regression...really?

Preach on

I don't think you'll find many people around here saying they have regressed from last year. Last year was bad bad bad bad, and this is no fun, but it's obviously better than last year. I think most people that claim regression would be interested if you would do the same study comparing '09s games 1-5 vs games 6-10.

Yes, most of the team has regressed. In their Stats class. And gosh darnit, if you were on the field thinking about the normality of the residuals instead of your cover assignment, you'd probably blow a few, too.

I say enough of this regression thing. It's costing us wins on the football field.

Last year, opponents often had much better field position due to the general awfulness of our offense (short drives and turnovers are contributors here). Therefore, it was easier for them to score points.

On the other hand, our rushing defense is clearly worse than it was (it went from middling to bad), while our pass defense remained bad. This makes sense, as our secondary sucks about as much as it did last year, our d-line is still good, but our linebackers have regressed, hence the increase in rushing yardage allowed.

But you can't tell most of these people anything because its like talking about religion. Emotions, not facts take over for most in these conversations/debates. With facts like this I think you chip away at the number of doubters little by little.

With all due respect, either you are misinterpreting the typical message board complaints, or this is a purposely disingenuous post. When people talk about "regression," they are generally talking about the team regressing from its performance in the first four games of the year, not from last year. People are upset that the team is getting worse, rather than improving, as the year goes on.

Aside from that, the analysis of the so-called defensive improvement is silly. Any slight increase in scoring defense is likely due to the improvement in our offense -- we are giving the other team less possessions than last year, so they score fewer points. The worsening numbers for the pass and rush defenses belie any suggestion that the defense has improved. A more accurate statistic would be points given up per possession, which I would wager is at least as bad as last year.

Last 5 losses, Michigan's defense has given up on average 435.2 yards and 33.4 points per game. Michigan has looked great against EMU, WMU, and Delaware St which has altered the stats.

Seth is right about the defense from this year to last years. Michigan's defense was hosed by their offense and the TO's in 2008.

While the Offense has shown great improvements in 09, they have still been very inconsistent. If we had the 08 Defense with the 09 Offense we would have 7 wins right now maybe 8.

Good post though our defense would be worse if it weren't for Baby Seal U. Maybe next year....

Well, it does seem like most of the Angry(!!!!11!!1!) people with DEMANDS are following the Psychological definition.

I wonder what callers "Dave from Phoenix" and "Linda from Grand Blanc" think about all of this???

.....is all too common from the shouters and the trolls.

5. Biology The return of a population to an earlier or less complex physical type in successive generations.

See it isn't #2 its #5

-Notice the the loss of a 5th digit off of both hands and feet
-The hair which is the obvious result of some toxic waste spill or massive dose of Radiation
-I have even heard from a good source that certain parts of their bodies glow in the dark though I haven't been able to confirm that myself.
-The conspicuous lack of knees or elbows, but the addition of a massive belly button and distended stomach
-Not to mention the lack of any reproductive organs which is actually a blessing as they seem to be devolving themselves into extinction.

So don't worry they will eventually all die off

I do think DSU needs to be taken out of comparisons between last year and this year. We did not play a team of such caliber last year and it throws off stats for comparison between the two years. I would imagine that taking similar opponents, we would see a slightly less rosy picture.

I responded to another comment on Brian's post, but people seem to be forgetting that between last year and this year we lost:

Morgan Trent (CB)
Terrence Taylor (NT)
Brandon Harrison (DB)
Will Johnson (DT)
Tim Jamison (DT)
John Thompson (LB)

The linemen in the list (and John Thompson) would have helped with rushing yardage, and may have given Brandan Graham more opportunity to impact the passing game. (This is nothing against Martin, Roh and Van Bergen. I think they've done a great job this year, and when they have a few more pounds on ...)

And then there is Morgan Trent ... As much as we dogged him after USC and App State, man we sure could use him this year.

I may be wrong, but I think the regression that the MGoBlog denizens have been talking about is the performance as of late (the last 5 B10 games) compared to the performance of the first 4 games of the year.

There is no doubt our team is better overall this year than last (although the last 2 games have been eerily similar to last year's), but we seem to be getting worse every week, which is frustrating.

That's still a good comparison though, and it gives a little more hope for the future.

I actually agree with the anti-Rich Rod people and the pro-Rich Rod people, and here's why. Anti: I do see regression over the course of the season. If the team played against Purdue and Illinois like they played against ND (i.e., poise, confidence, emotion), there is no way we lose those games. Even Tate said the team has been coming out flat in the 2nd half. That fire and drive has faded over the season, and it appears to be costing us games. Some of this has to be attributed to coaching, especially since we have seen it in the last several games. As for the Rich Rod supporters, I am officially one of those. I think the coaching staff has a long way to go, like the team. Unfortunately, they weren't able to produce the rebound we were all hoping for this year. But, given the youth on the team, the attrition and the injuries, they need at least one more season. With the huge number of returning starters next year, the bar will be set high, and rightfully so. Until the staff has a chance to reach that bar, I will withhold judgment and support my coach.

Sagarin rankings for the two seasons

2008:

Utah 5
Miami (OH) 157
ND 53
Wisconsin 61
Illinois 68
Toledo 134
Penn St 8
MSU 33
Purdue 78
Minnesota 75
Northwestern 44
OSU 14

2009 up to this point:

WMU 112
ND 32
EMU 179
Indiana 91
MSU 51
Iowa 16
Delaware St 189
Penn St 19
Illinois 90
Purdue 72

Games upcoming
Wisconsin 24
OSU 12

So it's almost like we replaced a top-5 team in Utah with Baby Seal U.

There's no doubt the offense is much better, even if you take out Delaware State. They've shown a lot more ability to throw the ball downfield this year, when last year it seemed like we never threw more than 15 yards the whole season. Running game is improved with QBs that can run/scramble.

Scoring defense is going to end up worse because I don't think we hold either Wisconsin or OSU to the same 26.4 average we have up to this point.

Not to mention the aggression of Rich Rod supporters, who seem to open every post in support of him with ad hominem attacks (and always have) on those who dare question the greatness that is Rich Rodriguez. Ahem.

The progression includes Delaware State. 'Nuff said. It also includes games against the completely farcical Eastern Michigan--arguably the worst team this side of Delaware State (I'm really not sure Eastern would beat DSU)--and again, by all accounts, worse versions this year of MSU, PSU, and Illinois.

You can shout all you want about progression and improvement; it's been conceded that the offense is more capable this year than lst. Wow. Last year was so goddamm bad that if the team didn't improve, Rodriguez should have turned in his headset and voluntarily resigned by now. Go ahead and compare this year's team to last year's, toss in all the stats and the weaker schedule (and many here have ALSO SAID CAPS ARE FUN that OSU is "worse" than last year--so how 'bout that?) and do whatever you can to justify the belief that the team is better--but they're not, really--it's a smoke and mirrors approach that covers up the fact that the team is still in trouble, and there isn't much help on the way besides another year of experience--the NEW battle hymn of the republic.

To remove the effects of the Delaware States of the world and the MAC, we should look at the Big Ten only Stats:

Rush offense decreased 12.1% from 153.5 to 135 yards per game. Somewhat surprising, but given the loss of Molk and Minor's traitorous ankle it can at least be somewhat explained.

Pass offense increased 43.9% from 137.4 to 197.7 yards per game.
Pass efficiency offense increased 29.3% from 90.5 to 117. Not DEATH.

Total offense increased 14.4% 290.9 to 332.7 yards per game. However, scoring offense only increased 7.7% from 22.1 to 23.8 points per game. The reason for that: Turnover Margin has doubled from -5 in 2008 to -10 in 2009. And we thought last year was bad.

Rush defense worsened by 11.1% from 172.1 to 192.1. Pass defense worsened by 7.9% from 230.8 to 249. Pass efficiency defense has actually improved 0.8% from 141.9 to 140.8. Total Defense has declined 9.3% from 402.9 to 440.2, while our scoring defense has improved 0.6% from 33.5 to 33.3.

I will take a deeper look at the numbers later, specifically yards per carry, sacks, penalties.

Also can anyone tell me how to put in a chart or table?

Taking the sequence of wins and losses this year alone, there is roughly* a 63% chance that the W-L record this year is not random.* If you base this analysis on our actual record, it suggests we've gone from good (1st 5 games) to bad (last 5). But this is not really valid, since the chances of winning each of these games was not equal. Thus we really need to look at the WL record relative to the point spread i.e., did we beat the PS or not? Based on this analysis, there's again a 63% chance that we've gone from good to bad. However, the "good" period was, more or less, the 1st 7 games of the season and the "bad" only the last 3 games. Why? Because the late season losses to MSU IA were actually wins vs the point spread (based on the score at the end of regulation); whereas, the early season win vs Ind was actually a loss vs. the point spread.

That is, actual WL recordL WM(W) ND(W) EMU(W) IND(W) MSU(L) IA(L) DSU(W) PSU(L) IL(L) PU(L)
or WWWWLLWLLL ie losses in 4 of last 5

WL record VS point spread WM(W) ND(W) EMU(W) IND(L) MSU(W) IA(W) DSU(W) PSU(L) IL(L) PU(L)
or WWWLW*WWLLL ie winning 6/7 games followed by 3 losses in a row.

Clearly, there are a lot of other; aspects of the question about whether we've regressed. But this analysis suggests that, just based on this season's WL record vs. pt spread, we may have regressed but the regression is only during the past three weeks.

*Based on a runs test for randomness. The 63% chance is based on the assumption of a normal distribution, which is not likely to be valid, so this is a very crude, approximate figure.

In 2008, Michigan ranked 104th--out 0f 119 teams, yes, I wrote it--with a turnover margin of -.83. I think we can all agree that this is not good and at least one of the reasons Michigan did so poorly last year.

In the first four games of the 2009 season--against opponents from near the bottom of the MAC, Notre Dame, and Indiana--Michigan was able to acheive a turnover margin of +.25. Sweet, a positive turnover margin that represents progress, that's a good thing. Had Michigan continued at that pace, Michigan would be ranked at approximately 52nd.

Sadly, that did not happen and--in spite of such a good start statistically--Michigan now sits at 106th with a turnover margin of -.9. So, yeah, that's regression. And not just a simple regression to the "less perfect" state of last year, but--in several ways--a regression beyond it.

First, the turnover margin is now statistically worse than last year's. Second, last year had inexperienced freshman who were being coached by an unfamiliar coaching staff committing turnovers. This year, we have sophomores who have been in the system for a year with the coaching staff commiting turnovers. Third,the coaches knew it was a problem last year and should have addressed it in the offseason.

So, the bottom line is that this year's team has fewer "excuses" for an unfavorable turnover margin and yet has a worse turnover margin than the 2008 team.

And, no, I do not hate Rich Rodriguez.