better team, you want as many possessions as possible, b/c the law of averages will catch up to the other team...An inferior team may stop you on your first possession and may score on their first possession, but (if they're truly inferior) they won't be able to keep it up. The more possessions each team has the greater the likelihood of your superiority asserting itself.
Disclaimer: I'm not a statistician by any stretch. I get a nervous twitch every time I think about SPSS...
But...how can one team change its style of play without changing its opponents expected outcome? As Brian says, "maybe they get blown out more often," this implies that the favorite's expectation and variance have been affected by its opponent's strategy; hence, their expectations are not independent of one another. So, that means that while the underdog has changed its probability distribution for points scored, it has also changed its opponent's.
Taking risk-loving chances (trading a lower expected result for greater possible reward) necessarily favors the opponent on average, and will give it greater opportunity to score points. For example, going for it on 4th down in questionable situations rather than punting leads to, on average, better starting field position for the opponent, resulting in more points.
Am I on to something here or did I write my first mgoblog post only to sound like an idiot? (or both?)
You're making a valid point. Variance can be good or bad. A 60 yard TD pass FOR your team is a variance increasing event that is good. A 60 yard TD pass AGAINST your team is a variance increasing event that is bad.
"So, that means that while the underdog has changed its probability distribution for points scored, it has also changed its opponent's."
Exactly. Good post imo.
"But...how can one team change its style of play without changing its opponents expected outcome?"
As you worked out in your comment, you can't. Your expected outcome and your opponent's expected outcome are inextricably linked. The only question is whether it benefits you or your opponent to increase the variance on that expected outcome. For favorites the answer is NO, but for underdogs the answer is YES.
Smart Football has been musing on this for a week or so, and they go over what might be good high-variance strategies for underdogs. Check it out.
if you were to cross-post some of this to bucknuts.....!
Love this post. As others have stated, you just do not see this level of analysis on most sports blogs, and it is refreshing to have some mental exercise while trolling the interwebs.
My one question about all of this variance talk (and I apologize if I make some dumb observations), but how does one determine who is or is not an "underdog" or a "favorite", and does this analysis take into account shifts that might occur throughout a season?
I presume that the title of "underdog" is bestowed upon the team designated by odds makers, but that to me seems extremely subjective. For example, the 2009 Sugar Bowl had Utah designated as a 9-point underdog versus Alabama. There is no objective proof or characteristic that "proved" Utah was the weaker team; in fact, the 18-point Utah win perhaps "proved" the inverse. Yet, depending on how the teams were viewed prior to the match-up, this was either a major upset or an overlooked team proving to everyone how good they are.
From the variance calculation above, Utah winning would push the "underdog" variance up and depress the "favorite" numbers. Now, if enough people had realized how good Utah really was at the time, maybe they are the slight favorite and suddenly the numbers above are reversed and trends completely reset.
Now, I know that App. St. was a clear underdog AND the worse team (perhaps not as great a difference as some hope, but still the talent disparity favored UM), and that Oklahoma is much better than Baylor, but I suspect that quite a few match-ups are far closer, and that at least some of the "underdog" designations have more to do with generating gambling money than any tangible distinctions. Furthermore, an underdog in the first week of season might become a favorite a few weeks later once people realize just how good that team is compared to their preseason expectations.
I would be interested to know if these models take this inherent subjectivity into account, or at least how they determine what qualifies a team as a "favorite" and as an "underdog".
You are right to say that nowadays, clear underdogs and clear favorites are fewer and farther between...but Brian points that out as a weakness of Carr's, as he was unable to adjust his game strategy to take into account the increased parity that hit college football after the schollie limits went into effect. Back in Bo's day, the underdogs and favorites were much easier to identify, and so it made complete sense for Bo to pursue low-variance plays and strategies. Unfortunately, Lloyd never changed from that stance, and it probably hurt his chances at wins, especially late in games.
You would be wrong, however, to point to outcomes after the fact as some sort of repudiation of prior expected outcomes. After all, that's the whole point of variance- if you can toggle it high enough, you can actually trigger a surprise outcome. We call them "upsets" for a reason, and afterwards they don't get retroactively renamed to "outcomes that should have been expected."
So when looking at Utah-Alabama, I would point you to the high variance strategy Utah employed on defense (mainly because I don't remember what Utah did on offense). They blitzed the shit out of Bama, and it paid off with a bunch of turnovers, IIRC. That's absolutely a demonstration of high variance plays working to change outcomes.
I agree that Brian's point about greater parity in college football - arguably, that has been the case since the early-90's when teams went to 85 scholarships. My question has more to do with the determination of an "underdog" in these calculations. True, if you knock the variance up quite a bit, the "underdog" has a better chance to win and the "favorite" opens up more opportunities for the lesser team to pull off the win. But I use Utah as an example of the somewhat-arbitrary distinction between "underdog" and "favorite." Let's say that Utah was treated as the "favorite" in the Sugar Bowl - suddenly, that blitzing style would be deemed detrimental because they are the favorites and should be limiting the "variance" in their game plan. But then again, if they did not blitz and use the spread and no-huddle offense, then they would be a worse team and would, in effect, become the underdog.
I guess my point is that these numbers effectively protray variance from "conventional" offense and defense. On defense, that means the 4-3 or 3-4 with limited blitzing; on offense, the pro-style. Teams that run a spread offense or blitz more than usual are deemed "outliers" because, compared to the vanilla sets, they introduce more variability into the calculation. I'm not sure, though, if this is a true measure of talent disparity/favorite vs. underdog as it is a measure of non-traditional offenses and defenses versus the standards.
We call them "upsets" for a reason, and afterwards they don't get retroactively renamed to "outcomes that should have been expected."
It all depends on what's the true expected outcome. Ohio State was favored by a TD against Florida in the 2006 title game and got absolutely waxed. If they had replayed that game the following week, does anyone think Ohio State would still be favored? There are times, not often, that Vegas and general public perception is just flat out wrong.
Utah's offense (in the Sugar Bowl) was no-huddle fast tempo and mostly throwing the football--not exactly a recipe for minimizing possessions and believing you're an underdog. They believed they were just better than Alabama (and FWIW, I think they win that game 6 times out of 10) and played their game accordingly.
Brian, I'm very disappointed in this post. Not once did you mention:
- What school has the hottest chicks
- Who the hot MGoBlog chicks are
- Chicks, man
I think the OT's you're referring to are started and dominated by guys who aren't getting any.
capitalistic plan on Brian's part to launch the highly profitable spin-off, MGoHorny.
Remember the Ferrari 250 GT California that Morris owned but never drove? Carr's offenses were a lot like that.
You watched TMZ today
I missed it. Was it anything worth watching?
well...it's always funny, but you always feel shameful for watching it. so, worthwhile? never.
anyway, Cameron from Ferris Bueller was on that's the only reason I say that.
I know what you mean about that show. They are funny, but it is always at someone elses expense.
I read a SI article about Gordie Howe. In it he described how the producers of that movie asked him if they could use one of his jerseys in the movie. He later said it was cool to see his jersey on the big screen and that he "scored a lot of goals in that jersey."
oh that's cool. Did not know it was an authentic Howe jersey.
I found the article again. I misquoted him. He said he had a lot of proud moments in that jersey. I assumed he scored more than a few goals.
In response to bronxblue, I think these variance-maximizing/minimizing strategies mostly apply when there is an obvious favorite. When the teams are close in strength they just have to maximize their expectation and hope for the best.
Having said that, Vegas is generally right about who the favorite is. If you could reliably pick out situations where they were wrong, you would be a millionaire.
Rock on Brian!
Yeah, Woody/Bo/Lloydball sucks. Unless you're Tressel.
I have been saying this for a good year now...I am posting this link for all of the stupid people I know. Of course, then they would have to read something.....but anyway, good to know there is someone out there who gets it too.....
In the new Star Trek movie Spock wants the logical choice where Kirk wants the illogical. As it is always is in ST so no big spoiler there. But it comes down to the old EXCEPTION/RULE argument. People are already arguing that the spread is the exception and a gimmick. Where Carr is the rule and more true. One winning season and a bowl win though and many flip flop from that stance. Some begrudgingly, some after their first beer. And so goes the legacy of a college coach. Thus the reason why Carr was a man of rules and loathed even contemplating the idea of exceptions. And we all know or could guess what were Carr's exceptions or things he thought College Football should be without. Carr had a formula, and it was from a different time. Perhaps a better time. In season one there was little merging, to use a driving term. There was a lot of quick exits and quick onramp entries though. It was a bumpy ride. MY HOPE.......is that as fans, and I am also hoping the two coaches( LC & RR ) now see is that the answer, THE REAL NEW DIRECTION OF MICHIGAN FOOTBALL lies somewhere between these two poles and not at either extreme. When you are fluid and loose you use what you have and whats thrown at you and when you're tight and rigid you do the opposite. And in football that translates, usually into more losses than wins. Ok I'll shush it, damn the green tea.
All these comments about the cupboard being bare are nonsense.
Loyd before he left had recruiting classes that never was lower than top ten and as high as # 2. Loyd didn't leave it bare, the administration in deciding to hire RR changed the whole system to such a degree that the palyers previously recruited did not match up well. This is not Loyds fault. Loyd was the besyt coach in the past 30 + years. He won a partial NC and bowl games unlike his predessor.
While I partially agree that the notion the cupboards were bare when RR arrived is foolish, it is also clear that Carr's last few classes lacked the depth and breadth of some of his earlier classes. Far too often, his classes featured one or two big-name guys followed by a bunch of middling players. I'm not saying that is a failing of Carr's - most recruiting done at the assistant/position coach level, so any blame for a drop off probably should fall on those coaches. That said, the team RR inherited lacked the type of depth and top D-1 talent that other elite programs generally have. That's a big reason the Horror and the subsequent final season happened - that team had some big names, but it lacked the good players across the board that step up when necessary. I think the shift in idealogy had something to do with last season, but this team lacked depth on both lines and in the secondary, and that falls on the past administration.
I still think Carr was a great coach and motivator, but remember that he won his NC largely with Moeller's kids, to be perfectly fair. His cupboards were clearly better stocked than the team handed to RR.
"That said, the team RR inherited lacked the type of depth and top D-1 talent that other elite programs generally have."
That is putting it mildly. And since I am new to this blog I am sure I missed all of the discussions on last season so I am not going to rehash it...I really want to move on from it anyway! However, I am encouraged by what I am seeing in the recruiting at this point. But waiting to see which players actually put their pens to the paper come February.... should be another interesting season!!
RR inherited a # 2 recruiting class in 2005, # 9 in 2006, # 10 in 2007 and # 6 in 2008. Don't use the excuse that the cupboard was bare. RR instead of easing the team in to the spread, jumped in with both feet. He is responsible for the 3-9 last year and has had a great deal to do with the departures, ignoring mant of Blue's traditions and wanting it his way or the highway.
lot of it was gone (per Rivals):
Antonio Bass 4*
Jason Forcier 3*
Eugene Germany 4*
Mario Manningham 4*
Chris McLaurin 3*
Chris Richards 3*
Justin Schilano 4*
Johnny Sears 3*
Mister Simpson 3*
Marquis Slocum 4*
Corey Zirbel (4*) was injured prior to the season.
The star of the class, 5* Kevin Grady, has been a bust.
That's a lot of talent gone from that class, but most noteworthy are the first two names, the QB's.
....have a headache reading both of your posts.
As another poster has recognized, the rankings of your respective recruiting classes mean little if the players which formed the basis for that ranking are no longer on the team. That's not Coach Rod's fault.
As for "jumping in with both feet" to the system that he has successfully operated for the past 20 years, well, flip it around. Imagine if he decided to keep the system as it was or modify it slightly to "accommodate" the 2008 system. The 10 new offensive starters would have learned the old system, but guess where we'd be this off-season? Instead of hearing about how good our offensive linemen are playing, how improved they are, hearing about the wideout's improvement, we'd be hearing about growing pains in learning a new spread offense. I'm personally glad that (a) that headache is further in the rear view mirror; (b) that Tate is going to have 5 returning starting linemen, and about 13-14 linemen in total, who now have over a year and a half in Coach Rod's system, in order to better protect him, and (c) these receivers will be more experienced in running Coach Rod's routes.
I recently watched the 2008 Michigan - Wisconsin game on the BTN. One play in particular stood out to me (well, other than the exciting comeback plays). It was on the drive when Minor, was major, and scored the long TD run when the UW linebackers blitzed. Either the play before or two plays before, there was a split screen with Threet and Coach Rod, and you can hear Coach screaming at Minor to get lined up in the correct position. Coach is going nuts. The end result is Minor catches a wheel route outlet pass and gains about 10 yards.
Coach Rod's system works when the players execute. That comes with experience. Your complaint about "jumping in" would have caused the 2009 season to be even more perilous than it currently is. The players got experience last year, there is a QB used to running this kind of offense now, and the offense will be much better this year as a result.
I LOVE seeing RR get all crazy on the sidelines....I expect that from a coach!!!
As was noted below, a lot of those kids were gone by the time RR arrived.
Furthermore, rankings are always somewhat subjective. A middling 3-star kid who receives an offer from a "name" school will usually make a sudden jump in said rankings, the logic being that if team X sees something in him, then he must be good. I call that the Notre Dame Effect - every year, Notre Dame will recruit some middling kid (like Mike Golic's kid this year) who, because mighty ND thought so highly of him, will suddenly receive a boost in the rankings to better fit with the rest of his class. Well, that phenomenon happened at UM as well, and it seemed that toward the end of Carr's tenure he was getting some "top classes" in ranking only. I'm not trying to knock the guy - he remains a very good coach and a great motivator/leader, but look at those last few seasons. Outside of 2006's magical run, I think the team was showing some atrophy in recruiting, with a few "big names" overshadowing a dearth in D-1 talent. You might disagree, but just pointing to the rankings as proof that RR screwed up a good thing is not enough.