MANBALL DOESN'T NEED STATISTICS. MANBALL ONLY NEEDS MEN TO HANDLE THE BALL IN A MANLY WAY.
Mid-Week Metrics: Mythbusters: Manball edition
Power. Strength. Toughness. Big Ten Football.
This is the new (old) Michigan football. What this actually looks like remains to be seen, but I wanted to test out some of the core tenants and clichés of the Manball philosophy to see if there power still rings true today.
Bring on the charts!
Check here for a run down of the background behind the methods.
Myth 1: Passing too much on offense makes your defense ill-prepared for the rigors of Big Ten play.
I tested this myth for both all college football and the Big Ten exclusively. If it’s going to be true anywhere, it’s going to be true in the Big Ten.
To judge how much a team passed, I looked only at first half plays where teams haven’t made half-time adjustments and should be executing their intended game plan and not reacting much to score and time considerations. I then compared the quantity of first half passes against the defensive success. First I looked at all of the FBS:
That’s a whole lot of buck shot and not a lot of trend. There is a slight trend toward more passing = better defense but the effect is not statistically significant.
But as I mentioned earlier, the Big Ten is different than the rest of FBS, it is the nativeland of Manball. So if you look at Big Ten teams in Big Ten games over the last eight years, does the picture look different?
Here at the least the slope is going in the “right” direction but the effect is still small and insignificant. Even if it was statistically significant, the difference between the low (10 passes per first half) and the high (25 passes per first half) is worth one game a season, an advantage sure, but nothing monumental.
Myth 2: Long Scoring Drives Rest a Defense
Unfortunately I don’t have any good tools to tell how rested a defense gets, but I can look at the outcomes of subsequent drives following a scoring drive of various lengths. Does a defense have better outcomes after a long or short scoring drive, does any of it matter at all?
Looks like the rest is more beneficial to the offense than the defense. Defenses give up 20% more points after a 15 play scoring drive by their offense than a 1 play scoring drive.
The usual correlation does not equal causation applies. Worse teams could be more likely to score on longer drives than good teams. Other issues could be at play but I felt comfortable that this overall myth does not hold true.
Myth 3: Running Teams Do Better in the Red Zone Than Passing Teams
I had two ways to look at this one. Is it about running the ball in general, or is it about running the ball once you are in the red zone? They are usually the same thing but I wanted to test out both to see if one rang more true than the other.
First, comparing how much teams run between the 20’s to red zone effectiveness, measured in [points on red zone trips]/[7*red zone trips]:
This looks a lot like Myth 1. Some slope but no significance. Even at a significant r sqaured, the difference between 30% rushing and 60% rushing is worth less than a touchdown in red zone production over the course of an entire season.
Here is what it looks like when you change the x-axis to reflect playing calling within the red zone:
Slope increases, as does r squared although there is still a ton of noise.
The case is not strong, and there is definitely more than one way to skin a cat in the red zone but I would leave the door open on this one:
Finding: Plausible, but evidence weak
Myth 4: Offenses With Running Quarterbacks Break Down As The Season Progresses
This one is probably not a manball myth, necessarily, but a good one to look at. Let's go straight to the you-know-what.
Did not see this one coming. Sure last year clouded my mind a little bit but I did not expect QB running offenses to be this dominant. That’s a very real gap between QB running offenses and non-QB running offenses.
The weekly data here is a bit noisy but it looks as though offenses built around running QBs peak in early November but are still pretty strong come bowl season. The overall trend roughly mirrors statue QB offenses although the statues do have a bigger uptick come bowl season than other offenses.
Myth 5: Offenses With Running QBs Have Worse Defenses
Not a lot of fancy numbers or charts on this one. Only real numbers of note are that the 100+ carry group from Myth 4 have an average defense of that is 0.2 points per game worse than then 0-99 group, that’s worth less than a game a decade.
Myth 6: Run Oriented Offenses Do Better In The Fourth Quarter
This is one of the key tenants of a run-based offense. The ability to hold the ball with a lead late. Unfortunately the NCAA doesn’t provide time stamps for plays and so I don’t have them in my database, making a good estimation of clock killing impossible to determine from my data. All I can provide analysis on is the ability of different combinations of run and pass to score points, not run out the clock.
Partially because objectives change in the fourth quarter, but the likelihood of scoring is the lowest in the fourth of any given quarter. That means all situations will tilt toward the negative in my analysis. What I can look at is how much teams run in the first three quarters and compare that with their overall performance in the fourth quarter when the game is within two touchdowns.
I hope I didn’t just give away the ending, but if you are going to be a running team you better come into the fourth quarter with a lead. One of the strongest correlations of the day points to strongly diminished returns in the fourth quarter for teams heavily invested in the run.
Finding: Busted without a lead, inconclusive running out the clock
What Does This Mean For The Future of Michigan Manball?
Right now the evidence still points to Manball being more of a philosophical theme than a practice of playcalling but that doesn’t mean it’s not going to happen either. Nothing I have seen indicates that it can’t win a lot of games but it is definitely far from a Decided Schematic Advantage. As all good Michigan fans know, Manball can be effective in most games as long as you have better talent and you aren’t playing from behind.
Manball is not an offensive scheme, it is something that applies to both sides of the ball, it applies to the special teams as well. It applies to the way the players practice. It is about giving your best shot every time, it means letting Coach Hoke "hear" you at practice, it means being accountable to your teammates. It is the defining philosophy of this coaching staff.
and thats why improving the defense is the most important thing that michigan needs to do. you get a good to great defense put together the type of offensive system becomes less important. offense is obviously important but if you get a d put together that can hold oppoenents to between 21-24 ppg then any type of effective offense should work in theory. of course the good to great defense won't happen this year or probably not even next year but building towards the future the defense is the most important part imo. good diary entry though.
All the dumb reasons that people point to as reasons Rodriguez failed in the Big Ten (Spread! Small! Too much Denard! Toughness! TOP!) were wrong and it really just comes down to having quality players with experience in a system for a few years with a capable defensive coordinator?
Once again, there are reasonable and unreasonable ways to bash RichRod. Too bad we get more of the latter than the former.
SDSU last year played an extremely easy schedule with 7 of their 9 wins coming against teams a combined 46 games below .500. Lets look at their game stats from their 4 losses and two wins against teams with winning records:
Air Force had 29 First downs to SDSU's 16. SDSU 31 rush 27 pass. Win
Mizzou 25 first downs to SDSU's 17. SDSU 33 rush 44 pass.
BYU 24 FDs to 12 for SDSU. 12 rush 36 pass.
TCU 27 FDs to SDSU's 7. 21 rush 26 pass.
Utah 22 FDs to SDSU's 29 (25 by pass). 21 rush 54 pass.
Navy 22 FDs to SDSU's 27. 41 rush 23 pass. Win
They had one other close game (single digit win) against 3-8 CSU. 39 rush and 23 pass with a 3.9 yard per carry average. Not the numbers you would expect from manball.
Wow. When did Brady Hoke and Al Borges say the word "MANBALL?"
Great strawman you're tearing down here.
as it mainly seems to make some people angry. Personally I have no problem with it, I think it's kind of funny (and almost certainly not descriptive of anything substantial) but I am tired of seeing the same arguments back and forth every time the word MANBALL appears in a post.
The season can't start soon enough.
If you want to determine whether resting a defense helps said defense, why would you just compare numbers following a scoring drive? It would be better to compare general time/length of plays in all drives -- for instance, 3 and outs, turnovers, etc. A one-play score on offense may fire up the defense and negate a general trend where length of previous offense drive helps a defense.
Moreover, it might be an aggregate phenomenon. A defense might not produce differently after a one-play drive vs a 15-play drive, while still performing differently being on the field 20 minutes in a half vs 10 minutes in a half.
Agreed, it would seem to get "tired" would take some number of plays. Certainly no defense is tired if their offense goes three and out in the first drive. Maybe a break down by quarter, or look at third / fourth quarter specifically? After x plays or drive? Of course later in the game coaching adjustments might effect the effectiveness of defensive squad unless you consider those to be a wash.
I wonder what the average number plays/time a good defense is on the field versus a bad defense.
I may be in the heavy minority on this, but if you could provide t-statistics and p-values when presenting regression results (in addition to the coefficient estimate and R-squared that are already provided), that would be great.
Wouldn't necessarily do much to change inferences from these results, but disclosure (up to the point of diminishing returns, at least) is good. If it's too much of a pain here, perhaps put full data analysis results on a separate webpage for those of us interested?
I am inclined to agree ^
Great post mathlete...
I think your myth #4 analysis shows the simple theory that having a running quarterback gives one a numbers advantage works out. RR was and is obviously right in this aspect (but not in his total approach to running Michigan). I loved the guy, loved the offense... but the D, special teams were incredibly lacking.
Brady's blow hard attitude about manball I think is more attitude than intended to give away the playbook. Borges seems to be the right sort to find the right blend of things from his press conference (and has been pass happy at SDSU). I expect good things! :)
It would be interesting to see a variation of #6, where instead of using offensive value you looked only at teams' won-lost records using a subset of games where they went into the 4th quarter behind by one score or less, tied, or ahead by two scores or less. I suspect that running teams did very poorly in games where they had to play catch-up, but if you're within one score you can still run your basic offense. I think two scores on the other end is appropriate because you'd think a ball-control offense would help you execute the old Tresselball formula of getting ahead by two scores and then sitting on the lead. If the lead is three scores or more, most of the 4th quarter is likely to be garbage time which would screw up the sample.
Without differentiating between types of offenses I don't see how this data is helpful. There are spread passing (Texas Tech of yore) and spread running (WVU/Mich under RR) types of teams. We probably ran the ball more last year than we will this year, does that mean we're moving to less manball?
I see your point, but in aggregate the results show that there is precious little correlation or causation between "tough running" and a better team. What it does show is that a good defense and a good, diverse offense will make a team successful; just trying to "man up" and run down a team's throat doesn't assure you of a victory.
While I'm appreciative of the effort put into this, I will say that you should be careful not to use an R-squared value to validate statistical significance, like you alluded to in Myth 3. Rather, it should be used to determine how much of the variance can be explained by the model.
If you were to run t-tests on specific variables, you could get a better idea of which ones were useful predictors for success or failure, and start to talk about statistical significance. Overall, though, awesome work. As a statistics nerd, I love reading through these in-depth analyses.
The r2 is a simple function of the f- stat that tests the null that all the regressors have no explanatory power. So it does have a dual use though unlike a t stat where it's pretty simple to see if it has an adequate p value at least I don't have that same intuitive feel.
Did anyone say a thing about how "passing teams" defend the run? Or how passing re as fare in the redzone? Hoke didn't. I'm unaware of anyone else. There's a ton of straw man burning in this post.
I'm still waiting for ANYONE to define "manball" in a way that is consistent with a word Hoke has uttered.
I'd like to give it shot because I think the petulance behind the term is justified to a certain extent. But I think this conversation would go smoother and faster if you could summarize what you think it is that Hoke said.
In the meantime, I'll look up the phrases that annoy me tho most and try to be thoughtful in my criticism of those phrases.
I know you'll be waiting on the edge of your seat.../s.
What Hoke said was that teams that exclusively zone block have a hard time getting their defense ready for man-blocking schemes. (this is where his basketball on grass comment came from)
Of course, that claim isn't what is tested here. Neither Hoke nor any prominent figure I'm aware of claimed that passing teams can't defend running teams. Even if they did, it's completely non-descriptive of the ground-based attack RR ran. Nobody claimed running teams can't score touchdowns. I'm unaware of this running QB claim ever being made. I don't see what any of this has to do with "manball", but at makes sense - "manball" seems to be an ever-contorting term that can mean whatever irks he user at the time. Brian used it as "punting at inopportune times". What does that have to do with a thing Hoke has espoused?
That, as far as I can tell, is the beginning and end of the "manball" meme.
I have also been completely been confused by the content around "manball" around these parts lately. While I have some academic interest in how well traditional power football works, I don't recally Hoke saying we would only run the power play. He only said that, as you say, Michigan wouldn't be "only" zone blocking.
Preaching toughness and physicality along the front lines is definitely not the same as saying we're going back to 3 yards and a cloud of dust.
On another note, and I'm pretty sure I'm in a small minority here, I'm amazed how many people can be so fanatical about tradition at Michigan (so many "get off my lawn, you kids" moments") and still have such hatred for the traditional offense that defined the team for so long.
Sorry Chitown, I said I’d try to be thoughtful; it seems that means that I cannot be concise.
While I think it is fair to point out that Hoke's rhetoric can't be thoroughly summarized with a single word, I do think that the term MANBALL is reasonable shorthand for people who have thought about what Hoke has said. I'd put Brian into that category. As for the people who run with the word without thinking what it may or may not mean: [screw] ‘em, they won’t be reading this.
You referred to interview most people, myself included, think of. However, when that particular comment is referred to in defense of Hoke, a key parenthetical phrase is typically leave out: “basketball on grass.” To me, that phrase is key to trying to gain insight into Hoke’s view because it implies a certain lack of respect for that philosophy – it’s not really football, it’s actually basketball. The interpretation from the interviewer who actually took the quote is consistent with my own (link). To wit:
When asked recently about the influence of Oregon’s offense, Hoke subtly revealed his disdain for the tactical shift Michigan experienced under Rodriguez. He is convinced that modern spread option offenses can be counterproductive to the core values of smash-mouth football and are, therefore, to be avoided.
"Right, wrong or indifferent, when you're zone blocking all the time -- when you're playing basketball on grass -- you practice against that all spring, you practice against it all fall and then you're going to play a two-back team that wants to knock you off the football" Hoke said. "I don't think you're prepared."
That rhetoric is in the same vein as omitting the word State from one’s vernacular. To me the word MANBALL is attacking the way Coach Hoke projects to think that certain schemes are inherently better than others. There seems to be a certain closed-mindedness embedded within the comment that really bugs me. I mean, REALLY bugs me. Like a mofo. Particularly within the context circa January 5 when that article/column was published.
In the context of that article, Hoke seemed to be trying to say that if you don’t practice against man blocking, then your defense is going to be a gaggle of bitches. Pat White’s defenses were a bunch of pussies. As were Woody Dantzler’s. And Tim Tebow’s. Vince Young’s, too. TCU, bitches. Boise St., hoes. Oklahoma. Mississippi St. Oregon. Auburn. C’mon, Brady…
Brady Hoke never attacked any of those teams, he just attacked they way they played.
I do not believe that Hoke actually believes any of that. His past behavior at Ball State and San Diego State as well as his subsequent comments bear witness to that. But on January 5, the past behavior stuff needed still needed to be mined and the subsequent comment stuff didn’t exist.
The rhetoric elevating smashmouth defense and castigating spread offenses is code for “the spread won’t work in the tradition rich B1G. We play MANBALL here.” I don’t think Hoke is this type of guy, but it could be a slick shive into RichRod’s corpse. A lot of Michigan folk are gobbling that shit right up.
It’s brilliant pandering, all while living in van down by the river.
Brady Hoke's been a coach for decades. I'm gonna go ahead and say he's probably not being closed-minded when critiquing his own personal preferences for developing his defensive mindset.
The subtle and non-so-subtle criticism of Hoke is an embarrassment, and I hope to God he kicks serious ass in Ann Arbor so this crap will stop.
That;s not quite what I said. I said he came off that way. It's just, like, my opinion, man. Hoke has walked back from there since then.
We'll see of the MANBALL philosophy will help Michigan's D finally stand up to the fast break style of play -- basketball on grass -- that has shredded us since it errupted onto the landscape.
Donovan McNabb. Antwaan Randle El. Armanti Edwards. Dennis Dixon. Juice Williams. I'm sure there are others I have suppressed into the dpeths of my mind.
Oh yeah, Vince Young.
God, please, make it stop!
Taylor Martinez? Nathan Sheelhasse? Minnesota Dude?
I think Hoke and his homies have the guile to do what Michigan needs them to do. We're on the same page in that regard.
If you want to attack that piece of Hoke rhetoric, that makes sense - I would agree that it is a dig on Rodriguez. But attacking the rhetoric is not the same as the rabble-rousing play calling complaint this site engages in - which you acknowledge.
Just come out and say what you think about Hoke. You call him a panderer who just wanted to put a "slick shive into Rich Rod's corpse." You then put a bunch of words and beliefs into his mouth that he never said. (He never said that the spread doesn't work. We also have no idea what the offense will do. Other than the quotes from Borgess about 50% shotgun, etc. which you ignore.) You then go out of your way to say that you don't think that Hoke really thinks any of the things you implied he does. So, you're not sayin', you're just saying? If you have already decided that you don't like Coach Hoke then say it. (Though you face an uncomfortable comparison to the "blue hiars" who decided that Rich Rod would never work in July of 2008.)
One other question - did you really sit through last year's Wisconsin game and not think that our guys weren't prepared for Wisconsin running the ball down their throats?
Dude, I know it's hard to beleive, especially for a pirate, but I like Hoke. All Michigan fans do. Even Brian. I just don't like this compartment of his rhetoric, that's all. And I never will. There are some things Lloyd said that I still don't like. Rich, too. Even Bo. I like all of those guys. Maybe you can't deal with such dichotomy, but frankly, my dear, I don't give a damn.
My comment was directed toward chitown since he asked for a description of what MANBALL means. I think I gave it a pretty good go.
What do I think of Hoke? I think he's a good coach with is best results still ahead of him. I also think he's a way better politician that coach Rodriguez was, which is an asset that will serve all of us very well. That counts double becuase of the frailness of the egos of Michigan fans.
The MANBALL thing was born in the immediate aftermath of the hire; the result of a process, The Process, that was another political charade. That context is key to understanding the term. As Hoke and Borges have spoken subsequently and Hoke's track record has been scrutinized more carefully, Hoke's own words from the article--which the author of that column also interpretted as subtle disdain for spread offenses--have been proven to be pure rhetoric.
Hoke was only pandering to, apparently, the likes of you. It was the right thing for him to do politically. It still is. And I still don't like it.
As for the Wisconsin thing, I don't think our guys were good enough to deal with one of the top 5 best offenses in the country. Hell, they weren't good enough to deal with a mediocre offense. No, they weren't prepared, but I don't think that had anything to do with not practicing against MANBALL.
My issue is that this site does not attack "MANBALL" as a piece of rhetoric, it attacks "MANBALL" as a real thing. First, "MANBALL" was "not shotgun", I guess (remember that bleeding pulp of a horse?), then it was "runs power". Then it was "runs on first down". Then it was "punt at inopportune times". As I've said, it's an ever-moving target that Brian applies to any football strategem he doesn't find intellectually stimulating (which, I think, is why he likes the Spread - he finds it more interesting to dissect). Now it's "passing too much doesn't prepare you to defend the run", "teams with running QB's don't have good defenses", and "teams with running QB's suck at the end of the year", all of which are utter nonsense, none of which I can recall a single person with this staff ever espousing.
This site has taken a SINGLE illustrative comment (the one I referenced) wrapped it in a mountain of coach-speak hyperbolic nonsense (show me a coach in history who doesn't talk about "TOUGHNESS" to the media) and proceeded to use it as a cudgel to bludgeon anyone without a shred of critical thinking ability.
Here's a simple fact: Brady Hoke isn't an offensive coach. What he says about the offense is nearly irrelevent (I won't say completely, because he obviously shapes the overall approach). Al Borges, on the other hand, has always run a west-coast offense that skewed towards passing at UCLA and SDSU, and towards rushing when he had Ronnie Brown and Cadillac Williams.
Yeah, I think that's a fair beef. What started as a rhetorical devise mutated into a hyopthesis that was tested and refuted. Then revised and retested. Then refuted again. Wash, rinse, repeat. Even though each hypothesis isn't directly out of Hoke's words or behavior, once the specter of MANBALL left the bottle, the theories were created in the minds of the proletariate, they must be exterminated.
Such is the process of winning over the hearts and minds, r-e-a-l-l-y s-l-o-w. I think that all of these excersises are worhtwhile, especially in light of the degree of criticism that Brian & company have put out there in regard to The Process and Hoke.
Brian stated an honest and candid opinion, which I and many others shared, and stuck to it. In the face of a massive and stout backlash. I commend him for that. It would be easy to issue a mea culpa and teleport to the Hoke Harrem, but he has done the responsible thing and gone about the process of killing the genie. Anything short of that would be flip-flopping and wouldn't work. He has acturally walked his conclusion back quite a bit, which I think you would acknowledge. Brian is doing the thing that makes mgoblog so damn good and has provided us something to talk about all the while.
Frankly, I don't see how stating something that is patently dishonest, admitting it was wrong, then replacing it with another patently dishonest thing, then repeated the process ad nauseum is "responsible".
Brian doesn't need to like Hoke, that's his business, and it's fine. But creating coaching philosophies out of whole cloth, ascribing them to Hoke, then admitting it was bullshit (then shifting to a new philosophy created from whole cloth that's equally nonsense) isn't cool. This is what people did to Rodriguez which Brian, rightfully, shat on them for.
The characterization of the whole thing as dishonest is where we seperate. I think i know what you mean, so I won't bust your chops overe it.
These ideas now exist, maybe not in your or my mind, maybe not the minda of the mgoblog staff but they most certainly exist. As such, they must be exterminated and it must be done on the front page of mgoblog.
Having said that, I will allow your objection to the boondoggle.
Alla prossima, chitown.
How much of the "ideas that exist" exist not because well read guys like yourself see said quotes and follow other actions and come to their own conclusions, but because Brian's been beating it like a cheap drum? I'd say the vast majority of fans are basically uninformed, and don't know or don't know what the hell MANBALL even means (in any way), and don't care, because they want some MANBALL. But the biggest group of those that are semi-informed think MANBALL might even be a problem is because Brian has created that conception. Then there's a tiny group like Brian and yourself who have come to such conclusion through honest thought and analysis.
I guess the point is that if the idea's other there and needs analysis it's as chitown said, only because he created it in the first place. Kinda having Dr. Frankenstein selling the pitchforks and torches to destroy Frankenstein's Monster. Or maybe Brian is writing to that small percentage who came to their worries on their own. But I think if that's the case he's not spending the ratio of time talking about it that equals the ratio of people who are out there who are concerned.
He doesn't have to love his personality, and certainly doesn't have to be a fan of their playing style (but actually seeing it before deciding that might be nice). Just don't turn into what you hated when people who didn't like Rich from the start were creating meme's that had no basis in fact just because the program didn't go in the way you wanted.
Oh no, you made fun of my avatar. I am really hurt now. I gues will forgive you since you are underwater, drowning because you are trying to hold up a sword.
With your clarification, I think I agree with you most of the way. Your assessment of Hoke as a coach is pretty fair. Though, I do not think that Hoke is a pandering politician. I think he does not like the spread as an exclusive concept - he wants the team to be able to go old-school big ten as well. However, he isn't going to turn Denard into Navare as some people assume.
What I was most reacting to was my impressive of where you were coming from. I was sickend by the people who never gave RR a chance. It doomed him from the start, and doomed the players and fans to three years of pretty bad football. My impression is that many of the purveyors of the manball meme are seeking to revenge RR by buring Hoke before he starts. I hope we can agree that would not be good. We don't need another three years like the last three from a W-L prospective. That would mean 6 years of pretty bad football and higher school seniors having to think back to 6th or 7th grade to remember when Michigan was good.
This is the best explanation I've read so far--great work MCalibur, and I can't believe this was downvoted.
If you will allow me to tack on, I think Brian and many others saw Rodriguez as the future of football, kind of the leader of a paradigm shift for Michigan. We were tired of the Carr teams that constantly underplayed their talent--Carr's last decade was full of Michigan being on the wrong side of multiple upsets, without any good ones to counter. When we hired RR, he was a guy who was modern, wasn't so conservative, was not beholden to 3 yards etc., and whose offensive scheme looked like something that could make us compete with Ohio State, Florida, USC, and the other successful teams we had struggled against up until that point.
Hoke's hire seems to reverse all of that. As MCalibur said, his comments reflect a close-mindedness about the spread offense that Brian in particular was fighting against for the past 3 years. Whereas Rodriguez was the spread guru, whose offense was terrifying when well executed, who had different wrinkles on each of his plays to stay a step ahead of the defense, who was a football intellectual Brian could identify with, Hoke's initial interview (quoted above) makes him look like a Bielema-like meathead* who can only talk a lot about toughness, power running, aggression--i.e. MANBALL.
The references to conservative punting strategy is related. Romer studied it, and I think even the Mathlete had a diary analyzing the punting percentages relative to going for it on 4th-and-short. RR seemed more open to going for it when appropriate, instead of playing the ultra-conservative strategy of, say, punting from the opp. 39. Again, it goes back to the close-mindedness some of Hoke's comments seem to imply. He might not be Carr 2.0, but his actions seem to indicate he might be in some frustrating ways.
*I don't actually believe he is a Bielema-like meathead. He seems open to letting Borges take control of the offense and doing what he wants with it, and maybe the rhetoric we took to be MANBALL was just pandering in the first place. We'll see when the team takes the field, and more importantly, how it does when things aren't going well. Solid performances in bowl games will prove Hoke's skill as a game and scheme coach.
Hoke made a solitary comment abount zone-blocking.
The rest of this is 100% conjecture.
Further, no matter how badly people want to bring Lloyd Carr into this conversation, what Carr did in 2007 is irrelevent. We didn't rehire him to run this team.
We were tired of the Carr teams that constantly underplayed their talent--Carr's last decade was full of Michigan being on the wrong side of multiple upsets, without any good ones to counter.
Carr's last decade started in 1998. This statement is so wrong I don't know where to start with it.
First off, it's a testament to Lloyd that Michigan rarely upset other teams. It says something when your team is always the favorite, doesn't it?
Second off, if you cannot find great victories (whether "upsets" or no), you're not looking that hard. They're all over Lloyd's resume and I don't have the time to list them.
If I had to break out a couple, well, tell me the one time when Rich Rodriguez bitch-slapped an opponent on the road the way Lloyd did it to Penn State in 1997, Ohio State in 2000, Notre Dame in 2006, or Florida in 2003 and 2008. If you're not choosing to remember those epic victories by our "boring, tired, conservative coach," you're just being myopic.
But if Brian built up an image of Rich as some intellectual who was working out pie charts in the football offices, he deluded himself into a bigger fantasy than I would have thought. The funny thing is, Rich and Brady are much more alike in demeanor, (as in folksy, football coach types who you're not expecting a lot of intellectualism from, but know football and can relate on a one on one basis) than either of them are with say, Lloyd, who was quoting Kipling and stuff that Alumni ate up but wasn't really a factor on the football field. Rich had his fair share of foot in mouth disease, and it's not beyond the pale that you could see Hoke tasting shoe leather at some point. But because Rich Rod made minor adjustments to plays between games, to throw off the opposition for a series the next game...that doesn't make him a genius, it makes him a football coach. I just wish there was as much genius thrown onto the defensive side of the ball, because then he'd still be here. Because if Borges offense sucks, no one is going to say "hey, Hoke is a defensive coach, don't blame him". They'll all be going down with the ship too. Because that's what the captain does.
Brian may enjoy breaking down a spread offense more, and apparently offense more than defense, and that's fine. And Rich is a really good offensive mind. But to act like bringing back an offense that was used quite often back in the and modernizing it may make him an innovator, but it doesn't make the offense "intellectually superior". And you're correct, that perception may be there. But it's funny how that perception is ok that we're moving back to "cave man football run by troglodytes", but the idea that "Rich was a hillybilly that didn't get Michigan" was so offensive. I'd think unfounded character shots would always be objectionable.
Also, I've never heard anyone say "offenses with running quarterbacks have worse defenses." Not once.
Your database is similar to what Tom Tango, et al. used for The Book about baseball, which is one of the most interesting sport analysis books I've read. Curious: could you translate PAN into win probability added?
On Myth 3, I'd be interested in how pass offense PAN correlates with red zone effectiveness, vs. how rush offense PAN correlates with red zone effectiveness.
On Myth 4, I had a slightly different take -- running QB offenses DO break down more at the end of the season. But they start so dang high to begin with, so the drop is faster/further (farther?) but they still end up better than other types of offenses.
First off, great work Mathlete.
I see two conclusions from your graph relating to Myth 4.
The first is that teams that run there QB 100+ times a season have a higher offensive value than teams that run 100 or fewer times over the course of a season. This conclusion is based on the y values associated with each of the three classes.
The second is that at the very end of the season (say week ~9-11 onward) the teams that run 100+ times see a decline in there offensive value relative to early in the season and teams that run 100 or fewer times see a slight increase. This conclusion is based on the slope of the three classes and is what addresses your stated myth, that offensive value decreases over time. You concluded that the myth is busted, but the slope (at least at the end of the season) doesn't bear this out visually.
Currently I would say that Myth 4 is inconclusive. If you fit a line through each class like with the graphs associated with the other myth and there is no meaningful difference between the slopes I would then say that Myth 4 is busted.
Otherwise excellent analysis and interesting results. And regardless it still appears QB running more is better, just not as much better at the end of the season and in bowls.
We Sucked with Rich Rod as our coach, for three years. I hope that the program can be mended as fast as it was decimated. Best of luck to him, good guy, but let's stop wishing we could return to something that was not working.
Yeah, I don't like to read articles before responding either!
This OP clarifies why score, and not statistics, measures success.
"You see, a team wins a ball game when they score more points than the other team."
- John Madden
I'll need to see a chart and some advanced metrics before I believe that old bit. I mean, really.
Great work as usual. I do wonder about the whole "running teams win games" mantra, since good teams tend to be good throwing AND running the ball; that's why they are good. Teams that are one-dimensional tend to be exposed by good defenses, and over time they suffer as teams figure out whatever they do well. For example, Purdue's spread attack used to confound B1G teams that weren't used to it, but toward the end even mediocre defenses were having some success slowing them down. Those Boilermaker teams couldn't run the ball to save their lives some seasons, and as a result teams started to load up on DBs and dare the QB to pick them apart. It didn't work with Drew Brees, but as be begat Orton who begat Painter who begat whoever, you saw it fail.
I won't make this about RR because I don't think the Mathlete was going that way, but these numbers prove that the failure of his reign wasn't on the offensive end. I have full faith that Borges & Co. will take advantage of the players available and produce a good offense; it will be up to Mattison to turn those players on defense into a cohesive unit.
Perhaps the Mathelete didn't mean that post that way, but the whole "Manball" theme is heavily slanted toward the Hoke = Bad Hire camp. The use of stats and charts and regressions to tell us that Rich Rod was really a blazing success while Hoke is a fool for his beliefs is getting a bit old. Ask yourself what is really the point of taking a bunch of things that Hoke never said and disproving them if you don't mean this a shot at Hoke?