Michigan Football, Team 139, 2018 Season

[Jeff Buchanan]

Typical Harbaugh on injuries but the comment on Dwumfor is encouraging. He reportedly tweeted yesterday afternoon that "he's good" which can mean a lot of different things. Betting it's not an Achilles Tendon rupture like the situation seemed to indicate it might be. Cart was definitely precautionary based on exam findings on the field. That made it look more scary than it probably was. In NFL lingo he's doubtful.

Also on Gary. We now know it's an AC joint problem. That is way different that a labrum injury and probably a lot less problematic depending on the grade. There are three grades. It's a I (a slight separation) or he wouldn't be playing at all. They are troublesome though. Henne and Navarre both had them at various times during their careers at M. I recall when they were most troublesome it really affected their ability to throw. A less impactfull thing for a DE but it's probably painful. It would not surprise me at all if he had a steroid injection after the NW game and that's why he didn't play v. Maryland. You rest for 7-10d after one of these. They are very safe and easy to do injections. I've even done them on patients. You can't do a lot of them though. Usually the ortho docs say once every 3 months. He'll play - 90% certain based on what we know.

I'm not sure about Evans.He did dress for Maryland, warmed up and then changed back into street cloths. Hammys are tough. Once it's pulled, it's really easy to re-tweak it. If it is painful on certain movements, it gets into your head more than do more damage.

Kemp will probably be available. No telling what he is dealing with. I haven't seen enough video of the injury or anyone offering an evaluation of the injury on line.

I'm suspecting Black could be ready to practice next week but that's pushing it. More likely we'd see him for PSU on November 3rd. If he's going to play at all this season, he'll play v. IU at home. If not, we won't see him until a bowl game, if at all.

[hack]

I hate my right hamstring. Fucking hate it.

[Jeff Buchanan]

Michigan to the Peach Bowl v. Miami?

Here's the Bowl predictions as of Sunday evening from SB Nation (put this on the Bulletin Board in the locker room ...... fuck osu):

https://www.sbnation.com/college-football/2018/10/7/17947186/bowl-game-projections-2018-college-football-playoff-predictions

[Jeff Buchanan]

Michigan has moved from 70/1 to win the NC before the season, to 25/1 at the start of last week, to 16/1 today. At 16/1, the team is the 6th most likely to win the title according to the oddsmakers.

While for fans, the optimism is mostly about feelings ball, objectively, M seems to be a contender. The play of the defense and Patterson along with the play of OL shaping up would give rise to justified optimism.

But let's be real here. The last time M faced a top 25 team, it lost. Wisconsin beat M last season in Madison. Harbaugh hasn't beaten osu, got ass-raped by PSU last season and has 1 win v. MSU since the 2014 season. There are chinks in M's armor in 2018 most prominently scheme variability that results in opponent defenses guessing right more times than they guess wrong. Even though it appears that pounding away has it's benefits over the length of a game, I seriously question this strategy v. teams that are going to press what have appeared to be in the last 5 games at least, a sound defense and an improving offense.

When you balance history against future accomplishments, the future gets a little less clear; a little less hopeful and a bit more justifiably pessimistic.

[drok]

Odds, feelingsball, improvement, etc...none of it means anything until we see what happens the next three games.

[Jeff Buchanan]

Agree.

[Rocky Bleier]

I hate my right hamstring. Fucking hate it.

I'll take it and put it in my spare parts collection if you don't want it.

[hack]

Can you swap me one that's looser? Can pay in AAbucks.

[Jeff Buchanan]

Ya'll know I love this stuff but the data points I'm going to point out here demonstrate that the likelihood of M losing at least two of the remaining games among Wisconsin, MSU, PSU and osu is high without a change in the typical Harbaughffense run/pass percentages. (The link takes you to the tables page. Click on Michigan at the top among the list of schools to get M's stats)

While M is ranked #2 S&P+ Defense it's #25 ranking on offense is middling to slightly above average. You don't win championships with those numbers. Holding back M's offense among Connelly's stats is Opportunity Rate. This is the percentage of carries in which the offensive line "does its job" and produces at least five yards of rushing for the runner. (Generally speaking, the first five yards are considered the line's responsibility, the next five are split evenly between the runner and the line, and anything over 10 yards is all on the runner.) M ranks 112th here......and in Rushing Marginal Efficiency, M ranks #57. Is M going to beat Wisconsin, MSU, PSU and osu touting those kinds of numbers? I don't think so but that's just my data driven, facts based self.

There's another section among these data points called Bill Walsh Stats, Percentage of First Downs Coming on First or Second Down. That's 64% but ranks 113th nationally. Looking at the metric Opportunity Rate, you could say the OL's performance in the run game is not good. In fact, it screams it's not good in the face of decent box score rushing numbers. OTH, if you look at Opportunity Rate for rush plays (not good) and combine that with Avg. 3rd Down Distance to Go (5y/#4) and 3rd Down Success Rate (57.4%/#4), you come up with a statistical basis for what Harbaugh does - pound the ball. But is M scoring doing that. The answer is some but probably not enough to beat good teams.

Passing game numbers that measure efficiency/effectiveness are strikingly higher than the rushing numbers mentioned above: the marginal efficiency of the passing game is ranked #19 compared to that of the rushing game ranked #57. IMO, this screams for more passing that takes Harbaughffense from the pattern of 60/40 rush/run to something closer to 50/50 or 55/45. The data points also seem to scream that doing this on early standard downs will improve the percentage of First Downs Coming on First or Second Down from the pitiful rank of 113th, i.e., move the fucking chains and score, instead of 2-3 series of 30-40yd drives, don't score and then punt intended to set up one big play later that might score. Doing this should also improve the data point Finishing Drives as measured by Points per Scoring Opportunity (Rank #36/5.1 pts). It's this sort of thing that would be against M's trends to date and do a good job of keeping the opponent's defenses in the coming gauntlet of games guessing.

https://docs.google.com/spreadsheets...JugvRR/pubhtml

[Detroit Dan]

Those stats are amazing! Thanks for the link.

I do think you cherry picked a few stats and results from previous seasons. The bottom line, according to these stats, seems to be that Michigan is ranked in the top 10, behind Ohio State which is in the top 5, and perhaps a tad behind Penn State, but ahead of MSU and Wisconsin.

So, in my view, the glass is half full. Yes, the offense could be better, but it is so much better than last year that I'm not complaining. Your comments on Michigan's scheme seem overly verbose to me. Basically, you'd like Michigan to pass more. In most years I would agree, but this year I'm not complaining about that. With a defense as good as ours, we almost always have a chance to win, unless we screw things up on offense. That philosophy didn't work out so well last year, however, and changes were needed. So far, the offensive changes seem to be working.

To get to the next level, Michigan probably needs better recruits, not necessarily better schemes. Ohio State, Alabama, Clemson, and Georgia are in the top tier right now, but we're right behind them in the next group.

[Jeff Buchanan]

Well, I don't disagree with you ..... talent trumps scheme, about 98% of the time. The 2% is accounted for by badly coached defenses and one of those might be osu's.

http://www.detroitlionsforum.com/for...11#post1600511

The verbosity you assign to my post, which is an accurate assessment of it, essentially says pass more. It got verbose by making sure I had the numbers to back up that assertion. I am pretty certain that Harbaugh has analysts on his pay-roll that rival anything he might have had at SF. They know what the numbers say and what reasonable conclusions about scheme can be drawn from them.

One of those conclusions that can be drawn from the data points in the Connelly stats is that pounding the ball works. So, yes, I may not have cherry picked the data, I simply came to a reasonable conclusion about what the data points to ..... pass more.

In one of Mgo's UFRs, Brian pointed out that v. NW, and I think we'll see it again v. Maryland, passing plays increased in number. A trend? Maybe but Brian was quick to point out, with Harbaugh, it's not likely to get to 50/50. My position is that if passing on standard downs, and in particular first and second down, doesn't increase and by extension score points on each drive/possession, M is going to lose games that it has the talent and potentially the scheme variability that it could otherwise win.

Now, observers like us have to ask, are the current data points that might suggest M ought to pass more to win more reflective of (1) an effort on JH's part to conceal tools in the tool box or (2) reflective of the reality of practice time to devote to an expanding play book - layers of certain plays or (3) is JH just stubbornly stuck on man-ball concepts. I really don't know at this point.

What I feel reasonably confident in saying is that M's scoring efficiency as measured by these data points doesn't improve, winning the division isn't going to happen:

RED ZONE	21- to 30-yard line success rate		38.1%	90
	20 - to 11-yard line success rate		28.1%	113

I think the picture is pretty clear how Harbaugh wants to score assuming M's drives don't end up in punts outside, say, the 40 (reasonable FG range). He's going to run on first down. If he gets > 4 yards, 2nd down will probably be a run and so on. I think M needs more scheme variability on 1st down not just in the running game but it has to incorporate more passing. The opponent's defense is able to guess correctly too often what M is going to do play-by-play. That's why M has too many drives that aren't getting points and their red zone success rate is poor.

I'll grant we definitely saw some of that scheme variability involving passing v. Maryland. Hope we see more of it and I think we will as the offense gets opportunities to rep those things. Hope is the operative word.

[Detroit Dan]

talent trumps scheme, about 98% of the time. The 2% is accounted for by badly coached defenses and one of those might be osu's.

Shhh. I'm trying to get Talent overconfident.

The verbosity you assign to my post, which is an accurate assessment of it, essentially says...

ZZZZZZZZ. Oh, what? Right, good point.

My position is that if passing on standard downs, and in particular first and second down, doesn't increase and by extension score points on each drive/possession, M is going to lose games that it has the talent and potentially the scheme variability that it could otherwise win.

I agree that we will need to pass more often to win against the better teams. Often, however, that is dictated by score. If we're behind, we'll obviously need to pass to catch up. But if we're ahead, we should keep doin' what got us ahead, which is mixing in a lot of unpredictable plays including, especially, passes. So we agree for the most part I think.

Hope is the operative word.

Of course, and I agree that it would be foolish to be overly confident after what we've seen in recent years in the 2nd half of the season.

This B1G QB comparison at PFF is interesting. Most interesting is that Shea has been improving steadily and is now about even with Haskins, but significantly behind Hornibrook and McSorely.

[Detroit Dan]

And one more thing -- I disagreed with that punk Brian Griese about Shea's interception in the Maryland game. You've got to play to win, and that means hitting some tight windows a la Alex Hornibrook, and having some trust in your 6'8" receiver to win a jump ball. It doesn't always work out, but you've got to at least try. The TD pass from Shea to Bell was a thing of beauty -- risky, but worth it as it turned out. Kudos to HARBAUGH!!! for turning Shea loose.

[Detroit Dan]

Average Michigan points per game through the first 6 games, by year of Harbaugh tenure:

2015: 29
2016: 50
2017: 27
2018: 38

So the offense this year is much better than last year, but much worse than 2 years ago. Looking at the trajectory of the program, this year seems most similar to Harbaugh's first year, 2015, with the transfer QB and improvement as the season progresses.2015 was a good year, but this year we seem a bit better because of slightly better talent, and a more cohesive coaching staff. Brown is an improvement over Durkin, Ed Warinner is a good OL coach, and Harbaugh knows better what he has put together.2016 was Harbaugh's best team, but suffered from bad luck (spot in the OSU game) and one fatal flaw -- lack of QB depth. We seem to be on track this year, with vastly improved QB and OL play, while the defense continues to be very strong.

The bottom line is the scoreboard, not the opportunity rate and other obscure stats which may have value, but should be consistent with the bottom line if they are to have value. The scoreboard is doing much better this year, mainly because of the passing game. We already have more TD passes than we had all of last year, for example. So I am optimistic that when we play the better teams, the passing game will be better than in 2016 and 2017. It will need to be.

[Jeff Buchanan]

Fancy stats have predictive value that increases as the season progresses. In that way, they have some utility in comparing teams nationally.

I agree with you that the score is the most important thing. There is no question the team is improved by eyeball check and by any statistical measure you choose to use.

There's nothing wrong with the passing game. It's solid ...... when it's used.

There's nothing wrong with the running game. It's solid ......but IMO over-used and at the expense of the passing game.

You and I agree fundamentally. Where we have different perspectives is on the use of the passing game schematically. My view is opponent defensive scheme dependent. i.e., if an opponent deploys a zone defense as it's base and continually uses it's LBs and Ss conservatively at the LOS, run the ball. If they are playing aggressive man coverage as their base with Ss cheating up, throw it. Teams that M has played this season have been very aggressive 80% of the time on standard downs. My view is that JH through Pep hasn't punished that approach enough with the passing game. They'd prefer to pound it for various reasons that we've discussed here- that approach has it's merits. It seems to me you are fine with that and I am not. We can disagree.