Like a lot of people in these parts, I felt a Michigan victory over Penn State was nearly a certainty prior to Saturday. And, like a lot of people, instead of enjoying the Halloween weekend comfortable with the knowledge that Michigan was at least going to a bowl game this year, I instead fretted about what exactly had gone wrong. Obviously, the defense is ridiculous, having managed to get worse despite the fact that the last two year's defenses were the worst in history and could only get better, but that's another story.

As I read through comments on the board, a common theme was that the offense is tremendous and that doing anything (like firing Rodriguez) to mess that up would be the worst decision that could be made. While I don't claim to know whether or not Rodriguez should be fired, I was having a hard time accepting the argument that the offense is fantastic. Somebody referred to the offense as #4 in FBS; I pointed out that while they are #4 in yardage, they're only #19 in scoring offense.  #19 is still good, of course, although that ranking has been dropping as they've been playing Big 10 teams that are not named Indiana.

I did some research trying to figure out why I felt compelled to argue with fellow fans who had nothing more obnoxious to say than that the Michigan offense is somewhere between really, really good and fantastic right now. I looked at the top 20 scoring offenses and tried to find something that would jump out at me as the source of my dissatisfaction. I didn't find it. The data (courtesy of ncaa.org) showed that Michigan doesn't score TDs as regularly as the heavyweights like Nevada, TCU, Boise State, and Oregon (all over 50%), but at a 42.4% TD rate, Michigan did rank 12th among the top 20. With 13% of UM drives ending in turnovers, they were pretty much middle of the pack (and ahead of Oregon -- the gold standard in my book -- at 13.3%), and having 4.3% of their drives end in failed fourth down conversions was again more or less middle of the pack.  I threw third and fourth down efficiency into the mix to see if something stuck, but it didn't. Michigan's numbers don't look drastically different than the other top scoring offenses.

	TDs	FG Atts	Punts	Tos	TO on Ds	Total		TD %	FG Att %	Punt %	TO %	TO on D%		3D%	4D%
Nevada	45	7	16	13	4	85		52.9%	8.2%	18.8%	15.3%	4.7%		59.4%	75.0%
TCU	50	6	27	8	4	95		52.6%	6.3%	28.4%	8.4%	4.2%		53.9%	63.6%
Boise State	44	10	17	9	5	85		51.8%	11.8%	20.0%	10.6%	5.9%		46.8%	44.4%
Oregon	58	7	28	15	5	113		51.3%	6.2%	24.8%	13.3%	4.4%		49.1%	64.3%
Stanford	43	14	16	13	2	88		48.9%	15.9%	18.2%	14.8%	2.3%		57.7%	81.8%
Utah	48	10	25	14	2	99		48.5%	10.1%	25.3%	14.1%	2.0%		54.3%	75.0%
Houston	45	9	22	16	3	95		47.4%	9.5%	23.2%	16.8%	3.2%		56.8%	62.5%
Ohio State	48	16	30	11	1	106		45.3%	15.1%	28.3%	10.4%	0.9%		44.3%	85.7%
Wisconsin	37	12	25	6	2	82		45.1%	14.6%	30.5%	7.3%	2.4%		51.7%	66.7%
So Cal	40	9	22	14	5	90		44.4%	10.0%	24.4%	15.6%	5.6%		52.0%	66.7%
Auburn	45	19	25	12	2	103		43.7%	18.4%	24.3%	11.7%	1.9%		50.0%	60.0%
Michigan	39	9	28	12	4	92		42.4%	9.8%	30.4%	13.0%	4.3%		46.5%	69.2%
Hawaii	45	17	30	15	4	111		40.5%	15.3%	27.0%	13.5%	3.6%		37.9%	55.6%
Nebraska	39	11	33	12	2	97		40.2%	11.3%	34.0%	12.4%	2.1%		39.8%	66.7%
VA Tech	37	14	30	9	3	93		39.8%	15.1%	32.3%	9.7%	3.2%		42.1%	50.0%
OK State	45	16	35	16	5	117		38.5%	13.7%	29.9%	13.7%	4.3%		42.6%	44.4%
Oklahoma	37	8	42	8	5	100		37.0%	8.0%	42.0%	8.0%	5.0%		45.8%	58.3%
East Carolina	38	11	36	16	2	103		36.9%	10.7%	35.0%	15.5%	1.9%		44.4%	84.6%
Tulsa	37	19	34	11	2	103		35.9%	18.4%	33.0%	10.7%	1.9%		49.2%	71.4%
Arkansas	36	9	35	16	8	104		34.6%	8.7%	33.7%	15.4%	7.7%		39.2%	42.9%

I thought of Denard's interceptions in the redzone and threw redzone efficiency into the mix, but that really just makes Michigan look better. They are only 43rd nationally with an 85.3% success rate (meaning TD or FG) in the red zone, but their TD percentage of 76.5% ranks 6th, well ahead of most of the top 20 scoring offenses. (Data below with the unwashed masses removed and only the top 20 scoring offenses depicted)  

Name	Gm	Drives	Scores	Points	Rush TD	Pass TD	FG	Pct	TD PCT
Wisconsin	8	41	37	242	25	8	4	90.2%	80.5%
TCU	9	46	41	265	30	6	5	89.1%	78.3%
Michigan	8	34	29	190	19	7	3	85.3%	76.5%
Southern California	8	32	28	183	10	14	4	87.5%	75.0%
Arkansas	8	27	25	154	10	10	5	92.6%	74.1%
East Carolina	8	34	32	195	11	14	7	94.1%	73.5%
Utah	8	39	35	216	19	9	7	89.7%	71.8%
Boise St.	7	39	34	215	18	10	6	87.2%	71.8%
Nevada	8	44	37	234	24	7	6	84.1%	70.5%
Oklahoma St.	8	36	35	205	14	11	10	97.2%	69.4%
Stanford	8	50	46	273	18	16	12	92.0%	68.0%
Houston	8	43	36	223	19	10	7	83.7%	67.4%
Oklahoma	8	43	35	220	15	14	6	81.4%	67.4%
Ohio St.	9	51	45	272	18	16	11	88.2%	66.7%
Oregon	8	42	37	226	21	7	9	88.1%	66.7%
Auburn	9	44	39	229	19	9	11	88.6%	63.6%
Virginia Tech	8	38	34	198	16	8	10	89.5%	63.2%
Nebraska	8	24	19	117	12	3	4	79.2%	62.5%
Tulsa	8	42	35	201	14	10	11	83.3%	57.1%
Hawaii	9	46	37	211	10	15	12	80.4%	54.3%

To cut to the chase, I wound up focusing solely on the last three games. I took ESPN's drive charts and just added the score at the time the drive started. That information is below:

OPP	START	QTR	POSS.	YARD	PLAYS	YARDS	RESULT	SCORE
MSU	14:55	1	3:23	MICH 25	9	65	Interception	0-0
MSU	7:14	1	5:49	MICH 10	13	73	Field Goal Good	0-0
MSU	14:05	2	1:52	MICH 12	3	9	Punt	3-0
MSU	11:28	2	3:21	MICH 40	9	60	Passing Touchdown	3-7
MSU	4:23	2	1:33	MICH 15	3	1	Punt	10-14
MSU	0:23	2	0:23	MICH 20	2	55	Field Goal Missed	10-17
MSU	12:32	3	2:25	MICH 34	7	58	Interception	10-24
MSU	4:55	3	1:22	MICH 36	3	3	Punt	10-31
MSU	1:28	3	1:32	MSU 42	8	41	Rushing Touchdown	10-31
MSU	13:16	4	0:53	MICH 20	3	11	Interception	17-31
MSU	7:14	4	1:33	MICH 25	3	1	Punt	17-34
Iowa	12:59	1	4:41	MICH 25	13	75	Passing Touchdown	0-0
Iowa	7:20	1	1:50	MICH 20	3	1	Punt	7-0
Iowa	1:21	1	1:42	MICH 28	5	17	Interception	7-7
Iowa	13:22	2	4:49	MICH 20	12	59	Field Goal Missed	7-14
Iowa	4:15	2	3:58	MICH 7	10	48	Punt	7-21
Iowa	14:55	3	1:45	MICH 23	3	4	Punt	7-21
Iowa	11:18	3	4:24	MICH 16	12	71	Fumble	7-21
Iowa	5:12	3	0:42	MICH 35	3	5	Interception	7-21
Iowa	1:45	3	3:40	MICH 15	12	85	Rushing Touchdown	7-28
Iowa	11:37	4	1:09	MICH 25	4	75	Passing Touchdown	14-35
Iowa	8:08	4	1:13	MICH 30	6	69	Rushing Touchdown	21-35
Iowa	2:47	4	0:59	MICH 37	3	-9	Interception	28-38
PSU	15:00	1	1:42	MICH 28	3	8	Punt	0-0
PSU	7:26	1	3:25	MICH 20	9	80	Rushing Touchdown	0-7
PSU	1:24	1	1:38	MICH 29	6	32	Punt	7-14
PSU	13:12	2	4:29	MICH 25	15	55	Field Goal Good	7-14
PSU	3:27	2	0:58	MICH 2	3	2	Punt	10-21
PSU	1:02	2	1:02	MICH 27	4	9	Turnover on Downs	10-28
PSU	9:57	3	1:59	MICH 20	5	80	Passing Touchdown	10-31
PSU	3:59	3	2:23	MICH 48	8	53	Rushing Touchdown	17-38
PSU	13:21	4	3:56	MICH 29	11	72	Rushing Touchdown	24-38
PSU	5:43	4	0:59	MICH 26	4	2	Turnover on Downs	31-41

Breaking that down by game situation (e.g., Michigan leads, game tied, Michigan down by one score, etc) yields:

Michigan leads
Number	TDs	FGs	Missed FGs	TOs	TO downs	Punts	3 and out
2	0	0	0	0	0	2	2
Score tied
Number	TDs	FGs	Missed FGs	TOs	TO downs	Punts	3 and out
5	1	1	0	2	0	1	1
Michigan trails by one score
Number	TDs	FGs	Missed FGs	TOs	TO downs	Punts	3 and out
7	2	1	2	0	0	2	1
Michigan trails by two scores
Number	TDs	FGs	Missed FGs	TOs	TO downs	Punts	3 and out
11	2	0	0	5	1	3	2
Michigan trails by more than two scores
Number	TDs	FGs	Missed FGs	TOs	TO downs	Punts	3 and out
8	5	0	0	0	1	2	2

Conclusions?

• In those (rare) instances when Michigan had a lead, the ball, and a chance to potentially take charge, they failed to move the chains at all.
• When starting drives with the score tied, Michigan has managed TDs 20% of the time and has committed turnovers as often as they've scored.
• When down one score the TD percentage picks up a bit, and missed FGs have hurt, although neither of the misses would have tied the game or given Michigan a lead (they trailed by 7 both times).
• When down two scores and needing to score to keep it close/get back in the game, the wheels come off, with turnovers happening much more often than scores. Witness the Iowa game: down 21-7, four consecutive Michigan drives occur with a chance to close to within 7.  The results?  Punt, punt, fumble, interception.
• In 25 drives where the score delta was somewhere between +7 and -14, the offense scored TDs on just 20% of those drives (less than half the season average of 42.4%), tried FGs on 16% (converting only half ot those), committed turnovers on 28% of those drives, and punted on 32% (with 75% of those punts coming after 3-and-out).
• It is only in the 8 drives that started when Michigan trailed by more than two scores that the offense really shined, managing TDs on 5 of those drives to go along with 2 punts (both after 3-and-outs, unfortunately) and one failed fourth down conversion.
• Since Tate Forcier was the QB for 3 of our TD drives in the last three games, that means the offense as led by D Rob has only tallied 7 touchdowns.
• Going 3-and-out on nearly 25% of our drives in the last three games *feels* excessive to me, but I haven't yet tried to determine how often that actually happens to other "elite" offenses.

In that data, I think I discovered the source of my discontent. Yes, the offense has put up pretty good numbers against MSU, Iowa, and PSU, but the fireworks didn't really start until we were desperately trying to mount a comeback. The offense has been sputtering when games have been close.

In the board topic where I first tried to present this, there were suggestions that examination of the failed drives would lead to extenuating circumstances, and given the relatively small sample sizes, that is a possibility. I've pulled the UFRs for the MSU and Iowa games but haven't yet tried to assemble them into this analysis.

I put most of this together while I was (supposed to be) working today, so I won't be surprised if there are some errors, although I'm confident the numbers are largely correct.  Like I said, I looked into this just to try to help me understand why I wasn't freeling very impressed by the offense despite the gaudy yardage statistics, threw it into a board topic to back up another poster's assessment, and moved it to a diary upon request. Also, in case it's not clear, this is not a suggestion that Rich Rodriguez should be fired, that Tate Forcier should be starting, that the spread can't work in the Big 10, etc. Rather, it's just an attempt to help those who are convinced the offense is spectacular understand why other Michigan fans (who also bleed maize and blue) are somewhat disappointed by what the offense has done in our three losses.