Measuring Michigan's dominance over their opponents with a bonus '97 comparison

Submitted by trueblueintexas on January 9th, 2024 at 5:07 PM

At the close of the game last night, one of the announcers said this was Michigan's 11th win by 21 points or more this season. I thought that was an incredible measure of how dominant this team was. 

Then I saw the thread about who the best team Michigan played this season and it got me wondering, was Michigan the best team each of our opponents played this season? Obviously, yes...but how to show it. 

Then it got me thinking about a good way to compare this team to the '97 team. 

So I started playing with numbers and here's what I can say... this team was incredible on both defense and offense against their competition, and....better than the '97 team overall. Here's one look at some numbers to show it. 

I wanted to see how close Michigan came to giving up the fewest points and scoring the most points against each opponent this season and how their overall margin of victory compared to each opponents worst defeat. 

Here's the info for how well Michigan's defense performed holding opponents to their worst scoring output for the season:

Points For is what the opposing team scored against Michigan.

Season Worst was Michigan's opponents worst game against their other opponents.

Delta is the difference between Michigan's result vs. the opponents worst scoring result. I.e. East Carolina scored 3 against Michigan but Navy shut them out. The difference between what Michigan was able to do against East Carolina vs what Navy was able to do was 3 points. 

Out of 15 opponents, Michigan held that opponent to their worst scoring output on the season 5 times (highlighted in yellow). In some cases you can see Michigan shared that distinction with another team...poor, poor, wretched Iowa offense. 

Overall Michigan gave up 10.4 points per game. 

The smaller the delta number, the closer Michigan is to being the best defense that team played. The bigger the number means another team did a better job than Michigan. Overall Michigan was very close to being the hardest defense each team faced this year. That can be seen in the individual delta numbers as well as the average (4.7). The one outlier was OSU's abysmal 3 point performance against Missouri. If you take that out, the team that held OSU in check best this season was ND at 17 points (albeit, ND still lost). If you throw out the Missouri result, ND's 17 cuts Michigan's delta for OSU to 7. That would reduce Michigan's overall average delta to 3.8. That number will be important later. 

Summary: Throw out some of the garbage time scores by Michigan's opponents and it is clear, week after week, this was the toughest defense any of our opponents faced all season. Save for Maryland when they went up against...Nebraska???

We all knew the defense was good. How about the offense?

Points Against is how many points Michigan scored against the opponent.

Season Worst is the most points that opponent gave up to another team on their schedule. 

Delta is the difference between Michigan's points scored against that team vs. what the best performing other opponent did against that team. 

6 times Michigan scored the most points against an opponent this season (highlighted in yellow).

Overall Michigan averaged 35.9 points per game.

Michigan's average delta of 5.9 points means they were close to being the most prolific offense each of Michigan's opponents faced this year. While not as close as the defense at 4.7, that is still incredible. Looking at the individual delta's, there were three times other teams scored about 2 TD's more than Michigan did against that opponent. Note: Washington gave up those 42 points and still won. My goodness USC's defense was bad.

So what does it look like when you pair the defense with the offense? Here's Margin of victory:

First three columns show opponent and the score of their game against Michigan.

Margin of defeat should be clear.

Season Long Largest Margin of Defeat is that opponents worst game of the year as far as getting blown out. 

8 Times!!! 8 Times out of 15 opponents Michigan gave that opponent their worst defeat of the season (highlighted in yellow). That is incredible. Even more impressive is the delta average of 4.1. On average, Michigan was only a little more than a field goal away from giving every opponent their worst loss of the season. Look at those individual Delta numbers. They are simply glorious. The weird outlier here is Maryland. This is where "events" come together to skew things. Maryland was a trap game and JJ was injured so Michigan only won by 7 which is in-line with how Michigan outscored OSU & Alabama as a reference point.  At the same time, Maryland somehow got blasted by PSU as an outlier. Drop that Maryland delta to 9 (the largest other difference) and Michigan's average delta drops to 2.3 points. That would mean Michigan was less than a field goal away on average from being each opponents worst defeat of the season.

Given the significant variances week to week in college football and the ability for a team to just not have it one week that Margin of Victory chart is simply amazing showing how consistently dominant this Michigan team was. 

Now you have a better idea of how amazing this team is and how to read the info, here are the complete charts for 2023 & 1997. You can decide which team you think is more dominant in their era. 

2023:

1997:

The '97 defense gave up slightly fewer points per game and had a slightly smaller delta.

See note from above about how the Missouri/OSU game skewed the defense data and without that the two defenses are basically equal.

The '97 offense wasn't anywhere as close to dominant. With an average delta of 18 points. That means on average, other teams scored 18 more points than Michigan could against the same opponents. 

The Margin of Victory numbers really tell the difference. Only twice did Michigan give their opponent their worst defeat of the season. On average, they were 12.3 points per game from doing so. 

For those wondering about the green shading. Those are the teams where Michigan was not their worst in any category. I.e. other teams held them to lower scores, and other teams scored more on them, and other teams had a higher margin of victory. For all other teams, Michigan gave them their worst day in some manner or another. 

And for Nebraska & MSU, well, let's just say they hope never to see this Michigan team ever again...just like those other 13 opponents.

Congrat's on an amazing season Team 144. Thank you for the many memories. 

Jonesy

January 9th, 2024 at 5:11 PM ^

I think all non playoff bowl games should be excluded from all statistical analyses. Pretty ridiculous that S&P had UGA as the best team after the Bama game since it thought they played the real FSU team. That's the worst example but that was not the real OSU team either and that was pretty much the case for every bowl with a team in the top 25 (if not more).

caup

January 9th, 2024 at 5:20 PM ^

I think the 2023 stats are even better than what you posted, because I think you need to throw out the OSU and PSU bowl losses, when those 2 teams were a shell of their regular season rosters.

Vasav

January 9th, 2024 at 6:35 PM ^

I wrote this diary last night but I think Team 144 is probably the best Michigan team ever, and definitely has the best resume ever - altho the latter is because the CFP forces you to finish your season against two top5 teams. Nevertheless, I give 2023 the edge over 1997 in particular because Team 144 only ever trailed outside of the first quarter once - in the Rose Bowl. 118 trailed 3 times in the 2nd half - Notre Dame, Iowa and Wazzu. 2023 controlled the games with a dominance that '97 didn't.

I further believe that this was the best Big Ten team in 55 years - since the 1968 Buckeyes. The '02 Buckeyes played 6 one score games, the '14 champs lost to Virginia Tech. The 1968 Buckeyes never trailed in the second half and only ever trailed in the first half twice - once in the Rose Bowl in a 1vs2 matchup.

I think 1948, 1933, 1902 and 1901 all compare pretty well to 2023 - but 1933 took a tie, 1901 was dominant but ain't played nobody, 1902 was very good but didn't play an eastern power, 1918 was werid, 1923 didn't get to play Big Ten/National co-champ Illinois, and 1948 was pretty damn good, but didn't have the opportunity to play a top 5 team. So even if '01, '02, '18, '23 and '48 were as good, 2023 had the best resume.

Other Andrew

January 10th, 2024 at 6:35 AM ^

“You don’t compare great to great” - Jim Harbaugh in his introductory press conference.

I’ve been thinking about this quote as these inevitable comparisons are arising now.

These are both incredible teams from two very different eras. Both shocked the world, albeit in different ways, the ‘97 team coming off of four straight losing seasons, the ‘23 proving you can win in the “stars matter” playoff era against the death star programs, built over a three year period.

None of the current team’s players were even born yet when Charles Woodson won the Heisman. For many of us there is a linear aspect to our fandom which connects these periods. But in the end it’s apples and oranges. Both are sweet and healthy. Both should be cherished for the magical seasons they were. Let’s just celebrate the hell out of this.

Vasav

January 10th, 2024 at 8:49 AM ^

This is true - at the end of the day, "flags fly forever " championship banners are what count. 2023 did something amazing, so did 1997, so did the other 10

It's still fun to compare and see kinda what we do value - I do think this team did have the best season of a Big Ten team since the 1968 Buckeyes. I know advanced stats say it's been a very good year, and the other 15-0 squads of the CFP era were maybe better. But truly, the important thing for all of those teams is that they were national champs

Reno Drew

January 9th, 2024 at 5:24 PM ^

Just on the eye test, up until the last two games, I was still thinking the 1997 team would have beat the 2023 team in a hypothetical match up.  I just remember watching the 1997 team and thinking nobody that year could beat them, and I tend to be a glass half empty BPONE Michigan fan who usually thinks the worst. 

After watching what they did to Alabama and UW, I'm now going with team 144.

Both great teams in their era!  Go Blue! 

MichaelCarras

January 9th, 2024 at 5:25 PM ^

The 1997 team was 6th in SP+, 4th in Sagarin. and 5th in SRS.

The 2023 team is 1st in SP+, 1st in Sagarin and 1st in SRS.

I know these debates are fun but the answer is obvious based on the numbers. This team is clearly better. The 2022 and 2021 teams were also higher rated than the 1997 team.

Also as a random aside, the 1985 team was number 1 in SP+ and SRS  https://www.sports-reference.com/cfb/years/1985-ratings.html

J. Redux

January 9th, 2024 at 7:07 PM ^

It gives me a great deal of joy to learn that Navy broke the East Carolina shutout streak that they were so proud of that they kicked the saddest of all sad field goals against Michigan.  Michigan deserved to break that streak, but at least the football gods punished them for their transgression.

Vegas Wolverine

January 9th, 2024 at 7:24 PM ^

Charles Woodson himself said a few days ago on the Rich Eisen Show that if Team 144 won the National Championship, they'd be the greatest Michigan team of all time. Won the playoffs that he never had the chance to go through. Went 15-0. He said they're the best of all time at Big Blue.

Jacedeuce

January 9th, 2024 at 7:40 PM ^

Small nitpick.


That means on average, other teams scored 18 more points than Michigan could against the same opponents.

Not quite. Since you input a 0 when Michigan scored the highest point total of the season against that opponent, you are skewing the average. The conclusion you can draw is that "On average, the highest point total of the season was 18 points higher than Michigan could score." Or you could input a negative number in that spot to account for Michigan outscoring all others against that opponent. In the case of the 1997 team, it was only a single game. Another way to arrive at the delta between Michigan and the toughest opponents would be to average Michigan's point total and average the highest point totals and calculate the delta between those numbers.

trueblueintexas

January 9th, 2024 at 10:35 PM ^

The math is correct. The comment needs better clarity/correction. For each of Michigan’s opponent’s opponent’s most points scored, you are trying to see how close to zero Michigan can get. (I.e. the closer Michigan is to 0 the closer they are to the highest scoring opponent’s opponent)

The average of Michigan’s opponent’s opponent’s most points scored was 18 points more than Michigan’s performance against Michigan’s opponents.

If that’s more clear, you are smarter than me.  

AlbanyBlue

January 9th, 2024 at 10:44 PM ^

Just from my recollections of that '97 offense, I don't think the '97 team would score much at all against the '23 defense. '23 wins in a lower-scoring game, probably by 7-10 points. Call it 17-10.

Durham Blue

January 9th, 2024 at 11:37 PM ^

The 2023 Michigan team easily passed the eye test for me.  I have been watching Michigan football since 1985 and this team right here is the best Michigan team I've ever watched.  There is no counter point in my mind.  Talent, strength, conditioning, speed, skill, football IQ, toughness, grit, resolve, confidence.  They checked all the boxes and passed with flying colors.

trueblueintexas

January 10th, 2024 at 1:42 AM ^

I agree. Been watching for a little longer and this team is the best I have seen. 
If I have more time I may run a similar analysis for some of the other great college teams over the years to see how they compare to this Michigan team. I doubt many will match the overall dominance of this team against their schedule. 

Mgoblue0205

January 10th, 2024 at 10:00 AM ^

The '97 defense gets a slight edge IMO over the '23 defense, mainly because of Charles. Some fans do forget though that we had 3 future pros at LB(Sword, Gold, Jones). The offense is night and day different. Even though '97s OL was loaded with NFL talent, we featured kind of a weak RB group. Howard was a solid college RB, Anthony Thomas was only a freshman. The recievers were also very average. Streets was a good college WR, but after that Russell Shaw and Co. were average at best. People say Michigan would've been crushed by Nebraska that year which is ridiculous. Just like what the defense did this year to a relatively one-dimensional Washington offense, that '97 defense would've stuffed Nebraska's one dimensional offense. I think OSUs '02 NC team is alot like '97 Michigan. Win with defense and a running game, with a QB that makes good decisions. Well, nobody gave OSU a chance against Miami if I remember right before that game. The perception is that because Michigan didn't blow teams out and Nebraska did, that Michigan didnt stand a chance...It's alot like the '02 Buckeyes and Miami although I hate comparing our team to an Ohio team.

MichaelCarras

January 10th, 2024 at 12:10 PM ^

Pretty good intuition on everything. Both teams were the number 1 rated defense according to Sport Reference. The 1997 team had the number 1 defense by a huge margin whereas the 2023 team was only marginally ahead of OSU this year. though absolute rankings between 1997 and 2023 defenses are close.  But the 1997 team had the 32nd rated offense just ahead of Kent State. https://www.sports-reference.com/cfb/years/1997-ratings.html   The 2023 team had the 11th best offense. https://www.sports-reference.com/cfb/years/2023-ratings.html

trueblueintexas

January 12th, 2024 at 1:18 PM ^

One of the reasons I did this exercise was to see if Michigan really did blow out teams on the same level as other competition did, and the emphatic answer is yes!!! This 2023 Michigan team beat the teams on it's schedule by about as much as they possible could. The data says within 2.5 points. That means, over the course of the season Michigan's margin of victory across all opponents was less than a field goal away from being the largest possible margin of victory. The Maryland game was really the only outlier.