ESPN Matchup Predictions - 2021 Season

[Golson5]

There is a lot of turnover on the offensive side. You cannot discount that fact. No one, not even the coaches, know how the offense will perform until they are actually in a real game. The defense has plenty of playmakers. How will they perform in a new system under a new coordinator ? I think it will take some time for the team to really gel. Looking at the schedule, I do not see any "gimmies". I think they are going to have work hard and play as error free as possible to have a very good year. I know, "mr. doom and gloom". In this case that is not true. There are too many new faces to make a determination one way or the other. I do think if they can keep the penalties and turnovers to a minimum, they can have a great year. Just have to see how it goes.

There are more than a few gimmies

 

[W8LNXYR]

Your first two statements are contradictory. If you get the math then it should make sense that being favored in every game does NOT necessarily mean there is a > 50% chance to go undefeated.

According to the math, if a team has a 90% chance to win they will beat that team 9 out of 10 times. The extreme likelihood is they will beat that team on the day they face them as well as any other team they have a 90% chance to beat. However, if they play 10 such games in one season the math says they are supposed to lose one of those games which means they are highly unlikely to go undefeated. And that's just the scenario where we are 90% favorites against all opponents. Obviously, those odds get worse once you include games where we are only 60% or 70% favorites. This is why the same FPI predicts ND's record will be somewhere around 8-4 this year even though they have us as favorites in every singular game.

Having said that, whatever metrics ESPN puts into their FPI suck. Their FPI has always been ridiculous, so I think we will do much better than they think we will. Maybe we won't go undefeated but I don't see us losing more than 2 games.

How many games has BK lost in the last 4 years where ND wasn’t the favorite? Sometimes the cumulative value of percentages in a full season doesn’t hold because of external factors, like the head coach.

 

[NotreDame du Lac]

They just recently made their calculated predictions which I think are largely based on their FPI ratings. Definitely some surprises, both good and bad. They have us winning every game, although some of them more closely than I would think. Virginia Tech is considered the most challenging opponent. Overall, we are pretty substantial favorites, as none of the games have a less than 60% of winning, which is encouraging.

Here is the game-by-game breakdown indicating our chances of victory:

FSU - 71.9%
TOL - 90.2%
PUR - 85.9%
WIS - 61.4%
CIN - 73.2%
VT - 61.0%
USC - 78.7%
UNC - 65.3%
NAVY - 96.7%
UVA - 66.6%
GT - 88.6%
STAN - 72.5%

Chances of going undefeated - 3.2%

Damn. Favored in every game but 3% chance to go undefeated. I can see why I’m not a math guy.

 

[NotreDame du Lac]

I was surprised that our chances aren't higher vs FSU. They have been on a miserable run, although playing them on the road in Primetime will have them stoked and bringing their A-game. And then having Navy as by far our easiest opponent struck me as odd, given how we sometimes struggle with them. I expect USC to be a tougher battle...maybe even the toughest game on the schedule. Also, I remember when we were slated to play Wisconsin last season, we were pretty significant underdogs, so I'm not entirely sure what changed their mind so drastically this season when a lot of experts consider this year's team to take a bit of a step back with so many key personnel losses.

It'll be interesting to see how these predictions change throughout the year.

They mean nothing. We would’ve rolled Wisconsin last year. These people are as smart as this forum.

 

[tkirish]

There are more than a few gimmies

yeah , poor choice of words on my part.

There are more than a few gimmies

Golson, you are correct. My "post" regarding "gimmies" is a bit of a stretch. Still, there are some tough games and even though some of the teams may be listed as "down", the "new irish" are kind of unknown. For me, this puts them in the "uncertain' category. Again, I am not say definitively that they will be very good or really bad. I'm just saying more than likely, they will have some growing pains the first 4 or 5 games and no game will be an automatic win. We shall see how it plays out. Go Irish!!!!!!!!!!!

 

[Bumpdaddy]

How many games has BK lost in the last 4 years where ND wasn’t the favorite? Sometimes the cumulative value of percentages in a full season doesn’t hold because of external factors, like the head coach.

I was just explaining the math. When you say "external factors" you are describing a metric that might or might not be included in someone's math formula. They almost certainly exist in some. There are FPI-like formulas out there that estimate the relative strength of each opponent and a coach's or program's normal success rate against such programs when they come up with their odds for each game. I don't know if the FPI does that (I doubt it) and if they do, I don't know how much weight they put on that particular metric when calculating the overall odds for each game. Regardless, the math is the math, which is why if your metrics say a team has a 90% chance to win against all 12 teams they face on their schedule then the math MUST also say that team is highly UNLIKELY to go undefeated.

What I'm saying is, if you don't like the FPI's estimates for ND's season, it's not the math that is wrong it is their metrics. Maybe ND should be 99% favorites in most of their games. I'm not sure what the actual percentage should be but I do know ESPN's FPI metrics have sucked since inception.

 

[deadmanramsey]

Damn. Favored in every game but 3% chance to go undefeated. I can see why I’m not a math guy.

Yeah, that's how it works given the percentage chance of each game. Here is the distribution of simulated outcomes I ran. 9 and 10 are the most likely number of wins. 325 of the time it came out with a perfect season (3.25%) which is what was expected. Obviously all this comes down to assuming the FPI predictor is reliable which is very debatable.

 

[spenna]

I think the odds-makers don't care about history but ESPN might. We have ND playing on the road in Florida, first game of the season coached by Brian Kelly. Those aren't favorable factors.

Especially with B B passing. They will be up for that game .

 

[spenna]

I was just explaining the math. When you say "external factors" you are describing a metric that might or might not be included in someone's math formula. They almost certainly exist in some. There are FPI-like formulas out there that estimate the relative strength of each opponent and a coach's or program's normal success rate against such programs when they come up with their odds for each game. I don't know if the FPI does that (I doubt it) and if they do, I don't know how much weight they put on that particular metric when calculating the overall odds for each game. Regardless, the math is the math, which is why if your metrics say a team has a 90% chance to win against all 12 teams they face on their schedule then the math MUST also say that team is highly UNLIKELY to go undefeated.

What I'm saying is, if you don't like the FPI's estimates for ND's season, it's not the math that is wrong it is their metrics. Maybe ND should be 99% favorites in most of their games. I'm not sure what the actual percentage should be but I do know ESPN's FPI metrics have sucked since inception.

Metrics are deceiving, just ask the N Y Yankees