This isn't a huge shock, and something many people predicted. Almost every computer model I've seen has the spread at ~15 points. Now, preseason expectations still tamper that a good deal, but this isn't something particularly weird either. In most cases (but certainly not all) the spread matches the computers reasonably closely.
Sagarin hasn't said lately what his ratings are based on, but you can generate something with approximately a 0.98 R^2 using just points scored/points allowed. My guess is the 0.02 remaining is a question of the "decay" factor he's using to weigh previous games less (something like 0.95^[weeks ago]).
It's pretty safe to say he's just using points.
(To compare, S&P+ and FEI give about a 0.8 R^2 with pure points)
Michigan has outscored opponents by an average of 13 points per game. Florida has outscored opponents by an average of 19 points per game. So Michigan is about 6 points worse to start.
Then add in strength of schedule: Michigan's average opponent has been about 1 point per game stronger than Florida's per Sagarin (that's what the 79.70 vs. 78.56 is - strength of schedule scaled to a points per game metric).
Florida has been 6 points per game better in Margin of Victory, but 1 point worse in strength of schedule. To wit, they should rate about 5 points better than Michigan.
Michigan's predictor ranking is 89.38; Florida's is 94.26. In other words, Florida would be favored on a neutral court (per Sagarin) by 5 points.
It adds up. The reason it seems fishy is just that Sagarin thinks the strength of schedule is much closer (only 1 point) than the margin of victory (6 points).
Using those weightings (which are based on making All American teams - a flawed, but decent metric), you get that a 5 star recruit is about 2.6 times more likely to be a star than a 4 star recruit, who is in turn around 4.2 times more likely to be a star than a 3 star recruit.
Now star power isn't the only thing that matters, but I don't see any reason to think the distribution for "role player" or "starter" is all that different.
I think that's a pretty aggressive interpretation of what's going on here. UM has as many fans as anyone, and they're as affluent as anyone.
It has more to do with the fact that they Florida has 19 4/5* commits to UM's 17, and of the top six recruits, Florida has five of them. I wouldn't trade UM's class for Florida's, but there's nothing wrong with this ranking.
Notre Dame's turnover differential was pretty epic. They were -1.1 turnover per game relative to their opponents, and a lot of that was in the red zone (where the value of those turnovers is heightened). In spite of that, they ranked around 20th in strength of schedule adjusted point differential.
They went 8-5 with really terrible turnover results, and below average performance in close games.
I hate to say it, but the "sharp" action is gonna be on Notre Dame in a lot of these games. That team is primed for a major turnaround.
Maxwell's obviously the wild card, but wiht the depth that they've built up in East Lansing, I suspect most of their attrition due to the draft/graduation will be surprisingly non-existent.
If Maxwell can play, they're probably the 2nd best team on the schedule after Alabama. If he can't, well then, uh, yeah.
But honestly, not super different from using a scholarship on a fullback. It means we might finish 3rd in the Rivals rankings rather than 1st, but scorekeeping aside, it's not a big deal.
You can afford to use 1/85th the roster on a bit of a redundancy.
For all these analyses, I'm always curious to know whether the quality of the results would be improved by averaging ESPN/Scout/Rivals. In a lot of other fields, that's obviously invaluable, and my suspicion is it would be here as well.
It's ranking different things. UM is 13 on their list overall, with the undergraduate part of the school ranking 28th nationally. The implication (and the correct one) is that UM has especially strong graduate programs, which boost it to 13th overall.
The disparity between the 13th and the 28th rankings is no more surprising than UM ranking 1st in recruiting rankings right now, but finishing outside the top 10 in the polls. They're measuring different things.
The term that should be used is "not repeatable", which isn't quite the same thing. And even more specifically, it's "not repeatable at the D1 level."
If you have Michigan play against a high school team, I'd very much so expect UM to recover 70%+ of fumbles. And if you had UM play against an NFL team, I'd expect the ratio to be reversed. It's a skill. I don't even think Brian disagrees with that. However, it's a skill everyone who plays at the D1 level is exceptionally good at, so there's a bit of a standstill.
If every D1 football team played 162 games a year, you wouldn't need to regress turnover rate quite so strongly. However, since turnovers are rare, and everyone UM plays against is pretty good at recovering them, it's hard to tell much from which way 30 or so bounces went.
As we know, the staff tells guys not to commit if they still want to take an official visit to another school. This should generally help with decommits, insofar as it helps Deontay Greenberry to Houston type situations from developing.
It doesn't completely eliminate decommits obviously, and probably slows down getting commits in the first place (Ty Isaac wanting to visit USC first for instance), but it does mean once they're in the class, they're probably not likelyto leave.
That said, Jared Gaither is 6'9", so given some random inaccuracy in high school measurement, Orlando Brown might be "only" 6'9" as well. Ryan Mallett's height varied by a couple inches here and there too.
You're using Rivals ratings right? Have you tried matching up their numeric ratings to likelihood of being drafted? I'm curious if that would help give a more granular result?
I also think it's also interesting to compare this to All-American rankings. While five star players are just under five times more likely to get drafted than three star players per your analysis, Dr. Saturday found that five stars were more than eleven times more likely to make an All-American team than three stars. That's a pretty significant gap. I can speculate about a couple reasons for why, but the most obvious is a preference for placing a 5 star onto such a team over a 3 star (while the draft is more of an obvious meritocracy).
Other than Andre Smith, none of Saban's Alabama O-Linemen are playing significant roles even for bottom feeders in the NFL. (Per Pro-Football-Reference's draft checker).
And even Smith is generally considered to be a bust - just a guy who has a job because of the team he plays for.
That he previously attended Cypress Ridge HS (Houston suburb) previously. I imagine there's a good chance that in spite of the Cass Tech pipeline, that the Texas schools will be major players in his recruitment.
If you start moving teams around based on head to head play instead of overall resume, then you just end up creating other issues. I didn't see too many posters here complaining about UM being ranked higher than MSU in spite of head to head play.
If he's an All-American, even for one season, I think calling him a bust will be incredibly unfair.
Yeah it stinks he wasn't able to be a major contributor earlier in his career here, but if he has a Nick Fairley type season, all will be forgiven I believe.
The Colts play a very conservative Tampa 2 style defense. While it is a 1 gap system, they rarely put their corners in man coverage, and blitz much less frequently than other teams. That's not something I'm really expecting to see here.
This isn't terribly shocking, but it's nice to see it comfirmed. I'd be interested in seeing what the data looks like at various other points too. I don't imagine there's anything magic about the number 10 - the effect should remain at a -9 net turnover stat, just less so, right?
It's the returning starters. They're bringing back 7 guys. Only one team brings back less (Boise). They're not just returning relatively few guys - they're returning basically half what a normal team returns.
It's the same reason Boise, which F+ normally loves, has dropped from 4th to 60th. That much turnover is crazy.
Add to that a very high number of returning starters (18 - among the highest in all of college football), and a very high recruiting grade (5th), and that's a pretty strong recipe for success.
How did a 7-5 Michigan team go to 11-2? Mattison was a big part, but a ton of returning starters was a huge part too.
MSU was an excellent team last year. They ranked 10th in F+, which was basically the same as their AP/USA Today/BCS ranking. They're bringing back 14 starters, and an OK recruiting ranking. What's wrong with dropping them to 17? Are they really clearly not a top 25 team? I don't see it.
Notre Dame looks stranger at 12 since they went 8-5, but that's just because F+ liked them a lot last year too. The Irish look very good on all the advanced metrics, because of a strong point differential, a tough schedule, and because of a statistical bias in favor of assuming fumble recoveries will equalize next year. I'm not sure they're wrong either. We were all comforting ourselves with Michigan eventually regressing to the mean in terms of turnovers. I think expecting Notre Dame to do the same sadly isn't crazy.
Tebow is sort of a joke, but he does also run a lot. When you consider QB runs+QB hits, he's getting hit 20 times a game or more.
Peyton Hillis is also 6'2". I agree if Isaac is a true 6'3", it would be weird to see him succeed, but I wouldn't disqualify him because of it. If anything, it says something about how good the rest of his game is.
I don't know what promises Gunner made to Miles - it's possible he pinkie swore and Miles really takes that seriously, but even so - Miles is at the top of his profession. How can a coach, who spends a couple hours a week giving non-answer to reporters, just get pissed off enough to rip a kid for no reason.
This isn't Gundy ripping a reporter. He looked a bit silly there, but I see why he did it. Miles just looks like a child here.
Recent Comments
He'll go back to college when Harbaugh goes back to the NFL.
Seriously, re-read those quotes. They read exactly like the quotes about Harbaugh.
If you were curious.
This isn't a huge shock, and something many people predicted. Almost every computer model I've seen has the spread at ~15 points. Now, preseason expectations still tamper that a good deal, but this isn't something particularly weird either. In most cases (but certainly not all) the spread matches the computers reasonably closely.
Sagarin hasn't said lately what his ratings are based on, but you can generate something with approximately a 0.98 R^2 using just points scored/points allowed. My guess is the 0.02 remaining is a question of the "decay" factor he's using to weigh previous games less (something like 0.95^[weeks ago]).
It's pretty safe to say he's just using points.
(To compare, S&P+ and FEI give about a 0.8 R^2 with pure points)
Here: http://tangotiger.com/index.php/site/streaks-in-baseball
They make a number of errors as detailed in that thread. Their result is largely implausible.
You need to be 21 to sign legally binding contracts there. As result, rather than have two LOI forms, the NCAA just adopted the MS rules.
But this is just by one of the FO metrics (S&P+). They usually use a combination of that and FEI to get F/+.
By FEI, Michigan is 15th and MSU is 34th.
FWIW, Irish fans feel pretty confident he's their's. Is he at a catholic school or something?
Which seems like poor form, so I downvoted to cancel it out.
Simple terms:
Michigan has outscored opponents by an average of 13 points per game. Florida has outscored opponents by an average of 19 points per game. So Michigan is about 6 points worse to start.
Then add in strength of schedule: Michigan's average opponent has been about 1 point per game stronger than Florida's per Sagarin (that's what the 79.70 vs. 78.56 is - strength of schedule scaled to a points per game metric).
Florida has been 6 points per game better in Margin of Victory, but 1 point worse in strength of schedule. To wit, they should rate about 5 points better than Michigan.
Michigan's predictor ranking is 89.38; Florida's is 94.26. In other words, Florida would be favored on a neutral court (per Sagarin) by 5 points.
It adds up. The reason it seems fishy is just that Sagarin thinks the strength of schedule is much closer (only 1 point) than the margin of victory (6 points).
Michigan football or basketball was ranked #1?
Just off the top of my head, you can use Matt Hinton's findings.
Using those weightings (which are based on making All American teams - a flawed, but decent metric), you get that a 5 star recruit is about 2.6 times more likely to be a star than a 4 star recruit, who is in turn around 4.2 times more likely to be a star than a 3 star recruit.
Now star power isn't the only thing that matters, but I don't see any reason to think the distribution for "role player" or "starter" is all that different.
I think that's a pretty aggressive interpretation of what's going on here. UM has as many fans as anyone, and they're as affluent as anyone.
It has more to do with the fact that they Florida has 19 4/5* commits to UM's 17, and of the top six recruits, Florida has five of them. I wouldn't trade UM's class for Florida's, but there's nothing wrong with this ranking.
Notre Dame's turnover differential was pretty epic. They were -1.1 turnover per game relative to their opponents, and a lot of that was in the red zone (where the value of those turnovers is heightened). In spite of that, they ranked around 20th in strength of schedule adjusted point differential.
They went 8-5 with really terrible turnover results, and below average performance in close games.
I hate to say it, but the "sharp" action is gonna be on Notre Dame in a lot of these games. That team is primed for a major turnaround.
Maxwell's obviously the wild card, but wiht the depth that they've built up in East Lansing, I suspect most of their attrition due to the draft/graduation will be surprisingly non-existent.
If Maxwell can play, they're probably the 2nd best team on the schedule after Alabama. If he can't, well then, uh, yeah.
But honestly, not super different from using a scholarship on a fullback. It means we might finish 3rd in the Rivals rankings rather than 1st, but scorekeeping aside, it's not a big deal.
You can afford to use 1/85th the roster on a bit of a redundancy.
For all these analyses, I'm always curious to know whether the quality of the results would be improved by averaging ESPN/Scout/Rivals. In a lot of other fields, that's obviously invaluable, and my suspicion is it would be here as well.
Any thoughts?
It's ranking different things. UM is 13 on their list overall, with the undergraduate part of the school ranking 28th nationally. The implication (and the correct one) is that UM has especially strong graduate programs, which boost it to 13th overall.
The disparity between the 13th and the 28th rankings is no more surprising than UM ranking 1st in recruiting rankings right now, but finishing outside the top 10 in the polls. They're measuring different things.
They might just draft fewer busts, and more of the NBA's total revenue would go to established veterans rather than speculative bets on 19 year olds.
Not that that's a good idea in and of itself, but in terms of the number of dollars going towards unproductive players, it could help.
The term that should be used is "not repeatable", which isn't quite the same thing. And even more specifically, it's "not repeatable at the D1 level."
If you have Michigan play against a high school team, I'd very much so expect UM to recover 70%+ of fumbles. And if you had UM play against an NFL team, I'd expect the ratio to be reversed. It's a skill. I don't even think Brian disagrees with that. However, it's a skill everyone who plays at the D1 level is exceptionally good at, so there's a bit of a standstill.
If every D1 football team played 162 games a year, you wouldn't need to regress turnover rate quite so strongly. However, since turnovers are rare, and everyone UM plays against is pretty good at recovering them, it's hard to tell much from which way 30 or so bounces went.
As we know, the staff tells guys not to commit if they still want to take an official visit to another school. This should generally help with decommits, insofar as it helps Deontay Greenberry to Houston type situations from developing.
It doesn't completely eliminate decommits obviously, and probably slows down getting commits in the first place (Ty Isaac wanting to visit USC first for instance), but it does mean once they're in the class, they're probably not likelyto leave.
That said, Jared Gaither is 6'9", so given some random inaccuracy in high school measurement, Orlando Brown might be "only" 6'9" as well. Ryan Mallett's height varied by a couple inches here and there too.
You're using Rivals ratings right? Have you tried matching up their numeric ratings to likelihood of being drafted? I'm curious if that would help give a more granular result?
I also think it's also interesting to compare this to All-American rankings. While five star players are just under five times more likely to get drafted than three star players per your analysis, Dr. Saturday found that five stars were more than eleven times more likely to make an All-American team than three stars. That's a pretty significant gap. I can speculate about a couple reasons for why, but the most obvious is a preference for placing a 5 star onto such a team over a 3 star (while the draft is more of an obvious meritocracy).
But it's still stunning to see how big March Madness is. Those 67 games generate more ad revenue than the 750 or so college football games do?
1. Need a center. Prefer smart players over physical beasts. David Molk to the rescue.
2. Need a wide receiver. Love 3 cone drill performers. Junior Hemingway had the best 3 cone result.
Brady and Zoltan get a couple of Michigan teammates. Win a Super Bowl. Done.
Other than Andre Smith, none of Saban's Alabama O-Linemen are playing significant roles even for bottom feeders in the NFL. (Per Pro-Football-Reference's draft checker).
And even Smith is generally considered to be a bust - just a guy who has a job because of the team he plays for.
I don't know if that makes it better or worse?
Going to announce soon. Per twitter.
But this nugget from Irish Envy was too good to skip:
"Just stated clearly that Neal wants AU and dad wants ND....come on dad pull the belt out!!"
Something about asking a father to beat a commitment out of his son seems like a very apt summary of the state of Notre Dame Football.
That he previously attended Cypress Ridge HS (Houston suburb) previously. I imagine there's a good chance that in spite of the Cass Tech pipeline, that the Texas schools will be major players in his recruitment.
If you start moving teams around based on head to head play instead of overall resume, then you just end up creating other issues. I didn't see too many posters here complaining about UM being ranked higher than MSU in spite of head to head play.
If he's an All-American, even for one season, I think calling him a bust will be incredibly unfair.
Yeah it stinks he wasn't able to be a major contributor earlier in his career here, but if he has a Nick Fairley type season, all will be forgiven I believe.
The Colts play a very conservative Tampa 2 style defense. While it is a 1 gap system, they rarely put their corners in man coverage, and blitz much less frequently than other teams. That's not something I'm really expecting to see here.
Huh? Index card?
Does anyone expect the Barwis-Michigan connection to continue once the RichRod generation of recruits have won multiple B1G titles and graduated?
I'm also curious - are these guys still in school? Mike Martin clearly has some NFL future. Has he graduated, or is he in full NFL draft mode now?
I don't really want analysts guessing at coachability as well. I'd rather they focus on technique, athletic ability, size, etc...
Coachability is obviously important, but that's just such a black box that Rivals/Scout aren't going to do anyone any good by guessing at it.
I'm just curious - I've been looking for a study along those lines?
This isn't terribly shocking, but it's nice to see it comfirmed. I'd be interested in seeing what the data looks like at various other points too. I don't imagine there's anything magic about the number 10 - the effect should remain at a -9 net turnover stat, just less so, right?
Would be nice to see how much less.
I know ESPN may be dumb, and fake 40s are fake, but that's not a great sign if he's a TE at that size.
I have to think he's a DE or he's gotten a lot faster (or bigger).
There are still more than 500 uncommitted recruits of at least three stars. I imagine they'll fill out that class somewhat?
It's the returning starters. They're bringing back 7 guys. Only one team brings back less (Boise). They're not just returning relatively few guys - they're returning basically half what a normal team returns.
It's the same reason Boise, which F+ normally loves, has dropped from 4th to 60th. That much turnover is crazy.
Add to that a very high number of returning starters (18 - among the highest in all of college football), and a very high recruiting grade (5th), and that's a pretty strong recipe for success.
How did a 7-5 Michigan team go to 11-2? Mattison was a big part, but a ton of returning starters was a huge part too.
You'd see they have 14 returning starters.
MSU was an excellent team last year. They ranked 10th in F+, which was basically the same as their AP/USA Today/BCS ranking. They're bringing back 14 starters, and an OK recruiting ranking. What's wrong with dropping them to 17? Are they really clearly not a top 25 team? I don't see it.
Notre Dame looks stranger at 12 since they went 8-5, but that's just because F+ liked them a lot last year too. The Irish look very good on all the advanced metrics, because of a strong point differential, a tough schedule, and because of a statistical bias in favor of assuming fumble recoveries will equalize next year. I'm not sure they're wrong either. We were all comforting ourselves with Michigan eventually regressing to the mean in terms of turnovers. I think expecting Notre Dame to do the same sadly isn't crazy.
It is what it is. Would have been nice though.
Just cause NSD to occur in August? The same insanity, just earlier?
What am I missing?
I mean, schools redshirt kids for a reason. It would be pretty crazy to prefer to get the age 18 season of a kid instead of his age 23 season?
Is it because they think all their kids are going to the NFL?
Tebow is sort of a joke, but he does also run a lot. When you consider QB runs+QB hits, he's getting hit 20 times a game or more.
Peyton Hillis is also 6'2". I agree if Isaac is a true 6'3", it would be weird to see him succeed, but I wouldn't disqualify him because of it. If anything, it says something about how good the rest of his game is.
I don't know what promises Gunner made to Miles - it's possible he pinkie swore and Miles really takes that seriously, but even so - Miles is at the top of his profession. How can a coach, who spends a couple hours a week giving non-answer to reporters, just get pissed off enough to rip a kid for no reason.
This isn't Gundy ripping a reporter. He looked a bit silly there, but I see why he did it. Miles just looks like a child here.
Mike Dyer
@MikeDyerMoeller senior WR Monty Madaris is "undecided" at this point, according to Moeller coach John Rodenberg
#enqr#MichiganState#UC