Sagarin Predictions: Week 6

Gopher07

Captain of Awesome
Joined
Nov 20, 2008
Messages
9,008
Reaction score
15
Points
38
No huge shifts in the formula as the Gophers were off this week. Probably the biggest change week over week involved Nebraska shifting further in our favor ... and there's a better-than-zero chance the Huskers are winless when we go to Lincoln. Smells like a trap game, if we can have such a thing. We also inched the spread slightly higher against Indiana and slightly lower against Purdue and Northwestern, but nothing notable.

This weekend is our first "likely loss" game of the year. As far as I can tell and remember, the last time we won a "likely loss" game was against Nebraska in 2014. Hopefully the week off gave the team a chance to get things right after Maryland and put in a few wrinkles to get the offense going against a really tough Iowa defense. Looking forward to the battle for Floyd on Saturday!

A reminder on how I arrive here, I take Jeff Sagarin's predictor ratings (found here), and calculate what the computer says the spread should be, for now through the end of the season. Negative numbers favor the Gophers, positive favor the opponent. I round to the nearest half point, and if neither team has a greater than 0.24 advantage, the game will be listed as a PICK.

vs New Mexico State W (Predicted -12.5; Actual -38)
vs Fresno State W (Predicted +0.5; Actual -7)
vs Miami (Ohio) -22 W (Predicted -22; Actual -23)
at Maryland L (Predicted -1.5; Actual +28)
vs Iowa +9
at Ohio State +28
at Nebraska -4
vs Indiana -4
at Illinois -8
vs Purdue -2.5
vs Northwestern -1.5
at Wisconsin +17.5

Final record: 8-4 (5-4)

Likely in the W column (>7-point advantage to Minnesota): @ Illinois
Toss ups (no team with a greater than 7-point advantage): @ Nebraska, vs Indiana, vs Purdue, vs Northwestern
Likely in the L column (>7-point advantage to opponent): vs Iowa, @ Ohio State, @ Wisconsin

NEW FEATURE!
Thanks to user matt, we now have predicted B1G standings based on the formula. I have also included some likely destinations for teams should they finish as predicted. Because of the Wisconsin loss to BYU, I'd say Iowa has the best shot to make the NY6 from the B1G West in the below standings. If the Badgers win the West at 7-2, they'd likely head into bowl season at 10-4 after a conference title game loss, while the Hawkeyes would be sitting at 10-2.

From the East, Ohio State seems to be an odds-on-favorite to make the CFP at 14-0, while Penn State looks good to make the NY6. Michigan is next in line but making the NY6 might be difficult if they lose to Ohio State and Penn State in addition to ND.

B1G West
Wisconsin: 7-2
Iowa: 7-2 - maybe NY6
Minnesota: 5-4
Northwestern: 4-5
Purdue: 2-7
Illinois: 1-8
Nebraska: 1-8

B1G East
Ohio St: 9-0 - likely CFP
Penn St: 8-1 - maybe CFP / likely NY6
Michigan: 7-2 - maybe NY6
Michigan St: 6-3
Maryland: 4-5
Indiana: 2-7
Rutgers: 0-9

Crossover Games
West: 6-15
East: 15-6

Interesting B1G Notes

- Wisconsin has 1 toss up left. Notable games: +4 @ Mich, +8 @ PSU, -11 @ NW, -12 @ Purdue.
- Iowa has 0 toss ups. Notable games: +11 @ PSU, -9 @ Purdue, -9 @ Minn.
- Ohio St has 0 toss ups. Notable games: -12 vs Mich, -14.5 @ MSU.
- Rutgers is +6 vs Illinois this week. After that, closest games are +18.5 vs NW, +25.5 @ MD, +31 @ MSU…
- Minn and NW are projected very similar - go 1-2 in crossovers and beat the same teams in the West. The difference is the Minn vs NW game which we are projected to win with homefield advantage.

Previous weeks:

Preseason
Week 2
Week 3
Week 4
Week 5
 

I'd be thrilled to go 8-4 (5-4).

After two tough games, the Gophers are going to have to make hay while the sun shines. To me, those five games (after OSU) are going to define the season.

... Or, we could lay a beating on Iowa this week, and elevate expectations.

Always enjoy seeing this weekly feature! Thanks, Gopher07!

JTG
 

Sargin doesn't seem to account for critical injuries / ominous OL performance and etc.
 

As far as I can tell and remember, the last time we won a "likely loss" game was against Nebraska in 2014.

I have to think that the Holiday Bowl in 2016 vs. Washington St. falls into this category, no?
 

I have to think that the Holiday Bowl in 2016 vs. Washington St. falls into this category, no?

Yeah that had to be an upset. I always say / think Washington St. was ranked, but they were not at the time we played.
 


I have to think that the Holiday Bowl in 2016 vs. Washington St. falls into this category, no?

That was one of my favorite games to watch in recent history. Recipe for disaster with a lot of key defensive personnel out and a strong passing attack for the cougs. Instead of whining and making excuses, the coaches found a way to scheme around the missing players and pull out the win.
 

I have to think that the Holiday Bowl in 2016 vs. Washington St. falls into this category, no?

I think you're right, but not sure. After the bowl game Washington State was a 5.5 point favorite on a neutral field, but that almost certainly dropped as a result of the game so it might've been >7 points prior.
 

Great stuff gopher07, thanks for all the effort. I continue to believe there are five wins to be had on the schedule, and this seems to confirm that notion.
 

I have to think that the Holiday Bowl in 2016 vs. Washington St. falls into this category, no?

What was the spread? That's roughly what he's using to define likely loss/toss-up/likely win, with the threshold being 7 points one way or another.
 



That was one of my favorite games to watch in recent history. Recipe for disaster with a lot of key defensive personnel out and a strong passing attack for the cougs. Instead of whining and making excuses, the coaches found a way to scheme around the missing players and pull out the win.

And if I remember correctly, at the last second the backup center couldn't play ... so it suddenly fell onto Mitch Leider's brother, who was woefully undersized against the big Pacific Islanders on Wazzou's DL. But he played the whole game, and stepped up!
 

What was the spread? That's roughly what he's using to define likely loss/toss-up/likely win, with the threshold being 7 points one way or another.

I think the Vegas spread was WSU -10.5
 

I have one small bone to pick with this. It's about the predicted record ..... my gripe is that it counts ALL "toss-up" games as victories for us, because all of them (going into this week) are a small advantage for the Gophers.

OK ... but doesn't that defy the definition of toss-up??? A toss up means, the expected probability is 50% . So, it would make a lot more sense if you gave use HALF those wins, not ALL. And that would predict a record of 6-6. Which is much more realistic.
 

I have one small bone to pick with this. It's about the predicted record ..... my gripe is that it counts ALL "toss-up" games as victories for us, because all of them (going into this week) are a small advantage for the Gophers.

OK ... but doesn't that defy the definition of toss-up??? A toss up means, the expected probability is 50% . So, it would make a lot more sense if you gave use HALF those wins, not ALL. And that would predict a record of 6-6. Which is much more realistic.

It is defined. Look closer.
 



I have one small bone to pick with this. It's about the predicted record ..... my gripe is that it counts ALL "toss-up" games as victories for us, because all of them (going into this week) are a small advantage for the Gophers.

OK ... but doesn't that defy the definition of toss-up??? A toss up means, the expected probability is 50% . So, it would make a lot more sense if you gave use HALF those wins, not ALL. And that would predict a record of 6-6. Which is much more realistic.

You're right.

For the purposes of simplicity I take pure win/loss when laying out a final predicted record, but if I were really enterprising I would lay out some kind of probability distribution based on the predicted outcomes of each game - and given we have four relatively close games in our favor along with one relatively not-close game, we're probably looking at 3-4 additional wins on top of the 3 we already have, meaning we're landing somewhere in the 6-7 win range when it's all said and done.
 

You're right.

For the purposes of simplicity I take pure win/loss when laying out a final predicted record, but if I were really enterprising I would lay out some kind of probability distribution based on the predicted outcomes of each game - and given we have four relatively close games in our favor along with one relatively not-close game, we're probably looking at 3-4 additional wins on top of the 3 we already have, meaning we're landing somewhere in the 6-7 win range when it's all said and done.

Well you wouldn't have to go that fancy. You could still bin the games by < -7, [-7,7], >7, as you do now. And could still give automatic win/loss to the likely bins. Just asking for 50/50 split on the toss-ups! You can even give the good guys the benefit of the doubt if the toss-ups are odd :)


I'm just saying, even if we are predicted to win all four of those games, on an individual game basis, the probability that we win all four ... is gonna be pretty low. Same reason why parlay bets are tough.
 

Well you wouldn't have to go that fancy. You could still bin the games by < -7, [-7,7], >7, as you do now. And could still give automatic win/loss to the likely bins. Just asking for 50/50 split on the toss-ups! You can even give the good guys the benefit of the doubt if the toss-ups are odd :)


I'm just saying, even if we are predicted to win all four of those games, on an individual game basis, the probability that we win all four ... is gonna be pretty low. Same reason why parlay bets are tough.

Lighten up Francis!
G7 puts this out every week and hasn't changed the format in years. It's a prediction not the actual result. I think what he has done and how he presents and represents everything in his initial post each week is exactly the way it should be.
 


The games are later in the season, but hard to believe the bookies will favor us over Northwester or Purdue. An 0-6 Nebraska could play their game of the year against us and Indiana is much improved. If Iowa drubs us, Illinois may be our best shot at adding a W. On the other hand, if the coaches pull some rabbits out of the hat and the team is sky high Saturday, the season might pivot. Having our two best runners out, plus Winfield, and Zack limited in mobility, makes that unlikely.
 

The games are later in the season, but hard to believe the bookies will favor us over Northwester or Purdue. An 0-6 Nebraska could play their game of the year against us and Indiana is much improved. If Iowa drubs us, Illinois may be our best shot at adding a W. On the other hand, if the coaches pull some rabbits out of the hat and the team is sky high Saturday, the season might pivot. Having our two best runners out, plus Winfield, and Zack limited in mobility, makes that unlikely.

Why exactly is it hard to believe a game between two bad teams would favor the home team?
 

Sad thing is Saragin can’t factor in injuries. How many points are Rodney Smith and Antoine Winfield Jr worth?
7? 14?


Sent from my iPhone using Tapatalk
 

Yeah I'm not sure yet that I believe going to Champaign will be easier than going to Lincoln.


MAN ... it would suck if we end up losing both those road games, yet win all three home games vs Indiana, NW, and Purdue. Because we're not going to win at Ohio St, and we likely won't win at Madison. So that puts the almost full weight of the season on this homecoming game vs Iowa. Win ... and we go to a bowl game, in that scenario. Lose, and it's a repeat 5-7 season, sitting at home.
 

MAN ... it would suck if we end up losing both those road games, yet win all three home games vs Indiana, NW, and Purdue. Because we're not going to win at Ohio St, and we likely won't win at Madison. So that puts the almost full weight of the season on this homecoming game vs Iowa. Win ... and we go to a bowl game, in that scenario. Lose, and it's a repeat 5-7 season, sitting at home.

If we win those three, we're bowl-eligible (with six wins).

Still, a win against Iowa would put a smile on my face that would last about a week.

JTG
 

Doh! I make mistakes like this all the time. That makes my point about the Iowa game incorrect and moot.

Thanks for correction
 




Top Bottom