The Greatest Wisconsin Scoring Debate: Whom do the stats support?
This issue has been discussed ad nauseum. Certain posters, myself included, think that our offense performance against Wisconsin was not as impressive as the numbers indicate, since we were shut out in the first half and only put up points when we were already down by 21. Others, including our illustrious leader, think the above position is idiotic, as obviously Wisconsin was not easing up on us. In his most recent post, Brian mocked those of us in group 1 for just going with our "feelings, man" instead of presenting any hard data.
But, unless I'm missing something, I've never seen any "hard data" supporting the second theory, either. Instead of just dismissing the opposing argument as an unsupported "gut feeling," shouldn't people claiming that it's just as hard to score when you're down by 21 as when you're tied actually present some data to that effect?
I believe in statistics. I'm not some cantankerous Murray Chass who believes that my gut feeling is more important than newfangled math. At the same time, I acknowledge my limitations -- I wouldn't know where to begin in gathering the data to demonstrate whether or not defenses "let up" when they're up big. (If I did, this would be a diary, rather than a board entry). But I would love to know whom advanced statistics actually support in this area, rather than simply being talked down to by people who claim I don't understand math, but don't present any numbers in their argument. Anyone aware of any actual statistical analysis in this area that addresses the question at hand?
I think you can say that until you are blue in the face...
Well for 1 how about something along the lines of Hoke never being hired for the job? I think that qualifies.
Edit:
Some other gems include Denard will definitely transfer, the recruiting class will be a disaster, we won't be able to attract good assistants...
show that some good things happened under RR and using FEI to justify that UM had the second best offense in the country with only FGA, red zone randomness, and fluky TOs, all not related to the offense, being what pulled the team down.
Anyway, being out here in the bay area, I saw lot of Stanford and Oregon games, along with some BSU and Nevada games, and those four teams along with Auburn had better offenses than Michigan.
yeah, we lost the fucking game. These threads are useless. All the stats in the world can't capture the intangibles like momentum, how the teams/coaches respond to a 3TD deficit/lead, substitutions, etc. Fuck 2010...that was so last year...we got fall camp in 20+ days. Hoke will be pointing all over and destroying stuff.
Last season was soooo 2010.
The reason people like Fuzzy Dunlop keep beating this dead horse is because they are undoubtedly right and can't stand Brian and his cult followers suggesting otherwise. Bad news, MgoBlog nation. We SUCKED last year. That offense did not get it done against top tier B1G teams, until the game was out of reach.
If you're stats tell you otherwise than either (1) your model is not sophisticated enought or (2) your interpretation of the data is wrong.
This like the Rich Rod debate. No matter what excues you make because Brian liked him from the beginning, it doesn't change the fact that homeboy didn't get it done.
I love the site, but Brian derided Hoke and the dude is hauling in a sick class and unifying the program. I FEEL like he's going to win and will soon back-up that up with stats about wins and losses.
GO BLUE
I think sucked is too strong of a word. I think people are latching onto what Denard did and projecting how amazing he was onto the rest of the offense. They act like RR had delivered us a finished offensive product that was the best thing ever when that isn't the truth. We had an amazing sophmore QB, a solid offensive line and not much else. The WRs were inconsistent. One game Stonum or Roundtree would be unstoppable, the next game it seemed like a constant stream of drops. I don't think any of our WRS really made any DBs afraid of 1v1 like Edwards and Mannginham could. On the RB side we had no analog to Slaton/Devine. Assuming of course we got Dee Hart we might have found ourself one but keep in mind that meant we'd using a true freshman as our primary back. I'd argue though that after QB, the RB is most important focus of RR's offense and he shoud have had a guy like Slaton or Devine in place early on, not in this fourth year/third full recruiting class.
I think the offense was really still a couple years away which really wasn't all that great given RR had been there three years. Couple that with the fact the defense was a black hole of doom, RR got fired.
A word of warning to anyone actually attempting to do the study Fuzzy suggests.
I think there would be a great deal of noise that might have to be filtered out especially with the talent disparity in college football. Wisconsin's defense could let up considerably and still be able to throttle a significant portion of their schedule. Michigan most probably represents the best team Wisconsin held in check for a half. If there was a let down, Michigan's offense would be the first and most able to take advantage of it.
In other words, you could demonstrate that was no correlation between score and Wisconsin's defensive performance over the course of the entire season and still not have answered the question. Also, there probably were games where Wisconsin took out key starters which would certainly lead to a decrease in performance but prove nothing about the Michgigan game.
More importantly, is there ANYONE who actually doubts that when a team gets up big they may call of the dogs either by relaxing their intesity , through play calling, or substitutions?
If someone wishes to contend that then I invite them to step forward. Otherwise, we don't really have an argument.
More importantly, is there ANYONE who actually doubts that when a team gets up big they may call of the dogs either by relaxing their intesity , through play calling, or substitutions?
I think the more fair statement would be, "is there anyone who actually doubts that when a team gets up comfortably they may call of the dogs..." And I think everyone would agree with that (unless we're talking about old school Spurrier). but when Michigan scored to pull within 24-7 with 26 minutes left in the game that Wisconsin felt like its lead was safe and the game was over? And yet Michigan scored on its next two drives, both against the first team D. That wasn't because Wisconsin let up. It was because Denard = Denard.
Although I do agree with you. There's very little way to do statistical analysis of this kind of thing; small sample size + much noise.
I don't think Bielema let up significantly until well into the fourth quarter.
I think that we are putting too fine a point on this debate. However, here is the scoring progression.
24-0
24-7
24-14
31-14
31-21
38-21
41-21
41-28
48-28
What I see is Wisconsin at 24-0 relaxing. Michigan getting a bit of life and then Wisconsin dropping the hammer again.
This is not the game I would use to make any argument about Michigans offense. I would use Iowa or Penn State. I have no idea why Fremeau and Cook are fixated on this game. The Wisconsin game was never a game. The only conclusion that I can draw from it is Wisconsin was way better than Michigan; so much so that Michigans defense would not have been able to stop Wisconsin if they were told the play ahead of time.
Wisconsin held OSU's offense in check in the first half too. Statistically OSU's offense isn't that different than UM's. OSU averaged 50 yds less/game (UM was first in the big ten in that catagory and OSU 2nd) but OSU averaged 6 more pts/game than UM.
I've been reading this debate and am actually a little surprised it even exists. Wisconsin played a certain game to get a 28-0 first half lead. All the evidence I need to know they changed their philosophy at that point was their drive at the end of the first half, in where they had the ball in the red zone with under a minute left and did something they would never, ever do at that spot on the field if they were really putting a premium on points: they threw the ball into the endzone for an interception. As effective as their passing game was last year, they would normally just pound it in from that spot, especially against a defense that wasn't really able to stop the ground game at all. That's exhibit A that they let their guard down a bit and lost discipline.
it is also 100% accurate to say they were content to trade scores in the second half. It's precisely what they did against OSU as well. They got a big lead and were fine giving up points as long as (1) the drives took time and (2) they could answer with scores when they needed to. This happened all year long. Because of the cushion, teams did not have time to catch up so long as they didn't turn it over, the defense kept everything underneath and the offense could answer with scores of its own. Remember, one of Michigan's second half scores came on a very short field after a Wisconsin WR fumbled after making a catch for a first down. Bielema saw that and realized he could only lose by playing as aggressively as he did in the first half. He shut it down. I can think of a great throw and catch that got behind our defense, other than that they gave Denard exactly what they gave Pryor: 5-7 yard runs that took time off the clock.
This is also why we ran the ball every single play, save one (the fumble), in the second half. The strategy was to just take time off the clock while maintaining a cushion on offense and make Michigan work for their scores so there was no time to mount a comeback. Wisconsin's offense was actually pretty balanced last year. They were good at throwing it out of play action. The fact that they never did this in an entire half should tell you they had focused on sewing up the win in the second half, both offensively and defensively.
Put it this way: I have a vested interest in beating Michigan. In a few recent Wisconsin wins (and one disastrous loss) that got close in the second half I threw things and behaved in such a way that on one such occasion my wife took my daughter and her playdate shopping because I was scaring them. At no point during last year's game was I nervous. The way they were playing fit a blueprint I had seen before and I knew the game was totally in control. That's not statistical evidence, but the "look test" can't be dismissed altogether either.
None of this is to say that Michigan's offense wasn't good last year. It was. With all due respect, however, I don't think it was great. When you average fewer than 7 points in the first half of every game against a decent opponent (PSU, MSU, Iowa, WIS, OSU and MSU), only to come back with points once the game was out of reach, that is a statistical trend. You can't discount the notion that your second half points were the product of a changed mentality by the opposing defense.
As an opposing fan, Michigan scares me a lot more with a talented running back running counters behind a bruising offensive line, a drop back NFL caliber QB throwing to tall speedy receivers and a good tight end, and a 4-3 defense with 300 lb pluggers at tackle and 220 lb safeties. Denard carrying the ball every play and a "light and quick" defense? Not so much. That's nothing against Denard. He's a rare talent. It's like Drew Brees, arguably one of the 5 greatest QBs to play in our conference in the last 20 years. I realize his version of the spread was different than what Rodriguez did, but I still remember them moving the ball up and down the field at will against us, only to run a half back option pass from inside the 5 that was promptly picked off. Those teams never did much against Michigan's physical, athletic corners. The spread relies on maximizing space. When you get close to the goal line the amount of space shrinks. You can beat the spread. On the other hand it's difficult to beat a team that can pound the ball, throw it out of play action, play defense and control the clock, especially a team that can recruit kids like Michigan can. Mixing in other elements is fine, but I do not think you can have sustained success without that physical element as part of your game.
In sum, without having seen what Brady Hoke can do I'd say getting back to Michigan's traditional principles and tweaking them is the right idea. It was never really broken, was it? Maybe in need of some paint but not a down to the studs renovation. I know that has been debated ad nauseam here but that's an outsider's take.
Sorry for the length . . .
Wisco fan thinks his team mostly "let up" and not that UM's offense was "that great."
Why would any Wisco fan argue to the contrary?
While he will find support in people in one camp about our offense, it is probably the least unbiased source to add to this conversation that started out looking for data/stats/evidence.
(P.S. - he may also want to add in the future that Wisco "cooled" on recruits that UM got over Wisco and that the refs totally blew any call that "cost" Wisco a game.)
You think any OSU, Michigan State or Mississippi State fan is going to argue to the contrary of that either? Heck, after the last 2 years an Iowa fan would probably say Tate Forcier gave them more problems that Denard - in 2009 Tate played the first 3 quarters and part of the 4th before Denard took over and in 2010 Denard played about the first 3 quarters before Tate took over. Denard led UM to 14 points and Tate led UM to 42 pts.
Good lord. How whiny.
Why did you put "letting up" in quotes? That phrase doesn't appear anywhere in what I wrote and it's not at all what I was arguing. I never claimed Wisconsin "let up" or didn't want to score there. I have no doubt they were trying and wanted points, otherwise they would have taken a knee. Their general tendency in that situation, however, would have been to go to their strength and run the ball, which was working rather well, and at the very worst come away with 3 points. Instead they did something uncharacteristically riskier and came away with none. I guarantee you their OC was angry at himself for getting sloppy with a big lead and would have freely admitted to making a call he would not have made in a closer game.
What about a basketball player who throws a behind the back pass or tries for a reverse dunk with a big lead. Is he letting up? No. But his willingness to try something he wouldn't feel comfortable trying in a closer game is indicative that his mental approach with a huge lead is different than it was at 0-0. And that might be relevant in a debate about the defensive ability of the player guarding him at that point in the game.
All of this by the way, was just one of several examples I offered from the perspective of a fan of the opposing team in question to show that I thought UW's approach to the game might have changed. Does that mean it was easier for Michigan to score? Who knows. It's all just good natured speculation among people who follow this stuff more closely than we should but will never know what really goes on in the heads of those who play the game.
But this is all "flamebating" and "extremely irrational."
So this thread is still going? Anyone else enjoying the nice summer Friday (offer not valid to Southern Hemisphere MGObloggers)?
The funny thing is, over a 100 posts in and there hasn't been any real effort to answer the main question I was asking:
In general, is it easier to score when you are down by 20+ points than when you are tied?
Perhaps what is funny is your definition of "real evidence." I have seen many reasons given in this thread as to why it is easier to score when down by 20+ points. Its just that they are not statistical. If you understand statistics, you will understand that it takes very sophisticated statistics to measure relationality, interrelational processes, dependent variations, mutual processes of becoming, and the like. Those processes as embedded in the Wisconson game have been described here ad nauseum. I suggest adopting a redefinition of "real evidence", and your world will be opened.
I do'nt think the relative score matters as much as the quality of the opposing defense for the most part. Play calling my be affected, but think about it, if you're down 60-40 in the 3Q it's probably going to be easier to score than tied 3-3 in the 3Q, but the converse is that if you're tied 40-40 in the 3Q it's going to be easier to score than down 23-3 in the 3Q. Given being an equivalent amount of time into a game, you ability to score going forward is best expressed by looking at the success you've had up to that point in the game. How many points your defense has given up is mostly irrelevant unless you assume the other team is putting in weaker players or conceeding yards for time, which isn't really something you can do against a spread team in Jet tempo.
Here's my minor contribution. I went through and averaged the points scored and given up for every UM opponent last year (excluding their game against UM and FBS opponents). Then I compared how UM did against those teams with the expected (average) points for and against for each team all year.
| Score | Opp Expected Points | UM Performance | ||||
| Opponent | UM | Opp | For | Against | Offense | Defense |
| Uconn | 31 | 10 | 24.6 | 23 | 8 | 14.6 |
| ND | 28 | 24 | 26.5 | 19.6 | 8.4 | 2.5 |
| BG | 65 | 21 | 21.2 | 30.7 | 34.3 | 0.2 |
| Ind | 42 | 35 | 24 | 34.9 | 7.1 | -11 |
| MSU | 17 | 34 | 27.6 | 24.2 | -7.2 | -6.4 |
| Iowa | 28 | 38 | 27.4 | 16.9 | 11.1 | -10.6 |
| PSU | 31 | 41 | 21.3 | 23.9 | 7.1 | -19.7 |
| Ill | 35 | 35 | 29.4 | 21.4 | 13.6 | -5.6 |
| Pur | 27 | 16 | 18.9 | 29.4 | -2.4 | 2.9 |
| Wisc | 28 | 48 | 38.3 | 21.5 | 6.5 | -9.7 |
| OSU | 7 | 37 | 38.9 | 14.9 | -7.9 | 1.9 |
| Miss St | 14 | 52 | 24.5 | 19.4 | -5.4 | -27.5 |
| Average | 29.41667 | 32.5833 | 26.88333333 | 23.316667 | 6.1 | -5.7 |
I adjusted the +/- to make all + numbers good for UM and - bad. So on the whole, UM scored 6.1 points per game more than their opponents had given up on average while allowing 5.7 more than the opponents had scored on average.
I think on the whole, this shows that UM's offense was good and its defense was bad, not exactly ground breaking stuff. It's encouraging that UM scored more than average against Iowa and Wisconsin, but it's also notable that they came up under average against MSU, OSU, and Miss State. I think it's also notable that the defense did better than average 5/12 times, but three of those were in the first three games of the year and the other two were the slop fest against QB-less Purdue and against OSU where giving up 37 can't really be considered a win.
I don't pretend to understand FEI, but I think this is an interesting stat. UM scored almost a TD more than average on the season, and that's pretty solid, but I doubt it's #2 in the country (no way am I wading through that for other teams).