They should hire you to run these numbers before naming the MVP. It seems like you have also made a case for Ellington: 67 vs 27 for the Clowney hit. (Two/too/to/tew, rabble rabble).
The Weekly Maths: Prediction Review and Outback Recap
Back in August I posted a Stock Watch article on the teams I thought where significantly over or under valued entering the season. In the spirit of accountability here they are again with appropriate amounts of gloating and denial as to their accuracy.
Like Buying Pre-Ipod Apple Stock
Underrated: Ohio State, Notre Dame,
Overrated: Michigan State, LSU, West Virginia, Arkansas
For all three of Michigan’s main rivals, I thought the season would deviate from preseason conventional wisdom and for all three I was surprisingly dead on. All three of them were pushed to extra limits of their deviation due to some extreme luck/variance but all predictions proved true. This is what I had to say entering the season:
Their defense will keep them afloat but unless Michigan St breaks in a new crew on offense at an unprecedented rate, the offense will be this team’s limiting reagent.
If the bounces go Notre Dame’s way this season they have a shot to be a top-10 team.
The Buckeyes are set up for Urban to get credit for an upswing they probably would have had anyway, but it will probably take some significant first year growing pains to keep Ohio from a great theoretical bowl game.
Like the BCS Championship Game.
I correctly pegged LSU to be good but largely out of the title picture, West Virginia’s defense to be a tire fire and Arkansas to be an out and out disaster.
Like Buying Enron Stock in the Summer of 2000
Underrated: Texas, Missouri, Tennessee
Overrated: Kansas State
Texas was the most consistently great team of the 2000’s but the 2010’s have been a different story. I predicted a return to the past glory but with three consecutive data points now, I think its finally time for me to admit that Texas ain’t what it used to be.
I thought Missouri and Tennessee could break into the middle of the SEC this season but between the two of them they finished with two fewer SEC wins than Vanderbilt and Tennessee saw its orange-pant clad coach on the outs.
Never bet against Bill Snyder, never bet against Bill Snyder. Living in Kansas most of my life I should have known better. I thought this year’s squad would be better but have a worse record than last year’s lucky team but I was only half right. The Wildcats were substantially better and came within a rough night in Waco of playing for their first national title.
Like Cash Stuck Under the Mattress
Overrated: South Carolina, Boise State
Both of these teams essentially finished about were they were predicted to.
In Hail To The Victors I predicted 9-3 (6-2). The numbers behind were more in the 8.5 range. That seems about right. Michigan finished within the low end of the range that I (and I think many others) expected entering the year.
Outback Bowl Thoughts
A lot of big swing plays (obviously)
The biggest plays of the game:
1. –67% Thompson to Ellington for the game winner
2. +42% Gardner to Gallon to give Michigan the lead late in the fourth quarter
3. –27% Clowney
4. +20% Connor Shaw sacked to give South Carolina 3rd and 10 on their final drive
5. +19% Craig Roh stops Dylan Thompson for one yard in the final minute
6. –17% Michigan fails to convert the final two point attempt
7. –17% Shaw hits Sanders for a 31 yard touchdown
8. –15% Thompson to Byrd picks up a first down right before the final touchdown
9. –13% Nick Jones is brought down at the 4 after a 70 yard completion in the second quarter
10. +12% Michigan blocks a 42 yard field goal attempt
As you can see from the chart above, there are a lot of big drops and a lot of slow climbs. In some ways this was the reverse of what a lot of us expected. Going in the conventional wisdom was that Michigan might hit some big plays but would be unable to sustain drives. In fact, South Carolina churned out big plays all game long while Michigan put together several nice drives.
Denard Robinson: +1 EV and +0% WPA on 24 plays
Devin Gardner: +7, 58% on 48 plays
Jeremy Gallon: +7, +53% on 17 targets (8.5 yards/target)
Connor Shaw: +11, +27%, 36 plays
Dylan Thompson: +8, +67%, 12 plays
Ace Sanders: +8, +30%, 9 plays (+5, +10% on three punt returns)
I really thought Sanders should have gotten game MVP.
Game Theory Note
I really didn’t like going for two in the third quarter. With that much time left there are too many scenarios where losing that point can come back to get you. While the touchdown alleviated the direct impact of the eventual loss of two points, I think its entirely possible Mattison’s approach changes knowing that a field goal would only tie the game on the final drive.
I’ll get deeper into the numbers throughout the offseason but here are some key areas I see for next season:
- Schedule. No Alabama. All other losses from this season are home games in 2013.
- Not breaking in a new quarterback. Michigan is in the rare situation where they lose a three year starter but are able to transition to an experienced replacement. Definitely a silver lining for this season.
- Key returners outweigh losses. Denard and Lewan will certainly be losses, but its hard to pick a position group that will be worse in 2013 than 2012.
- Borges unleashed. I still have some major reservations about him, but being more in his comfort zone can only be a good thing.
- Quarterback depth. If Devin struggles or gets hurt it becomes very scary. We all have high hopes for Shane Morris, but I’ll have a post later this year showing the total lack of success from true freshman quarterbacks.
- Final year of Rodriguez recruiting/attrition mess. A lot of the high level metrics see next year at a similar level to this one in terms of talent with 2014 being the year that the Hoke era really begins in terms of upper class recruits.
- Offensive line. They’re gonna be young.
Hit up the comments or twitter if there is anything you would like to see from these posts over the course of the off-season.
Sorry to threadjack but Pete Elliot has passed away. Fromer Michigan QB and All-American. Also lead Michigan to the 1948 Rose Bowl.
I don't know if its possible for the OL to be worse than it was this year. Schofield will move over and be fine, right tackle will probably be a downgrade. But the interior has nowhere to go but up.
But, I think there's a good chance that Braden could hold his own vs. Schofield, at least when it comes to the run game. Braden's huge. If all goes well, Michigan could be a heavy right-handed team.
you could show how Morris will deviate from the norm and be different from all other frosh. You're the Mathlete, you can do that to numbers, right?
Can you point me to a link to explain expected win odds? I would like to hone my gut on this sort of analysis but I didn't feel 70/30 at half getting the ball back in the third.
Ace got decleated by the media hype for Clowney. Small things win games. Bring on Norfleet in 2013.
Great post as expected.
The numbers are developed based on the last ten years of games between evenly matched FBS teams, accounting for down, distance,possession, timeouts and score.
I enjoyed these maths. Your method looks to have a high predictive value. Kudos