ESPN/Allstate Playoff Predictor after Week 2 (TL:DR- makes no sense)

Submitted by RealElonMusk on September 11th, 2023 at 2:24 PM

Playoff Predictor after week 2 

 https://www.espn.com/espn/feature/story/_/page/cfbplayoffpredictor/cfb-playoff-predictor

Hopefully the team sees this because this is some serious Disrespekt!   

Michigan made the playoffs 2 years in a row and this is the best team of the last 3 and Michigan is the 10th most likely team to make the playoffs?   Alabama with a loss to Texas at home is 11% more likely than U of M?  

F&$@ Allstate and ESPN!

 

 

 

 

 

 

 

 

 

 

 

mi93

September 11th, 2023 at 2:29 PM ^

Still way too much love for osu (and psu for that matter).  Once ND beats osu in two weeks, minds will change.  And SC@CU in 3 weeks could be a much bigger season turning point than anyone would have guessed 2 weeks ago.

Sooooo much football to be played still.

Mich1993

September 11th, 2023 at 2:34 PM ^

SC is going to destroy CU.  Without Nebraska's quarterback constantly turning the ball over, NU would have won.  Impressive given the turnover that CU is not terrible, but I'm not convinced they are better than mediocre.  SC's defense is bad, but I don't think it can be that bad.

Perkis-Size Me

September 11th, 2023 at 3:04 PM ^

They look the part so far, but they played a West Virginia team that is almost certainly firing its coach this year, along with a vastly overmatched Delaware team. I'm not sure how much we can accurately decipher from those games. 

I don't think there's any question about whether or not Penn State will be good this year. The question will be whether they're good enough to be better than the 10-2 ceiling they usually hit. We probably won't know until they have to play OSU next month. The Iowa game in two weeks might give us some indications for their offense, though. 

Vasav

September 11th, 2023 at 3:17 PM ^

In their defense though, it's hard to keep track of all the contenders in 5 leagues. We know OSU has looked shaky, but these guys look at stats. They see M only scoring 35 and think "hmmm?" And the running game not being what it was. They can't and don't know context beyond that.

Like, i know there are 8 ranked Pac12 teams and know that CU is a surpsie and Wazzu and Oregon St are nice stories. I couldn't tell you anything about UCLA right now. I know that Bama and LSU both have lost a game against a solid team, I know Ole Miss escaped a solid Tulane. I'm guessing every other SEC West team is undefeated but have no idea. These guys probably at least know some stats about them.

NOLA Blue

September 12th, 2023 at 9:39 AM ^

Ole Miss escaped a Tulane who was playing without their star quarterback.  The SEC is in rough shape when a second string qb at Tulane is giving them a run for their money.  Go Greenwave!

(Also, let's bring Tulane and their research prowess into the B1G.  Conference championship games in NOLA would be fabulous!)

gobluem

September 11th, 2023 at 2:32 PM ^

Why on earth people even LOOK at shit like this so early in the season, much less get bothered by it, is completely beyond me

 

It's been 2 weeks. Nobody ain't played nobody, pawl.  Any sort of computer model is next to worthless before Week 6 or 7. 

1989 UM GRAD

September 11th, 2023 at 3:14 PM ^

I find this stuff interesting...as I find data fascinating...but not sure why we need to get so upset about a predictor that's based on just the first two games of a 13-week season.

shrug.emoji

Blau

September 11th, 2023 at 3:49 PM ^

I'd say you probably enjoy algorithmic equations and weighted value manipulation more so than plain old data as whatever percentages being used by ESPN to begin the process are also somewhat skewed based on perception and expectation. I get that UM's non-conference is pretty bad and # of wins vs ranked opponents will be hard to come by but it seems this tool appears to assume losses for some teams and wins for others before calculating odds without taking recent success into account?

Objectively, why would UM have a 36% chance of beating OSU at home while currently being higher ranked, bringing back more returning players, having beaten OSU the past two years, and having what I assume similar scoring offenses/defenses thus far this year? What's the weighted value it's assessing here or does it just perceive OSU to be the better team? If that's the case, we've seen what happens the past two years.

I get that this sort of thing changes weekly but I agree this tool is only useful the last week or two of the season.

ehatch

September 11th, 2023 at 3:22 PM ^

This thing was trash last year. At one point when Texas had 2 losses, they had Texas as a very high percentage to get into the playoffs (it was during the TCU game). Despite the fact that no team has ever gotten in with 2 losses. They would have needed the following to happen:

Alabama to lose 3 games, since Bama had the head-to-head

Georgia, Michigan, and Ohio State would have had to lose at least 2 games 

TCU would have had to lose another game besides to Texas. 

USC would have had to lose another game

I think Oregon only had 1 loss at the time, meaning they would have had to lose another game. 

Could those things have happened? Sure, but it was nowhere near the percentage that they were suggesting. I would not put a lot of stock into anything that "predictor" had to say, especially this early in the season. 

AZBlue

September 11th, 2023 at 3:42 PM ^

As others have noted, ESPN's fancy metrics are not always accurate -particularly early in the year.

Pretty sure I saw something on twitter that ESPN's FPI system still has OSU and 'BAMA as 1-2 in the country at this point.  (maybe it was reposted from before the weekend?)

 

Side note -- Like others I saw a LOT of Tucker stuff on twitter over the weekend.  I think it worth noting that the ONLY two folks (of note) defending Tucker and/or preaching to "let the process work to reveal the truth" were.........(drum roll)..............................James Yoder and Zach Smith  (YTZS)...  Sad indicator of how bad things are for Coach Tucker.

matt1114

September 11th, 2023 at 4:26 PM ^

4 team playoff is going to be crowded this year, and we are realistically only going to play 2 ranked opponents. I think that's why osu is getting an edge, as they are predicting a ND and PSU win, which will be 2 wins against Top 10 teams. I don't think osu beats ND, however ND will vault into the playoff talk should they win. A lot of this is way too early, as many of these teams play each other so it should sort itself out. Michigan OSU winner likely gets into the playoffs. USC has a TOUGH schedule remaining against ranked opponents, and if they run the table, or even lose 1 of the 6 ranked games, I could see them making it in. If Alabama wins against Tennessee and LSU back to back, that will push them back into the conversation. 

Honestly, we need Rutgers, MSU, or Minnesota to be Top 20 when we play them. This year is going to be so much about "strength of schedule", and we really don't have that strong of one compared to the potential of others when we beat both PSU and OSU.