Fremeau Efficiency numbers are out, Michigan ends regular season at #32.
It's been a season with more than its share of end-of-game dice rolls, and instead of arguing about which wins could or should have been losses and which losses could or should have been wins I thought it might be worth turning to the finer-grained drive-by-drive analysis by the smart folks at Football Outsiders.
The new FEI numbers are out: Michigan is #26 on defense, #38 on offense, #32 overall.
That is not a typo; I did not mistake an 8 for a 3. Michigan finished the regular season #38 in offensive FEI and the offense on the whole was only marginally worse than the defense. In fact the relative positions of the offense and defense weren't much different than 2012; both were down from last year by about the same amount. (In the index itself, not in the ranking.)
Something curious I noticed for the first time today: on a drive-by-drive basis Michigan's defense was #26, but on a play-by-play basis they were #48. I wonder if that's related to Mattison's emphasis this year on "keeping the ball in front and inside", not giving up big plays and forcing the offense to grind out a drive? They get a lot of successful plays, but things go wrong before they get to the endzone? There's probably a diary here if I ever find the time to look at it.
Some historical context might be useful. The database only goes back to 2007, unfortunately--I'd really like to have some Carr seasons that weren't marred so badly by key injuries. But here's how the seven Michigan seasons in the database stack up. Note that in the overall and offensive FEIs positive numbers are good, but on defense negative numbers are good. In all three metrics, zero is the FBS mean.
Overall:
Season |
FEI |
FBS rank |
B1G rank |
2011 |
.202 |
9 |
2 |
2007 |
.141 |
22 |
2 |
2012 |
.133 |
26 |
5 |
2013 |
.118 |
32 |
5 |
2009 |
.007 |
59 |
9 |
2010 |
.005 |
55 |
6 |
2008 |
-.036 |
70 |
9 |
Offense:
Season |
oFEI |
FBS rank |
B1G rank |
2010 |
.652 |
2 |
1 |
2011 |
.472 |
9 |
2 |
2012 |
.267 |
25 |
3 |
2013 |
.237 |
38 |
4 |
2007 |
.001 |
56 |
7 |
2009 |
-.027 |
66 |
8 |
2008 |
-.167 |
81 |
9 |
Defense:
Season |
dFEI |
FBS rank |
B1G rank |
2007 |
-.477 |
11 |
1 |
2011 |
-.414 |
16 |
2 |
2012 |
-.338 |
26 |
5 |
2013 |
-.299 |
26 |
4 |
2008 |
-.111 |
44 |
6 |
2009 |
-.005 |
60 |
8 |
2010 |
.419 |
109 |
11 |
Would anyone have gueesed that the '09 defense would grade out better than the offense?
I suspect we're going to see some bashing of Fremeau's method in the comments. I hope it's well-informed; I'd love to know more about how he calculates these.
December 3rd, 2013 at 12:43 PM ^
December 3rd, 2013 at 12:45 PM ^
I don't really know Fremeau's method, I just think this shows the "danger" of taking one statistical model and reading too much into its results.
December 3rd, 2013 at 12:49 PM ^
too much, other than the fact our offense has gotten consistently worse under Borges
December 3rd, 2013 at 12:57 PM ^
its like in 2011 we were awesome, and then in 2012 we lost a rimington award winning center, junior hemingway, and denard got hurt and then in 2013 we lost 3 senior interior oline men and had to start freshmen. almost like every year is different because you have new unknown players and circumstances....
December 3rd, 2013 at 1:13 PM ^
because I'd love to break 2012 into pre- and post-injury. One of the things this chart screams at me is DENARD. He was even more valuable than I realized.
December 3rd, 2013 at 5:18 PM ^
it's almost like the defense has also gotten worse each year. it's almost like the roster has gotten younger each year. it's almost like michigan's roster has a huge structural problem that is going to take at least a full cycle to repair.
but who are we kidding, that's all crazy talk, right?
December 4th, 2013 at 11:06 AM ^
True, but I'd counter that we also added Minitron, one of the best RB commits in the country, still had two NFL-quality OTs, and a much better QB at throwing the ball and can do a reasonable facsimile as a runner.
December 3rd, 2013 at 12:58 PM ^
To be fair, according to that chart, so has our defense under Mattison.
December 3rd, 2013 at 2:18 PM ^
matriculating through the program, or attriting out as the case may be. When they're second years whatever you get is sort of a bonus. But by now they're supposed to be the heart of the team, and what do we have?
- Jackson
- Dileo
- Gardner
- Black
- Ash
- Ryan
- Furman
- Avery
That's everyone that's even on the depth chart. Some fine players on there, but it's a pretty sad haul for an entire recruiting year.
December 3rd, 2013 at 12:50 PM ^
Anybody, and I do mean ANYBODY, who actually watched the 2009 team play would know there is no way in holy hell the D was "better" than the O. The fact that this analysis somehow suggests that it was just shows how flawed the conclusions they draw from their data set must be.
This doesnt come close to passing the eye test IMO.
December 3rd, 2013 at 12:59 PM ^
Neither side of the ball was good that year. We had some good outings out of conference, but In Big Ten play our ppg and yardage were barely different from 2008.
December 3rd, 2013 at 1:01 PM ^
Since this equalizes to 0, the indexes are relative to the year. It is quite possible that our defense was worse than the offense in 2009, but relative to other defenses, not quite as bad. If everyone is giving up 400 yards a game, giving up 450 isn't as bad as if everyone else were giving up 350.
December 3rd, 2013 at 1:27 PM ^
I watched every game, and while I agree the O was better than the D, it wasn't by much.
That team failed to adapt, progress, or score in BIG play.
We talk about half time adjustments (or lack there of) from the current staff. But the RR offense never seemed to play well in the second half of any conference game.
December 3rd, 2013 at 1:41 PM ^
The D wasn't necessarilly better than the O. It's just that other teams had poor defense also so Michigan's ranked up better.
December 3rd, 2013 at 12:57 PM ^
Should have been a diary.
December 3rd, 2013 at 1:05 PM ^
These aren't quite final numbers; there's a bowl game to play. I'll try to remember to put up a diary in January.
December 3rd, 2013 at 1:08 PM ^
It is a shame that the database does not go back farther actually - the NCAA archive actually has some of the summary information displayed here, such as FBS rank. For that at least, we can go back into the Carr years:
For fun, we'll just go back to 2003. For total offense and defense for Michigan, per the NCAA archives:
2007 - #68 offense, #24 defense
2006 - #38 offense, #10 defense
2005 - #55 offense, #36 defense
2004 - #46 offense, #33 defense
2003 - #15 offense, #11 defense
December 3rd, 2013 at 1:17 PM ^
Among other things, FEI is tempo- and schedule-adjusted.
December 3rd, 2013 at 1:48 PM ^
I know - that's why I sort of wish Football Outsiders would embark on a project to take their ratings back at least another five years so most teams could see the variation over a more meaningful period (and indeed, over a few coaching regimes in a lot of cases).
At least for me, I think one of the advantages to these numbers is that they've always seemed to illustrate just how nebulous the boundaries of coaching changes can be when compared to performance. That's just one things I've seen in them though.
December 3rd, 2013 at 1:14 PM ^
http://www.footballoutsiders.com/stats/ncaadef
Other stats are also goofy:
S&P Defense - 48
S&P Offense - 41
S&P Overall - 41
F/+ - 35
December 3rd, 2013 at 1:38 PM ^
Defensive rankings of the weaker teams on the schedule:
- CMU #111
- Indiana #110
- Akron #88
- UConn #55
Connecticut was awful, but it was the offensive side of the ball that was the tire fire (they were #114, even worse than Purdue). That defensive ranking slots them right in between #54 LSU and #56 Georgia. It was a decent defense, handcapped by bad field position given up by their offense.
Worth remembering, when you evaluate that game.
Weirdly, Michigan played seven teams ranked between #25 and #43. MSU was #3.
December 3rd, 2013 at 2:34 PM ^
Connecticut gave up the following point totals in games this season:
33 (Towson!), 32, 41, 41, 62, 31 and 38.
This is what I will remember when evaluating that game.
December 3rd, 2013 at 2:46 PM ^
Stanford and Arizona State are the #3 and #4 teams in the country? That's about all I need to see in relation to this ranking system.
December 3rd, 2013 at 3:22 PM ^
We have been really bad in the Special Teams FEI in the last few years.