In trying to analyze the ACC basketball landscape for my own blog, I came up with a thing I decided to call Impact Score. It measures (a bit abstractly, since the numbers don't have units) a player's positive or negative impact on a basketball game. I figured it would be handy for the B1G, too. It's a simple formula:
(O-rating - 100) * Min% * (Poss% / 0.2)
All numbers are from KenPom, naturally. Explaining each:
-- O-rating minus 100 assumes that an average player would have an O-rating of 100, and generate one point per possession. The difference between a player's O-rating and 100 is their impact, for better or worse, compared to an average college player.
-- The number is reduced by the percentage of minutes the player plays. No impact when he's not on the court. All else being equal, a player who plays the whole game has twice the impact of a guy who sits out half of it.
-- Poss% / 0.2: There are five players on a court, so theoretically they should be responsible for about 20% of their team's possessions, and in fact, most players hover around that number. All else being equal, a guy whose poss% is 22% has a 10% greater impact than someone at the 20% average.
Example: An O-rating of 110, with a 50% min% (roughly corresponding to 20 mpg) and poss% of 22% has an Impact Score of 5.5 (10 * 0.5 * 1.1.)
Obvious caveats include the fact that this is for offense only (O-rating, duh) and that this kind of thing only roughly corresponds to success in the upcoming season - returning players is only the start of any such analysis, and this wouldn't take into account newcomers at all. Nor does this measure pure, sweet, beautiful grit, lovingly mined from the Grit Mines of the soaring mountains of the Yukon and shipped farm-fresh to your TV, so you won't see Aaron Craft on any of these lists.
One benefit of this over just looking at O-ratings is that while O-ratings tend to have wild-looking outliers for players who don't play much, this softens that up quite a bit. Most O-ratings, at least in big-conference play, are between about 90 and 120, while benchwarmers might be anywhere from 40 to 150. But because they have miniscule minutes percentages and often very low possession percentages as well, their Impact Scores barely register.
All that said, here are some tables to make sense of the Big Ten.
First, last year's total IS's for each team:
This is a pretty good approximation of the order in which you'd find them in KenPom's team O-ratings. Even though Nebrasketball was a force last year, it wasn't exactly the case on offense. Northwestern's offense stunk. Penn State is a little high, mainly because they had a couple stars who played all the time and got little support.
Next is the teams ranked by the total scores of their returning players:
It's certainly possible, of course, to go up from year to year - all you have to do is lose a few players with O-ratings below 100. Sometimes that really does mean addition by subtraction. Sometimes it's sneakily indicative of losing a defensive stopper - this isn't a great example because I'm not accounting for season-ending injuries, but Leslee Smith had a 97 O-rating and thus a negative Impact Score, but was a defensive hoss for the Huskers last year. Nebraska would actually even be a little higher if I was counting players missing the season with injuries. But again - all offense here.
Next four charts will be player-focused. First, the top 20 players in the Big Ten last year, not counting anyone from the interlopers:
|Gary Harris||Michigan State||13.48|
|Roy Devyn Marble||Iowa||12.98|
|Branden Dawson||Michigan State||11.66|
|Glenn Robinson III||Michigan||11.63|
|Adreian Payne||Michigan State||10.41|
|D.J. Newbill||Penn State||9.51|
|Tim Frazier||Penn State||8.79|
|Lenzelle Smith||Ohio State||8.77|
J-Mo is a perfect example of how this is supposed to work. Morgan was an incredibly efficient player with an O-rating of 128.2 - up in the stars, higher even than Stauskas. But with relatively low usage, his actual impact is measured better here. Still high, but not, like, the Big Ten's most elitest player. Iowa's Marble is another good one - his O-rating is good, not great, but wicked high usage means that Iowa will miss him more than, say, Illinois will miss Jon Ekey.
Next, last year's "worst" 20 players - or at least, those with the lowest Impact Scores, meaning they were both inefficient and highly used.
|John Johnson||Penn State||-3.19|
|Trey McDonald||Ohio State||-2.13|
|Amedeo Della Valle||Ohio State||-1.88|
Many of these guys are simply freshmen, fortunately lightly-used, and part of the future plans of their teams. Some are not actually bad players, just sort of stiffs on offense. But, some were truly awful last year. Jaylon Tate took 23 three-pointers and hit exactly one. Bad three-point shooting is common throughout this group. Northwestern's piteousness on offense shines throughout this chart. And the big winner here, Tai Webster, literally turned the ball over more than he scored (58 TOs, 34 buckets), couldn't shoot worth a damn from any range, and played almost 23 mpg.
Now for the conference's best returning players. Interloping teams are now included.
|Branden Dawson||Michigan State||11.66|
|D.J. Newbill||Penn State||9.51|
|Travis Trice||Michigan State||6.97|
With three of the top five returning players, it's easy to see why Wisconsin is such a prohibitive favorite.
Finally, the 20 worst returning players. Most of the usefulness of the list really is in the top ten - once you get down to #15 or so, any negative impact is pretty benign, not a hell of a lot below average, lightly-used, or worth a damn on defense - but I figured I wouldn't just change up the number at the very end.
|John Johnson||Penn State||-3.19|
|Trey McDonald||Ohio State||-2.13|
|Alvin Ellis||Michigan State||-1.35|
As always, Northwestern reigns supreme.
Clearly, this does nothing to take into account freshmen, transfers (in, that is) and coaching changes, individual improvements from year to year, as well as the chemistry aspect of putting together a basketball team. But I think it makes a great starting point for analyzing a conference (which naturally is why I waited until after the season started and the preseason stuff is complete.) Also, it's helpful, I think, to see who's missing from the top 20 - it helps identify volume scorers and just plain overrated players - possible examples being Terran Petteway and Aaron Craft. And this kind of thing might be useful at midpoints of the season as well, though there's been less time to tamp down the spikes and valleys inherent in the O-ratings.
In a universe where Bill Murray actually gets the gopher (source).
1. R. Pitino def. R. Pitino
The best game of college basketball’s opening weekend – one of those now-customary military appreciation contests in a weird venue (a hangar in Puerto Rico this time) – featured Rick Pitino’s Louisville Cardinals and his son, Richard’s Minnesota Golden Gophers. It was a predictably sloppy game, partially because of the heat that required several stoppages in play to wipe perspiration off of the court, partially because it was the season-opener for both teams.
In any case, it was an ugly game. There were a combined 60 fouls in the game, the two teams combined for more made free throws (49) than made field goals (45), and half-court offense for both teams looked like far more difficult than it should have. All of these early-season gimmick locales effectively dilute the quality of play because of the unusual environments and this game fell victim to that also. It wasn’t competitive either: Minnesota held an early 14-10 lead, but Louisville ripped off a 19-2 run. The Gophers didn’t get the deficit to two possessions for the rest of the game.
In the end, Louisville won because Rick was playing with a stacked deck: future lottery pick Montrezl Harrell was outstanding – efficient, with 30 points on 17 shot equivalents – and showcased not only his aggressive and physical interior play, but a good-looking jump shot as well. Minnesota’s backcourt of Deandre Mathieu and Andre Hollins is great in its own right, but they were simply outclassed by Chris Jones and Terry Rozier. Few teams have a guard tandem that can defend like those two and even fewer have a big man who’s comparable to Montrezl Harrell.
2. #lukewarm #gopher #takes
Not a whole lot can be drawn from a game like this – Louisville might be the best team Minnesota faces all year and the game was in a structure built to store helicopters in the Caribbean.
Still, a few observations:
- Mo Walker > Elliott Eliason. Minnesota’s senior center tandem split the minutes evenly, but Walker got the start and outplayed Eliason against the Cardinals. Louisville didn’t bother to double Walker on post-ups, and he took advantage – he also ran the floor well in transition and had a nice drive-and-dish from the perimeter.
- The four spot was a disaster. Understandable, when guarding Harrell, but the Gopher four men struggled on offense (against a zone) as well. Joey King should hold down the position but he didn’t play well and Bakary Konate, Charles Buggs, and Josh Martin didn’t impress in relief.
- Nate Mason should contribute this year. Minnesota’s freshman guard had an unenviable opponent in his first college game, but he held up decently well: 10 points (but four missed free throws), 5 rebounds, 2 assists (and 2 turnovers). Pitino should just rotate Mathieu, Hollins, and Mason at the two guard spots.
- Auto-benching Mathieu was stupid. When Deandre Mathieu picked up his second foul at 9:21 remaining in the first half, Louisville led by just three points – when he got back after halftime, the lead was thirteen. Mathieu didn’t play particularly well but he had just 3 fouls per 40 minutes last season and didn’t foul out anyways. Don’t automatically bench guys.
- Other than that: Andre Hollins looked good and Carlos Morris didn’t; The hyper-aggressive Louisville trapping zones and half-court 2-3 zone gave Minnesota a ton of problems (and it’s easy to foresee the younger Pitino molding the Gophers into a defense like that in time); Minnesota’s jerseys were awful.
[After THE JUMP: In the Navy, where something something weird guys.]
This is going to be a short diary, with more a focus on college football generally than Michigan football’s last two weeks of the regular season. I say this every time, but I mean it. Also, please make sure to read my comment regarding Frank Clark and what has come out in the past couple of hours.
Best: A Michigan-free Football Saturday
Virtually everyone here is a diehard Michigan fan; I kinda doubt you’d be posting and reading this blog if you weren’t. But it should come as no surprise to many that watching this version of Michigan football the past couple of years is not an overly-entertaining affair.
So I’ve come to enjoy watching football on bye weeks because I can just enjoy football for what it is, with no real rooting interest or aneurysm-inducing reactions to what’s going on during the game. I can watch Miss. St. vs. Alabama, OSU vs. Minny, or FSU vs. Miami and just enjoy watching other teams without having to be overly critical of What It Means or The Future of the guys on the sidelines. Plus, I got to see Notre Dame blow a game to Northwestern (!) and Auburn get waxed by Georgia and get a bit of revenge for that insanity last year.
Best: Seriously, Irish?
You get down with your bad self, Pat!
Nothing is more fun than seeing ND lose due to their own hubris, and a piece of me is happy to one-half of M00N get a little bit of a gift.
Worst: Smart football?
So this conference really does suck at football.
Okay, take a step back.
This conference sucks at smart football compared to the rest of the country.
Minnesota and OSU was amazing in so many ways, from the two turnovers by Jalin Marshall (one on a probable score and another on a muffed punt that set up a Minny TD), to the worst onside kick I’ve seen in a long time (the Minnesota kicker both kicked it softly AND directly at the OSU hands guy), to Mitch Leidner going 7/19 for 85 yards and 2 INTs, 36 of those yards on one completion to Maxx Williams. But at least those are football-type events; Cole Stoudt completed 3 passes to Clemson players and 3 to Georgia Tech DBs, and two of these were returned for TDs.
No, what was amazing was Jerry Kill deciding late in the 4th quarter, deep in OSU territory and on 2nd down, to kick a 34-yard-field goal while being down 10.
Obviously, you need the FG and the TD, and on 4th down I absolutely understand taking the points versus going for a TD. But he had a bit over a minute left, and while he had no timeouts (which he used to slow down OSU’s previous drive), there was little danger in at least trying to throw to the endzone once or twice. You need that TD at some point, and passing from the 17 at least gives you some options; trying to throw from the 40 (or wherever they’d be on a successful onside) with under a minute wasn’t going to be any easier. And while Minny’s kicker had missed earlier, it was a 50-yarder doinked off an upright with more than enough distance. Instead, the Gophers ran up, spiked the ball, then trotted out the FG kicker for a chipshot and then failed to recover the onside kick. It was baffling enough for the announcers to question it as it occurred, and is even more inexplicable in retrospect.
I hate picking on Kill because he is a good coach who has helped Minnesota recover from the dregs to be a decent mid-level conference program, and his medical condition is obviously quite draining and it is a testament to him that he is performing as well as he is. That said, this was a crazy end to a game that probably shouldn’t have been that close anyway.
Best: Running Back ConfUrence?
I know PSU is known as Linebacker U (yes, and for other things). But for all the crap that the conference gets for being less innovative than others, but what it does have is the most complete and impressive collection of running backs in the country, with 4 in the top 10 nationally, including the top 2 spots.
Melvin Gordon broke LaDainian Tomlinson’s FBS record for rushing yards with 408 yards on 25 carries (!), which I’m not sure I could pull off playing NCAA Football 2014 on easy mode against Southern Directional State U. And Nebraska is no slouch defensively; they were comfortably in the top-25 defenses by FEI coming into the game, and were only giving up 3.7 yards per carry. By comparison, Tomlinson put up his yardage against Texas-El Paso on 43 carries(!). Wiscy has had some great backs in its history, but Gordon seems otherworldy right now, and for once I’m happy that the weird scheduling quirks at the conference kept him off the schedule.
On the other side of the field, Abdullah struggled running the ball, especially once Wisconsin started to pull away, but any other year he’d be in the conversation for best back in the conference despite being injured the past couple of weeks. Befitting the tire fire that has been Indiana this year in conference, Tevin Coleman put up over 300 yards running the ball and still finished over 100 yards behind the day’s leader while IU lost by 22. Similarly, David Cobb has been a major force behind Minnesota’s renaissance this year, and yet I’m not sure most people realize he’s the eighth-leading rusher in the country. And you have guys like Langford at MSU and Elliott and OSU putting up really good numbers as well.Gordon and Coleman could return next year, though I wouldn’t hold me breath on either.
As a Michigan fan it is obviously distressing to see so much talent at the running back position seemingly everywhere else except in Ann Arbor. I don’t know much about Fred Jackson and his coaching style, but it is clear that there is talent in the region and Michigan is either (a) whiffing horribly, or (b) failing to develop it into competent on-field performers. Still, it’s been fun watching these guys play so well, and I’m going to try to enjoy it because I’m not sure we’ll see this concentration of talent anytime soon.
Worst: M.C. SECher
Michigan was never going to sniff the college playoff this year, and as soon as Oregon trounced MSU and OSU got blitzed by VT nobody in the conference stood much of a change of competing for a spot; I guess OSU has a slight chance if they win out and a bunch of teams ahead of them lose, but that seems unlikely. Not only is my team not involved in this first year, but the two teams that are my mortal enemies are most-likely out, so hate-watching isn’t on the table either. So instead, I’ve been watching the past couple of weeks play out nationally with about as much detachment as one person can have for the sport.
So this weekend #4 Alabama beat #1 Mississippi State at home, #15 Georgia blitzed #9 Auburn at home by 27, #24 Texas A&M lost at home to a Mizzou team that lost to Indiana at home and probably won’t finish the year with a win against a ranked team, and #17 LSU was shut out by 5-5 Arkansas a week after nearly upsetting now-#1 Alabama. This came a week after A&M upset Auburn on the road.
Now, on the surface these results should be exciting; highly-ranked teams playing each other in important games are what college football is supposed to be all about. Well, that and money. Lots and lots of money. Oh yeah, and maybe education.
But as has been the case all year, it feels like a shell game is going on in the SEC and the media (read: ESPN) is all too happy to keep it going for the watching public. This year has seen an unprecedented number of SEC teams ranked, and in a down year for college football some of the best teams do reside in the Southeast. But these rankings have become more and more dubious as the year has gone on and we realize that teams like LSU, Auburn, and A&M aren’t sleeping giants but instead deeply-flawed outfits with inconsistent performances.
The problem is that because the perception remains that the SEC is the dominant football conference by leaps and bounds, you have this weird feedback loop wherein everyone has “good” wins and “good” losses because everyone is ranked, even though objectively they probably shouldn’t be. I know the fashionable term is SEC bias, but it isn’t even that; it’s just laziness and an inability to look beyond the numbers next to the logo.
LSU has two wins against ranked teams on their schedule – 28-24 against Wisconsin to open the year wherein the Badgers forgot to give the ball to Melvin Gordon at halftime and let LSU score 21 points, and a 10-7 fugly win against Ole Miss that had “drama” the same way a couple of drunk housewives throwing crap at each other on a Bravo show have “character development.” And yet, they were #17 in the country going into the weekend because they had losses to highly-ranked Auburn, Mississippi State, and Alabama, who themselves benefited from “good” wins against the usually-ranked Tigers.
Miss. State was #1 in the country before this weekend because they beat 4-loss LSU, 3-loss Auburn, and 4-loss A&M. Since then, they struggled to beat Arkansas and had a meh performance against Kentucky, but are still in the playoff hunt and will basically play an elimination game against Ole Miss to end the year, a team that had an okay win against Boise St. to open the year (don’t be fooled by that 35-13 score; it was an ugly game until well into the 4th quarter), a legit good win against Alabama, a victory over that same A&M team, and “good” losses to LSU and Auburn.
All of these teams are basically trading off wins; you go up and your opponent goes down, and next week you’ll trade places again. People call it “competitive balance” and that has some merit, but at some point balance becomes mediocrity, and I’m starting to think that is the case for the SEC this year. There are no dominant teams in the country; FSU is a paper tiger at #1 given how they’ve looked for weeks now, and the Pac-12 and Big-12 keeping get in each other’s way. So I get that the SEC feels like a safe bet because of their recent dominance. But actually watching these games, it feels like this year is the worst type of one for the new playoff format. Nobody really deserves to be #3 or #4, and if this was a skins game in golf you’d almost rather they roll over the 2 spots to next year when, maybe, there would be more deserving candidates. But what I do know is that barring a major immolation by a couple of teams, the SEC will probably get a 2-loss team into the playoff because of musical chairs in the polls, not because of any amazing evidence on the field.
Worst: F*cking Ohio
[Note: This report about Clark’s situation came out as I wrote this entry. Based on the details included therein, this is a bad situation for all involved and, if true, should lead to Clark’s immediate dismissal from the team. I’m leaving in my thoughts here because I feel the themes remain true.]
So I don’t know what happened to Frank Clark this weekend; seems like nobody does beyond the basics. He was arrested in Ohio for a misdemeanor domestic violence offense. It’s the most serious misdemeanor offense charge, but the heaviest penalty is up to 6 months in jail and up to $1,000 in fines. I’m still a registered attorney in Michigan as well New York state, but I never practiced in this area of law and have no experience in Ohio, so I’m not able to comment on the overall seriousness of the charges. But what I do know is that Frank Clark did something significant enough that police were called, he was arrested and charged, and (to my knowledge) is awaiting a bail hearing/bond posting. I’m sure there are threads out there in the ether where people claim to know more, but until I see something more credible than boojack321_UMSUX, I’m not going to take those claims with much salt.
It doesn’t surprise me that people have overreacted and started to concoct their own narratives regarding what happened this weekend; this is of course the internet, and the Michigan fanbase is as neurotic and reactionary as any others. I’ve read people calling for restraint until we hear more, others claim this is yet more proof that Hoke has lost this team, and (sadly) a number of people denigrating Frank Clark and distancing themselves from his association with Michigan. I remain steadfast in the insane belief that we don’t know enough to pass any judgment one way or another, but I’m also a philistine.
In no way should this be taken as some greater indictment of Hoke or the program; Frank Clark's actions on a bye week, miles away from campus, are his own alone. This is not Hoke's fault, nor the school's, nor anyone else's; these are the actions, whatever they are, of a single person, and in no way should be try to read more into them than that.
By this time tomorrow we should know much more about what happened and proper Internet Justice will be handed down, but it is safe to assume that Clark won’t be suiting up this weekend barring an extremely rare reversal of the legal process, and I’m guessing his final game may well have been against Northwestern. If so, it will be a sad coda to a career at Michigan that has been marked with some ignominy (the laptop incident) as well as flashes of brilliance on the field.
Obviously I hope justice is served, and that once all of the facts come out the proper resolution is achieved. That said, and I know I’m just spitting in the wind by this point, but I hope the UM community pumps the brakes a bit on the crucifixion until we know more.
So Michigan beat Hillsdale comfortably after some early-game jitters. The offense looks pretty good, and the defense has its issues with so much inexperience in the front court but it can absolutely turn it on in spurts.
I thought Walton looked great running the offense, and Caris messed around and nearly got a triple-double for his efforts. The freshmen looked like freshmen, but you can see the potential there. In particular, I thought Dawkins looked really explosive out there in his limited minutes, and should be a nice second-team scoring and energy option.
It’s early, but it was nice to watch a Michigan team take care of business and look fun and interesting while they did so.
Best: Two More Weeks!
Like most people, I’m ready for this season to end. The Clark situation is one last kick in the balls, but it should be over soon. As I said last week, I’m not excited about UM to make a bowl game if they do, but whatever, I’ll still be hoping they pull off the win against Maryland and get to 6 wins. But this year could not end any sooner.
1. The Four Factors
|Expected Pts||Conversion Rate||Bonus Yds/Play||Red Zone|
*TD’s are treated as 7 points regardless of PAT outcome
So Michigan had the field position for 35 points…and scored 10. Only 11 times this season has an offense left more than 25 points on the table in a game. With over 6 touchdowns worth of points left on the field, #M00N was the third worst game of the year in terms of points scored versus field position. Clemson-Louisville was slightly ahead and Saturday’s lead-in game of Penn State versus Indiana was the worst of the year at 43.5 points below expected.
|Expected Pts||Conversion Rate||Bonus Yds/Play||Red Zone|
|Offense||26.0 (78/11)||59% (108/12)||2.0 (121/13)||4.7 (91/10)|
|Defense||26.2 (45/5)||63% (18/6)||1.7 (13/3)||4.2 (10/1)|
(National Rank/B1G Rank)
Saturday’s game pushed Michigan to roughly neutral on field position for the year. The 2010 vs 2014 season mirror image continues in earnest. The offense can’t convert 60% of it’s first downs, is bottom ten in explosive plays is an absolutely anemic unit. The defense has struggled against the better teams, but is still top 20 in conversion rate and has done very well in limiting big plays and red zone scoring.
If you are wondering how Michigan isn’t last in Bonus Yards in the Big Ten, despite being in the bottom 8, we have our opponents from Saturday to thank. Northwestern is second worst nationally, behind only Wake Forest which is threatening to set records for their lack of explosiveness.
2. Individual Performances
Devin Gardner: –9.8, –26% on 28 plays
Drake Johnson: –0.8, –2% on 10 plays
De’Veon Smith: +1.2, +13% on 18 plays
Devin Funchess: –3.3, –7% on 7 plays
The –9.8 for Gardner is before the opponent adjustment pushes it down another point. Gardner’s 3 lowest rated games of his career are Northwestern, Utah and Michigan State from this season. It’s like he was on a Saw movie and said if you don’t want to be killed by your offensive line, you have to fight to survive by starting over in a new system in which you will fail miserably. I am thankful Devin has fought to survive, but I can’t believe what he’s become.
In terms of the running backs, I don’t know what to say at this point. The statement from Hoke saying he is going with a hot hand is just silly. Hot hand is a pretty controversial concept to prove over large stretches, to think that one RB is performing better than another based on a handful of carries that are largely dependent on the guys in front is just silliness. This football season, man.
3. Game Chart
6. +12.7% Michigan stuffs Northwestern on second and goal (late Q4)
5. +13.1% Northwestern’s 4th down pass falls incomplete (mid Q3)
4. +14.9% Jehu Chesson recovers the muffed punt (mid Q3)
3. –20.5% Raymon Taylor’s pass interference gives Northwestern a 1st and Goal (late Q4)
2. –21.0% Devin Gardner is intercepted (-5.2%) and returned 79 yards (-15.8%) (late Q3)
1. +40.8% Frank Clark Shuts down the two point conversion (late Q4)
Michigan’s best play on offense was the 16 yard completion to Jake Butt on the final field goal drive, worth about 5.5%. No rush or pass play made the top 6 for anything good that happened to either offense. Raymon Taylor PI was the closest thing.
4. Dumb Punt of the Week
Since I have been slacking the last few weeks we’re going to go back a bit to pick what may be the dumbest punt of all time, at least in the first quarter.
Arkansas has a hard luck season, they are in the loaded SEC West and haven’t won an SEC game in over two years. So far this season they lost then #6 Texas A&M in overtime and then #7 Alabama by 1 after a blocked PAT. So with a shot at the #1 team in the country in Mississippi State, they would pull out all the stops to get the win, right? It was early and Arkansas had a 3-0 lead and was driving. After getting the ball down to the MSU 32 yardline, the Razorbacks faced a 4th and 1. Obviously since this is DPotW, you know what happens next.
Arkansas would go on to lose 17-10; I’m sure they would love to have another possession with less than 40 yards to go to have a shot at OT, but that went out the door with the punt team. Since 2003, the only case I can find of a more egregious punt was in 2009 UConn punted on 4th and 1 from the Ohio 31.
Submissions are no closed, Dumb Punt of the Year voting is now closed.
As you may have heard, Michigan may be looking for a new head coach and there is message board rumblings that it could be Jim Harbaugh. Not that he needs much introduction but I did want to put together a couple of quick graphics on how his Stanford teams performed.
Team Performance: Stanford vs Michigan
Harbaugh years in yellow, y-axis is opponent adjusted points versus average.
You can’t get a much more dramatic change than Stanford experienced in the four years under Harbaugh. They went from the bottom of the Power 5 and nearly 9 points below average prior to him joining to over 15 points above average in his final season, with improvement demonstrated in each and every season. He even left the program in strong hands with Offensive Coordinator David Shaw taking the reigns and keeping results at a still high level. From an on-field perspective, it doesn’t get any better than this. Took a program not doing well, consistently improved them, built up a strong staff that could continue at a high level after he left, but still demonstrating that he added some value while on the sideline.
On a side note, that blue line should silence any remaining voices about the viability of Hoke’s retention, independent of season outcome. That is four consecutive years of production decline, going to take some pretty amazing performances for 2014 to even end close to 2013, which obviously wasn’t great.
Recruiting Performance: Stanford versus Michigan
Stanford is a tough place to recruit at a high level and Michigan has traditionally operated at a different tier than Stanford. Harbaugh wasn’t able to turn around recruiting momentum immediately the way a James Franklin type would, but you can see by the 2009-2010 results that once the wins started coming, Harbaugh was able to raise the Cardinal recruiting profile, which has remained high in his absence. Based on this limited evidence, it would appear that Harbaugh is likely not an elite recruiter, but one more than capable of recruiting at a high level.
Basketball season will be more fun than football!
I thought we might take a moment here to look at a few differentials that normally get tracked (well, I normally track them because I am odd like this) and how they trend with scoring margin in games. We did this last year, of course, and for someone reason I called it “point differential” because I was on a roll with the word “differential” and I grew too lazy to change it. Anyway, it’s scoring margin.
First, we’ll look at yards per play, that differential and scoring margin. You’ll be pleased to know that we do in fact, on average, outgain our opponents this year despite everything that you have seen and the ugliness of some of it. By how much? The average is all of 0.76 yards, or about 27 inches per game, so we have this going for us. The average scoring margin, which will be the same in all three charts you’ll see anyway, is 0.3 points. By the slimmest of average margins, we somehow find a way to barely win half the time in 2014, if you will.
So, with two games to go and every opportunity for what I just mentioned to be completely untrue by the end of the month, here’s where we stand with this data:
Much like last year, the worst individual performance for yards per play differentials was against Michigan State, except that this year’s number is worse than last year’s – it was -2.62 last year, and we slowed down by nearly a yard this year. Overall, you can see that the two sets of data have a decent relationship though.
Before I introduce third down differentials, let me say that a most #B1G thing happened to us on Saturday in that ever so hideous win. We were, in fact, 41.67% LESS successful than Northwestern on converting third downs, so this is a metric where your worst individual performance in the season to date can still work out for you somehow. The average differential for the season to date on third downs is 1.20%, so we’re barely winning this one typically as well.
Here’s how that and scoring margin look together:
The relationship isn’t quite as strong as yards per play perhaps, but it is there. If you want an R-squared value for this, it is actually 0.668 (sample size…grumble grumble, of course), which is not too bad for all of the confounding factors of a football game.
With first down differentials, we find something similar to third down. On average, we manage to get 52% of the total first downs in a game, so by the slimmest of margins again, we come out ahead here. I keep saying that in this diary – I should qualify it by saying also that there are perhaps 2-3 games which are responsible for these averages looking like they do, but this is normal in most seasons, it seems.
Anyway, here’s the first down differential charted against scoring margin:
For kicks, the R-squared on this is 0.653, so as I said, very similar to third downs.
Alternative title: You want a 1 year wonder named Mullen over this guy?
It has been interesting to read all over MGo msg boards, and hear on the podcasts the conventional wisdom that Les Miles peaked a few years ago and has begun to regress significantly. Asking a coach to remain at 1-2 losses indefinitely is something only Bama fans can expect. And as sports lovers know the difference between a 9-3 year and a 11-1 year can often be 3 -5 plays.
So I thought I'd take a look at how Miles LSU teams offense + defense data stacked up the past 3 years (inclusive of 2014 which is far enough along to be useful) versus the first 7 years. The data will probably surprise those who have bought into the mythology of Les Miles dropping off of late.
Columns 3-5 are three measures of offense, of which I favor columns 4/5 for rigor (FEI began in 2007). Same for columns 7-9 for defense, with emphasis on colums 8-9.
|W/L||Tot Off||oFEI||oS&P+||Tot Def||dFEI||dS&P+|
|First 7 Years||62.3||28.4||19.6||11.6||16.4||12.3|
|Last 3 Years||62.3||29.7||21.7||12.0||18.7||15.0|
Miles' teams are remarkably consistent. All 6 offense/defensive measures show a variance of 0-3 in the past 3 years versus the first 7 years. That's a rounding error in the big picture of 128 slots.
Now of course in the end the W/L record matters. But if you analyze a coach only on that you will miss the forest for the trees.
- Miles 2012 team had 3 losses - by 8 to FL, by 4 to Bama, by 1 to Clemson. They also had a handful of close wins. LSU could have been anywhere from a 12-1 to 8-5 based on a few bounces.
- Miles 2013 team had 3 losses - by 3 to GA, by 3 to Ole Miss, by many to Bama. They only had 1 "close" win. LSU could have been a 12-1 team with 1 blowout loss to Bama very easily.
- The 2014 team has an offense which would make OSU's and UM's look like graybeards - essentially every skill player went to the NFL post 2013. Even with all thoses losses, if they played in the Big 10 (with weaker competition) their offensive FEI would be 3rd best in the confernce. There has been 1 blowout loss to Auburn and 2 close losses to the #1 and #4 teams in the country.
Miles has never had an elite QB other than Russell who he inherited from Saban. He has not really had elite RBs either - usually that offense is a manball style with good OLs and aNFL type WR here and there (who lack QBs to get them the ball regularly). And it's paired with a normally elite defense. Miles built at LSU what Brady promised in Ann Arbor.
Now if you want to argue Les cannot repeat that record here (where the conference is worse and he will be LOLing at 8 of the 14 Big 10 teams - "wait the Big 10 has 8 Vanderbilts?"), or he relies on JUCOs too much (as does Mullen), or he can't speak well enough in public forums, or he has no friends in the UM AD - well that's another conversation to have.
But the mythology of LSU regression of late is not born out by the data. LSU has essentially had the same team the past 3 years as it had the previous 7 on average - it has just picked up an extra close loss or two here or there vs elite years. That's football.
And I can make an argument that Miles today carries a lot less risk than Miles of 2008-2009 when it looked like he carried Saban's players to success and then once Miles' recruits filtered in things fell off. Of course 2010-2013 proved that not to be accurate but post 2009 you did not know that.
If the Harbaugh Hail Mary fails, it would be silly to not put Miles in the next tier of candidates. Miles has a better situation than Mullen at LSU but has a decade worth of great to elite seasons (vs 1!), runs a style of offense suited for the players recruited at UM (vs a run spread Mullen brings of which we are eliminating the last few players recruited for it this year), loves the damn school, and the "success due to JUCO" argument is the same for both. And if interested UM's average recruiting ranking during the Miles era at LSU is #11; LSU's is #9.
Hard to understand any logic calling for Mullen over Miles unless age trumps 15 other coaching measures Miles owns over Mullen. And 5-7 years of a proven commodity is better than a 1 hit wonder who if a failure won't be here 5-7 years himself.
Technical notes - I am using 3 measures.
(1) The NCAA's total offense + defense - which is a very simplistic measure that ranks teams on nothing other than total yards gained or allowed per game. This benefits teams in bad conferences over teams in good conferences as any halfway decent unit will put up PED laden performance. To that end if you focus solely on total defense, the Big 10 is in a golden era. In year 2014, 4 of the top 10 defenses are from th Big 10 (Wiscy #1, UM #7). And 8 of the conferences 14 teams have defenses ranked in the top 25 in the entire nation. Of course that reflects on the horrid state of offenses in this era led by a bunch of bad QBs. Plus a lot of MAC games in OOC. So I don't love this measure that much.
(2) Football Outsiders FEI measure and (3) Football Outsiders S&P+ measure. Both of these incorporate various strength of schedule adjustments, garbage time adjustments, blowout adjustments etc to give a more comprehensive view of a unit other than yards gained or yards given up. Read more about them here and here.