trying to delete...
Considering Michigan State
Why? We're 3-9 and have no bowl game, that's why. But also because I just found the stats interesting.
Michigan State played Georgia in the Citrus Bowl, prompting dual Georgia/Michigan fan Michael at Braves & Birds to ponder World War II (most things cause him to ponder World War II):
Michigan State came into the Citrus Bowl (I refuse to use the new name) in a position not unlike the one that Japan in 1940-41. MSU was opposed by a slumbering giant, an opponent that was complacent and unprepared for war at the outset, but an opponent with far greater talent and capacity for a long-term fight.
Long story short: the giant woke up in the second half and dropped bombs. Only Georgia's disinterest in the game kept Michigan State's 2008 from being a mirror image of Notre Dame's 2006 minus the undeserved BCS bid, a superficially pretty record marred by epic beatings at the hands of the few top ten teams on the schedule.
Here's a table.
|Team||Total Offense||Total Defense||Margin||Total Offense||Total Defense||Margin|
|Team A||352.4 (67th)||357.9 (40th)||-5.5||339.5 (8th)||375.4 (8th)||-35.9|
|Team B||416.8 (42nd)||345.5 (32nd)||71.3||431.1 (1st)||394.1 (7th)||37|
Which team went 9-4 and which team went 7-6? It's obvious. If team B went 9-4 I wouldn't be posting about it. Michigan State's offense fell off a cliff and the defense basically stayed level and the end result was a significant improvement. As a bonus, the real numbers are actually worse than what you see above since they don't include the Citrus loss in which State was outgained by about 100 yards.
Anyone who's read this blog for more than a couple months knows the a-ha moment that's coming up: turnover margin! Except the 7-6 team was slightly better than the 9-4 team that critical, near-random category. No dice there. Neither can the explanation be found on special teams. Punting was about a yard better this year, and while the punt returns got a lot better the kick returns got a lot worse. There's nothing in the stats that offers an easy explanation as to why Michigan State seemingly got much worse but won more games.
The nearest thing to an explanation I can come up with is the distribution of turnovers. MSU was in -7 in turnover margin against Ohio State and Penn State, both epic losses; in all other games they were +9. Since a large chunk of that yardage gap also came in those aforementioned epic losses, State played a large number of games in which they were on the whole equal with their opponents and won the turnover battle and therefore the game. State won three games in which they were outgained, sometimes badly:
|Opponent||Yards For||Yards Allowed||Margin||TO Margin||Final Score|
State had no games that went the other way; they had their crappy days against teams they were highly unlikely to beat anyway. The Spartans were a 6-6 or 7-5 team—again—that had the breaks fall in the right way for them to leap up a couple spots in the weakest Big Ten in memory.
This is the long way of saying I'm not particularly afeared of Dantonio. While he seems like a better coach than the last two jokers at State (an honor also shared by Clay Aiken and jars of peanuts), Braves and Birds nails his ceiling:
Mark Dantonio is Jim Tressel without the talent base. Exhibit A: punting in the first quarter on 4th and 1 from the Georgia 39. Exhibit B: an offense built around running the same guy over and over between the tackles. (At least Tressel came out of the dark ages with Troy Smith.) Exhibit C: a kicker who attempted 25 field goals this year. Exhibit D: an on-field persona that makes Ben Stein's character in Ferris Bueller's Day Off look like Sam Kinison. With the way Dantonio's team approaches offense, I'm constantly reminded of the Japanese officer who said in 1944 that Japan didn't need radar because its soldiers could see perfectly well.
World War II analogies are appropriate because the last time Dantonio updated his thinking was during the Battle of Midway. At best he makes Michigan State into a Wisconsin or Iowa level program, and even that seems pretty doubtful.
on fourth and 1 in Georgia territory in the first quarter told me all I needed to know about the game: Same Old Spartans. What do you have to lose at that point?? The odds tell you to go, common sense tells you to go, your sense of adventure tells you to go.
Brian Hoyer sucks.
This was the thought that kept running through my head as he overthrew open receivers and panicked at every hint of pressure. He's just terrible.
Dear Big 10 Fans:
Welcome to the "OMG WE SUCK WE HAVE TO REBUILD!!11!!!" party. We've made some bean dip, and we humbly suggest leaving your defensive coordinators at home.
Good stuff. I can't wait for Dantonio to point out to the media that MSU once led this game. "It was 6-3! Look at Georgia's sideline!"
I need to thank Brian for poiting this out all year.. 1st before the Mich game as it was eyepopping how bad the Sparty run game was considering the Ringer hype..this has lead to nice money collections betting against the fraudelent Sparties...that being said I think we need to give Dantonio some slack...he has upgraded the recruiting and he seems to have his team playing better fundamentally on defense...if he would game plan for all his opponents like he does for Michigan he may have a better chance...the Tressel comparison with less talent seems perfect. We need to give him time to see what that talent will get him.
The thing that cracks me up is the Dantonio is a saint stuff the Sparties like Valenti keep spouting while out of the other side of their mouth they call Tressel dirty...then he says Mayhew is worthless because he worked for Millen...can't have it both ways.
I disagree about turnovers. Good teams force them and take advantage of them. It is a coached advantage. "Random" TOs like unforced fumbles can be limited by coaching. Or being Mike Hart.
Turnovers are inherently random. No team proved it more than another Big Ten fluke, Minnesota. Good teams prevent offenses from moving forward - if the offense coughs it up along the way, gravy. But if you get routinely gashed like MSU and Minny and eventually end up watching that team cough it up randomly to you - you are lucky. That's it.
Luck is when preparation meets opportunity.
Turnovers are highly variable, yes. But not random.
Oklahoma (2008) - #1
Florida (2008) - #2
LSU (2007) - #2
Michigan (2006) - #4
These are not "just lucky" teams. Feel free to disagree.
That's going back three years focusing on national championship teams and our own maize and blue. Far more thorough than cherry-picking two "random" Big Ten teams.
Feel free to add some analysis of your own, instead of the internet equivalent of rolling your eyes.
how much of the TO issue can be traced to coaching or offensive scheme?
along the same lines, i also am convinced that this team, with the same personnel, wins more than 3 under lloyd. certainly not 8-9, but i think LC would have won 5-6 (maaaaaaybe 7, probably not) b/c the D would have been better (not saying a lot) and the ST wouldn't have fumbled 73 kicks. LC would have put fewer FR to return kicks, i think... and we'd have bitched & moaned about how he refused to put young guys with homerun potential back deep.... but they would have had fewer TOs... leading to another win or 3.
for the sake of argument, anyway.
"Am I disappointed? Hell no. We're 9-4! I'd be disappointed if we were 3-9."
When reached afterwards, he said he "wasn't referring to any team in particular, especially Michigan."
He's gonna make it fun.
TOs are random if you consistently have true FR returning kicks, running and throwing the ball instead of SRs.
so maybe it's not a direct coaching/scheme thing and a personnel thing?
either way, it would take a statistically rigorous analysis: touches/fumble (not necessarily lost) as a fcn of either:
- year out of HS
- games started
- games played in.
good luck with that, mgobloggers.
I think that "turnovers are random" are sort of a bastardization of what the numbers say. Obviously, skill plays a part in TO's - some RB's don't fumble, some QB's do a great job of protecting the ball, etc.
What is numerically true is that there is virtually no correlation in Turnover margin from year to year. So, a team that excels in the stat one year would be likely to gravitate towards 0. Similarly, a team that is poor in margin would improve towards 0.
anout 0 correlation from year to year.
y'all are gonna get tired of me saying this, but teams with DD -to margin almost always improve their record the next year.
becasue its next to impossible to be as bad the following year in turnovers.
Conversely, the teams with DD +to ratio usually regress, although I dont have as much, or concrete, evidence as I do with the other example.
I dont believe they are totally random either. Though I have no idea how one would go about explaining it.
Who was Team B?
a reasonable job of rebuilding a program that has been mostly dormant since Duffy. He may not be all things to all great minds, but he has the program headed in the right direction. He has gotten the most out of several players-like Jones the All Big 10 LB. He is recruiting well if not a Weis level.
I think it's best for UM fans to be'um damn humble until we have something to indicate we are not headed toward the abyss that MSU seems to be leaving. UM was beyond anyone's wildest guess of how horrible a transition can be. No one is suggesting that '09 is going to be pretty for UM. Yes, MSU can be laughed at for many reasons, but right now no one is causing its foes more glee than the humbled wolverines of AA.
You are a wise (blue)man.
the Domer Homer (go away).
If turnovers are inherently random (which I will stipulate simply because people on the internets have beaten me to death over it), then why are some teams consistently good in the +/- category, year over year? Example: for the last 5+ years, USC has been ranked in the top 20 in +/- category.
That's a team that is always good and has never "regressed to the mean" in this statistical category. And I imagine there are other examples, maybe not of teams consistently top 20 in +/-, but maybe they stay in the top half and never sink to the bottom. Am I to believe that this is simply a matter of the coins flipping the "right" way for USC and others the past 5+ years?
I understand the desire to paint turnovers as inherently random, because it enables us fans to not get worked up about a statistic that provides little (if any) information about the quality of a team. But I still have trouble simultaneously believing that turnovers are random, while observing that some teams have managed to beat this alleged "randomness" and consistently be good at it.
IMO, "random" is too strong a word to use when describing turnovers. Perhaps "externality" would better suit this category.
one team to refute a general rule? no one would disagree that mike hart had a talent for not turning the ball over. it's that in aggregate considering all teams, the correlation is nonexistent.
usc as a typical example of anything
It is generally true that younger offenses (especially young QBs) will commit more turnovers than more experienced ones. Veteran QBs are better at making reads, and veteran RBs are better at taking care of the ball. MSU had seniors at QB and RB, so it's not surprising that their offense didn't turn it over much. (Notice, though, that their backup RB, who rarely sees the field, lost a fumble against us.) It's also not that surprising that our extremely young offense committed an unusually high number of turnovers.
But on the whole, turnover margin for most teams is highly variable from year to year. I would not expect Minnesota to have that crazy margin in their favor next season, nor would I expect MSU to be able to win a bunch of games next year in which they are outgained by 100+ yards.
... Dantonio always sounds like a "turd in the punchbowl" emcee at a cheap-ass charity banquet.
Take this from the Freep:
“It was a big game for us,” Dantonio said. “We played competitive, and that was the first thing we had to do. After the Ohio State game and the Penn State game, we needed to compete at this level against this type of team. I think we did that today.”
Sort of like celebrating the new heights of low expectations. How about "we got beat by a better team" or "we're proud of playing on Jan 1, but we still have a lot of work to do."
They weren't even close to Georgia when the Bulldogs got rolling. Ringer with < 50 yards? That's a reality bitch-slap.
UM also has a lot of work to do but I think RichRod will get Dantonio's number quickly and I'm looking forward to his agony at the post-game pressers.
I think you can explain the turnover differntial thustly..they are random for 2 teams at equal levels..then they shift slightly away from randomness due to extreme disparities in talent(USC) freakishly good RB's at not fumbling(Hart, LSU RB's), extremely accurate experienced qb's(Bradford, Tebow), extreme coaching(1 way or the other), extreme youth at skill positions(Mich) extreme experience at skill positions(Oklahoma) you get better or worse then just plain dumb luck, but if you want to line up MSU vs Maryland and say MSU was +15 this year Dantonio can coach the piss out of turnovers or Maryland was -15 and say Friedgen should be fired he coaches to give up turnovers that makes no sense.
If I understand you correctly, then you're saying on the extreme ends, turnovers can be influenced by controllable factors (like coaching), but in the middle they are random. Ok, fine, that makes sense.
This kind of bolsters my point, though, that the "turnovers are random" meme is a somewhat inaccurate and insufficient explanation for what contributes to a turnover. There is a lot more going on there than just random acts of God, IMO.
*Turnovers are random for 2 similarly ranked (Offensive/Defensive/Special Teams) teams. W-L for two very similar teams (The MSU/Maryland example) may very well depend on the randomness of that distribution. The "Turnovers are random" meme applies to teams that overachieve based on their turnovers, or underachieve because they keep giving the ball away. You expect EMU or Army to almost always have a bad turnover differential because they aren't (usually) very good, just like you would expect USC to almost always be in the top 25 in positive differential.
Basically, you can use turnover differential to compare Apples to Apples (Oklahoma to Florida) but not to Oranges (Michigan)
and since it relies on extremes to overturn it, we should expect that to gravitate toward zero the following season.
i think that's a function of the inherent turnover of the college game. teams that turn it over a ton one year are usually sooooo young that the next year they return everyone and hence turn it over less.
the inverse is true for experienced teams in year one becoming inexperienced in year 2.
that and lightning never tends to strike twice.
I dont know which comment to "respond" to, so here's my take on the "randomness" of TOs as it pertain to your USC example.
I think good coaching with good players can create a situation where a team, year in and year out can be among the best in getting takeaways and in overall TO margin.
My focus on TOs has always been how do the bad teams respond. Those teams almost always bounce back becuase ball security and smarter throwing decisions become a major off season focus.
Sounds like that is coaching. To me, thats not random.
I think calling them random is a bit of a misnomer. What they are, are weird occurences in games that often--but not always--create misleading scores as it pertains to the statistic box score. People have equated that to being random, which is fine, but perhaps not a perfect definition for this specific usage.
Interceptions are not as random as fumbles. Interceptions are more of a function of the QB, OL, and WRs. However, look at our game vs Minnesota this year to realize how many bad throws should have been picked off.
at best, MSU will probably be better than wisco or iowa because they will be able to recruit better. look at msu's class for this year, i doubt wisconsin or iowa has landed a class like that since recruiting sites started
in 2005, iirc.
Okay, let me comment on a couple of things:
--- Turnovers are not entirely random, but they probably don't directly correlate to talent, experience, or coaching. When the ball pops out, it is luck whether that was a lost fumble or whether the team retained possession. When a QB throws a lame duck that a LB could have caught but didn't, that's luck and random in nature.
--- A better way to say it is that if you put two identical teams and played the same season over and over you would likely get a pretty large variation in the total lost turnovers each year, even if everything else was the same.
--- But most of you are correct, experienced players, Mike Hart, good QB/OL/WR, can help keep down turnovers.
For all of those reasons, I expect Minnesota to regress next year and Michigan to improve significantly. It isn't possible for either team to repeat their TO margin situation from 2008.
As for Mark Dantonio, the real subject of this post, I think he's the best coach MSU has had in a long time....even if that doesn't say much. The biggest area of improvement is in the discipline (penalties, being idiots) on the field and the confidence that the players have in themselves and him. He also seems to be a better (or more fortunate) recruiter.
At MSU, being the next Iowa or Wisconsin would be a big success. They'd take that, and it would shift some of the competitive balance on any given year within the conference.
But here's the catch...... If he turns them into a solid program and wins 9-10 games again next year or the next two years.....WILL HE STAY? Would he make a career at MSU or would he jump to another program? That's the real question MSU fans should be asking.
Seems awfully early to write off Dantonio based on two years. I don't expect him to get us to a national title game next year (or, really, within the next five years), but if the ceiling is 11 wins and a Rose Bowl (which may be a little optimistic, but not too much so), that's still a heck of a lot higher than it's been for a decade.
(And while I do have problems with some of the play-calls - the 4th and 5 fake punt, for instance, looked like a play designed to gain three yards in the most obvious fake-punt situation imaginable, and there's no excuse for the 4th and 1 punt - Dantonio's been a bit more aggressive on 4th down than most coaches.)
One last point: last year's MSU team had the opposite characteristics - lost close games, but dominated in their wins. If this year's team was, statistically, a 7-6 team that stole a couple of games they should not have, last year's was a 9-4 team that blew a couple of games they should not have. (In fact, they lost four games in which they outgained the opposition last year; only one game swung the other way.) Had the '07 team caught the kind of breaks this year's did, MSU might have been the surprise visitor to Pasadena last year instead of Illinois. (And yes, had the '08 team caught the kind of breaks - none - that the '07 team did, we might have been in no bowl instead of a New Year's game.)
So when it all averages out you're a 8-4 team in the Outback/Champs/Alamo bowl?
Admittedly its a step up from where you were the few years before Dantonio/where we are today.
Watched the entire MSU/GA game. Did anyone else notice the comments by both the onfield commentator and the guys in the booth about Dantonio's composure? They talked about it several times, like he was a ticking time bomb. I get the feeling he's kinda like the asshole boss that just yells a lot and gets production from fear. I agree with the other posters - RR will have him figured out in short order.
My take on turnovers, and Dantonio
On Turnovers - I thinks fumbles constitute the vast majority of the "randomness" in the "turnovers are random" meme. Sure, some WRs and RBs hold onto the ball better than others, and better OL protection tends to limit the number of fumbles by the QB, but overall the number of events that must go right for a fumble to occur are sufficiently unpredictable that you really can't predict how that translate from year to year.
As for interceptions, I think most people would agree that young/skittery quarterback+poor protection+falling behind = boatload of interceptions. The more of these characteristics found with your team, the higher likelihood of you having a bad turnover ratio. In particular, teams that play behind a great deal almost always have higher interception numbers than those more competitive teams. That's part of the reason why teams like USC tend to have great turnover margins - teams are always playing from behind, and as a result are forced to throw the ball more often, which in college inevitably leads to turnovers.
So I agree that, from year to year, it may be hard to predict those Minnesota-type turnover margins, I do think it is possible to forecast at least how many of those turnovers will come from the QB based on certain interception factors.
On Dantonio - I was at MSU during the JLS era, and I kid you not that a monkey with Turrettes would have been a better coach during his last year. So MD had nowhere else to go other than up. That said, he at least looks like a serviceable Big 10 coach, and with MSU's ability to recruit decent players, I fully expect them to be at least as good as Iowa/Wisconsin year-in and year-out, if not a solid #4 in the conference. His main problem is one that haunts every MSU coach - UM is just a bigger name both nationally and in the state, and thus it tends to get the best recruits and will consistently field a more talented team. Sure, there are years when MSU will beat UM and challenge for the Big Ten crown, but it will almost always be viewed as the second-best team in the state, and that outlook tends to serve as a self-fulfilling prophecy after a while. Plus, his play calling and reliance on an outdated offense (the single back slamming the line, unfortunately, doesn't provide an option B when your team is down 12) scares me a little, kind of like Bielema with Wisconsin.