UM Least Lucky to be Undefeated - ESPN Expected Win %
Narrative I keep seeing around the interwebs is that Michigan has struggled and doesn’t deserve to be a top 4 team. Seems to have been echoed by the AP voters this week as well, dropping Michigan below Clemson.
According to ESPN Analytics of expected win probability throughout the games, Michigan is the only top 10 team that hasn’t been expected to lose a game at some point during the season. FPI ranking (#) for reference
1) Georgia: 18.9% against Missouri (58)
2) OSU: 48.6% against Notre Dame (15)
3) Alabama: 20.4% against Texas (4)
4) Clemson: 18.5% against Wake Forest (24)
5) Michigan: 61.7% against Maryland (20)
6) Tennessee: 20.5% against Pitt (42)
7) USC: 24.1% against Oregon State (48)
8) Ok St: 29.4% against Texas Tech (44)
9) Ole Miss: 34.5% against Kentucky (36)
10) PSU: 10.6 % against Purdue (22)
*Manually pulled data, so I may have missed a data point that is worse for non-Michigan teams
So maybe for an offense that is designed to wear down a defense and dominate in the second half, we should wait until the second half before screaming that the sky is falling? Just a thought.
October 10th, 2022 at 11:16 AM ^
This is some whiny shit. cmon. every time we lose its a fluke? wtf.
There are always plays in games that flip things back and forth.
When stroud threw that pick 6 yesterday to make it 7-7, did osu pack it in and say.. see.. flukey play.. WR misread and gifted them 7? No. they pounded them into the dirt. without their best WR and best (i think) RB.
The punt thing was a fluke.. k... but the rest I'm not buying. Covid being in existence doesnt make it a fluke. Neither does weather.
October 10th, 2022 at 11:50 AM ^
I am pretty sure we were pounding them into the dirt last year, and the strip sack TD was dumping that dirt on their team's grave. Then the refs gave them hope, and we had some freshman mistakes when our starting QB got dinged up. They were also in desperation mode, and converted 2/2 2-point conversions. A whole lot of things had to go their way to win, and their best weapon was catching us subbing our players out on D. An effective gimmick, but not one that shows that they feel like they could beat us straight up.
And if you put ANY stock in the goofy covid year, when everything was a mess and many top players were either injured, sitting out, or just not invested, I don't know what to tell you. That is about the biggest asterisk ever, and pretty only useful for twitter memes.
October 10th, 2022 at 2:38 PM ^
Okay - you can see it that way.
We also didnt have to give up 2 long tds, 2 2-pt conversions, etc..etc.. i agree that it seemed unlikely, but that doesnt make it a fluke. tahts on our team and coaches.
They ran all over us. And we did nothing to stop it
October 10th, 2022 at 4:56 PM ^
I don't think every time we lost to them it's a fluke - I do think we've lost 4 close, one possession games where one play going differently would've made the difference in the outcome. And we've won 3 games where the closes one was a 9-pt win where they scored with 1 second left to say "defeated with dignity."
Dunno how much of close losses are bad luck or bad game management - Harbaugh's style to sit on teams works with bad teams, but when the teams aren't that bad it may lead to more close games. Nevertheless, we've been solid for 6 of those 7 seasons, and they've been solid for 3 of them. I'd rather take our program, even with the close-game luck/mismanagement.
October 10th, 2022 at 11:25 AM ^
Yes, tiny sample size, with some Covid noise and ref assistance. Kenneth Walker is not walking through that door, and we should be more ready for their attempts at tempo after playing against IU's ludicrous speed offense. MSU is just a bad team right now, and if we are willing to take advantage of the many open receivers that they are going to let run wild in their secondary, then we should win by 3 TD. If we decide to run into the teeth of their defense, we should still win, but it will be way closer and could go the other way.
October 10th, 2022 at 1:15 PM ^
You cant "if and but" your way to a 7 year history. At some point its on the coach. It sucks for sure but all those games had opportunities to take it out of luck/refs hands.
October 10th, 2022 at 11:32 AM ^
No it won't. I know it is hard to prove but the narrative then would be "they will play hard, they will get up for the game, look what happened in (pick year you want), it is their Superbowl". Until Harbaugh wins 5 NCs in a row and haul in top 1 class every year, there will be people with negative attitude. That is ok. It just means people care and love Michigan Football. Go Blue!
October 10th, 2022 at 10:33 AM ^
Post of the day. Would like to see the mgoblog crew examine this idea on a podcast.
October 10th, 2022 at 12:10 PM ^
- it usually takes us all 4 quarters to beat good teams
I might quibble on this one. We've delivered some absolute beatdowns - PSU a couple of times, Wisconsin a couple times, ND in 2019, Northwestern in 2015 (they were good that year)...
October 10th, 2022 at 5:01 PM ^
That ND team was top ten, Harbaugh's 2nd top ten win at M, and I believe his first "upset" (altho the Iowa team from earlier that season outranked us but I think we still were favored)
October 11th, 2022 at 9:06 AM ^
I would add in OSU and Iowa from last year to that list. Two very good teams that Michigan basically thumped from the beginning.
October 10th, 2022 at 10:28 AM ^
The problem with "an offense designed to dominate the second half" is that second halves are not as predictable as first halves.
My gripe with Michigan during the Harbaugh era has been a lack of urgency. This reaches its absolute nadir...pretty much every year against MSU. Remember that year that there was clearly inclement weather coming, yet we were content to sit on our scoring deficit to try to make up the margin in the second half?
Well, the weather hit, neither team did much, and so MSU sat on their lead until the end.
If your offense doesn't do much in the first half, it means that you're getting out-game-planned.
October 10th, 2022 at 10:34 AM ^
Or fans could still have complaints about the offense if it's struggling at times. Just a thought
October 10th, 2022 at 10:40 AM ^
I'm confused by this. I'm not a statistician but are they saying Georgia had a better chance to lose to Missouri than we did against Maryland?
October 10th, 2022 at 10:52 AM ^
Yes, these are expected win % during the game, so at one point in the Georgia vs Missouri game, Georgia only had an 18.9% expected chance to win the game. This occurred when Missouri was up 22-12 and Georgia threw an incomplete pass with 14:03 left in the game on 1st & 10 on their own 25 yard line.
For Michigan, our lowest was 61.4% against Maryland. This was after Tagovailoa completed a 26 yard pass to the Michigan 41 as Maryland led 13-10 with 8:56 left in the 2nd.
October 10th, 2022 at 11:24 AM ^
OK gotcha. That's what confused me, these were during game stats?
October 10th, 2022 at 11:36 AM ^
Yes, ESPN does it for football and basketball. Essentially just a win% associated with what a live money-line bet would be.
I like to check it to see how much I'm overreacting to things. When you're like "OMG Iowa is all of a sudden going to have an offense and score 3 TDs" and then you look at the current win expectancy being 98%, it helps you relax a little.
October 10th, 2022 at 12:13 PM ^
So Georgia had 2nd and 10 at their 25, down 10, with almost a full quarter to go? That doesn't seem that bleak. Some of these win probability numbers, I don't know. It may be that the average team only wins 18.9% in that situation, but Georgia clearly isn't an average team.
October 10th, 2022 at 1:34 PM ^
At that point Georgia had scored 4 FGs on 9 drives and essentially needed 2 TDs and to give up no points on their 3 remaining drives.
Yeah Georgia is the much better team, but that's still a tall ask. An average team would've been < 5% chance to do that.
October 10th, 2022 at 11:07 AM ^
Though competition caveats apply, this is primarily indicative of efficiency. We haven't necessarily been spectacular, but it's desirable to play well enough that you don't need to be.
"They're real, and they're efficient."
October 10th, 2022 at 11:22 AM ^
I read some summaries of Saturday's action yesterday. One article (I believe at ESPN), praised Clemson for rebounding from a bad start against a weak conference foe (10-3 at the half) to score 3 unanswered touchdowns in the second half (when their defense really came alive) to win a game in which they were favored by 26 point. They won 31-3 and covered the spread.
By contrast, Michigan was criticized for struggling in the first half against a weak conference foe (10-10 at the half) and for only scoring 3 unanswered touchdowns to win a game in which they were favored by 22 points. They won 31-10 and did not even cover the spread.
This was put forth as justification for Clemson jumping Michigan in the AP rankings. Whatever, keep winning and get to the playoffs and it will take care of itself.
October 10th, 2022 at 11:39 AM ^
I saw the justification that Clemson's defense has really come on lately, only allowing 23 points in their last two games. In contrast Michigan's defense has really struggled in the last two games, allowing . . . 24.
The poll voters make their decisions first, and then backfill the justification.
October 10th, 2022 at 1:13 PM ^
Yup, same principle. And there is a human tendency to make a decision and then seek confirming information. Having an ACC team in the CFP mix is sexy and their "recent statistical dominance" over Michigan helps to justify elevating them.
October 10th, 2022 at 1:08 PM ^
Probably unpopular opinion, but I'm fine with Michigan dropping and think it's justifiable to drop them further. These things sort themselves out, and while I (and analytics) still think Michigan is one of the best 4 teams in the country, other teams have better resumes up to this point.
Clemson, Tennessee, UCLA, Oklahoma State, Ole Miss have better wins, so far (btw, I don't think USC or OSU have better wins at this point). If Michigan takes care of business against PSU, they'll have the best win of the country and will be secure in the top-4.
October 10th, 2022 at 5:09 PM ^
i think they'll have one of the top 3 wins in the country. UGA's demolishing of Oregon is going to hold up until Oregon loses....and the winner of UT-Bama is probably on par with the winner of M-PSU
October 10th, 2022 at 1:23 PM ^
I suppose there are 2 ways to read this.
One way would be: every Top 10 team except U-M (and OSU, arguably) has had a "oh shit, we really may lose this game we need to turn it on now" moment, and has done what they needed to to wind up winning.
And outside of machines like 2020 Alabama, all very good and elite teams inevitably have that moment where they're in a challenging spot. 1997 Michigan had the defense stopping ND after the offense kept turning it over in the 4th Quarter, and the offense's final drive of the Iowa game. 2021 Michigan had the Nebraska & PSU road games.
Conversely, 2016 Michigan - they weren't truly tested at any time (the Wisconsin game, U-M's win probability was never below 50%) prior to the Iowa game. And they lost. In retrospect, it was a bit TOO easy for them through their first 9 games.
Net: that's only a few data points, but I'd argue it is GOOD for a team to be in a tough spot at some point prior to November.
As is often the case in real life, growth and bonding (with teammates and friends) may happen the most as a result of being in an uncomfortable spot.
October 10th, 2022 at 2:34 PM ^
In 2021 many were saying, "This is fine against [insert name of weaker opponent] but this style of play won't beat teams like Ohio State."
And yet it did, convincingly. Maybe this year will turn out as well, maybe not. We don't know yet.
Results are the only proof. Arguing over conjecture is a fool's errand.