ESPN's FPI predicts every game on Minnesota's 2021 schedule

nitramnaed

Well-known member
Joined
Nov 19, 2012
Messages
2,184
Reaction score
1,525
Points
113
Per Saturday Tradition:

Minnesota is hoping to contend for the Big Ten West crown this season, but ESPN’s Football Power Index is skeptical of that goal.

FPI projects the Gophers to win 6.1 games this season — an improvement over the 2020 season but a long way off from their 11-2 campaign in 2019.

Here’s FPI’s projected win probability for every Minnesota game this season:


  • Sept. 2: Vs. Ohio State — 12.3% (L)
  • Sept. 11: Vs. Miami (OH) — 88.3% (W)
  • Sept. 18: at Colorado — 47.3% (L)
  • Sept. 25: Vs. Bowling Green — 97.3% (W)
  • Oct. 2: at Purdue — 39.5% (L)
  • Oct. 16: Vs. Nebraska — 60.4% (W)
  • Oct. 23: Vs. Maryland — 62.9% (W)
  • Oct. 30: at Northwestern — 30.5% (L)
  • Nov. 6: Vs. Illinois — 77.7% (W)
  • Nov. 13: at Iowa — 26.1% (L)
  • Nov. 20: at Indiana — 27.6% (L)
  • Nov. 27: Vs. Wisconsin — 34.8% (L)
Despite projecting 6.1 wins for the Golden Gophers, FPI only views Minnesota as the favorite in five games. However, the metric shows P.J. Fleck’s squad as being in multiple relative toss-up games — including at Purdue, at Colorado and vs. Wisconsin.

According to FPI, Minnesota’s schedule gets much tougher in November. After facing Illinois in the first week of that month, the Gophers have to have Iowa and Wisconsin — the two favorites in the Big Ten West — and Indiana, who could be one of the best teams in the East this season.

Minnesota opens its 2021 season against Ohio State on Sept. 2.

 


I think our odds against Colorado, Purdue and Northwestern are better than shown. I think we are a better team than Colorado (though we have to play them at “altitude,” which will be an adverse factor). I think Purdue and Northwestern will be toss-ups. But I ain’t a computer …
 


It so often seems to me like the Midwest commentariat always rates Purdue as being better than us. Glad we play the games on the field......
 




When we get to the Nebraska game we'll know a lot more about this team and the rest of the League. If we get there at 4-1 (or dare I say 5-0) I bet the predictions will have changed.
 






I honestly have no idea why Colorado is favored.

Simple. We weren't very good last year, they were OK, and the game is at Colorado.

Computers can only process data that is inputted according to algorithms programmed. Nobody is going to pay a bunch of people to input data for 120+ teams in terms of starters returning or lost, significant transfers coming in, experience of players, etc., etc., etc, and pay a team of programmers to work out various algorithms to process all of this data just to satisfy a bunch of fans who don't pay anything for these services. So what we have for now is just a baseline of simple data.

After teams play about 8 games or so next season these simple algorithms and datasets will have significantly more predictive value.
 

Simple. We weren't very good last year, they were OK, and the game is at Colorado.

Computers can only process data that is inputted according to algorithms programmed. Nobody is going to pay a bunch of people to input data for 120+ teams in terms of starters returning or lost, significant transfers coming in, experience of players, etc., etc., etc, and pay a team of programmers to work out various algorithms to process all of this data. So what we have for now is just a baseline of simple data.

After teams play about 8 games or so next season these simple algorithms and datasets will have significantly more predictive value.
As a software dev, I can say that that a team of programmers would be unnecessary. You could get that algorithm done over a weekend. The design going in would probably take a week (by one person) of thinking it over. The longest task, and probably the only task involving other people (no programmers needed) would just be collecting data and debating player value (and other things like playbook vs playbook weight, coaching weight, and location weight etc).
 



I think our odds against Colorado, Purdue and Northwestern are better than shown. I think we are a better team than Colorado (though we have to play them at “altitude,” which will be an adverse factor). I think Purdue and Northwestern will be toss-ups. But I ain’t a computer …
I don't think that altitude will matter. Maybe if the contest was a 10K foot race, but not football.
 




You know what we are expecting when we open this thread on Monday morning, my friend. (We better be undefeated!)
Don't tempt me! (It does sound fun. Could probably use the Godot engine and turn it into a college football sim game...)
 

On a related note, EA will be returning to college football soon. Been too long.
 






FPI has indiana, northwestern, and Colorado overrated

FPI has Wisconsin underrated
 


2021 schedule is a buzz saw that we've seen coming for awhile. Not terribly surprised by this.
 




Top Bottom