ESPN FPI Predicts 8.3 Wins for 2017 Season
ESPN's FPI Predicts 8.3 wins for 2017 season
"...Projected results are based on 10,000 simulations of the rest of the season using FPI, results to date, and the remaining schedule."
1. OSU 11.8-1.0
2. Bama 10.6-1.9
3. Oklahoma 11.0-1.9
4. FSU 10.4-2.2
.
8. PSU 9.9-2.2
.
10. Wisc 10.6-2.2
.
17. Michigan 8.3-3.7
" If you’re wondering why the totals don’t add up to whole numbers – that’s because it factors in the odds of each team reaching and winning/losing the Big Ten title game." (collegespun)
April 18th, 2017 at 12:23 PM ^
how does the team adjust to being the "hunted" as opposed to "coming completely off the radar to win"?
Momentum and believing in onself is an INCREDIBLE thing with a bunch of coming-of-age 18-to-22 year olds. Not just in college football, but in life.
Penn State rode that wave in 2016. The week after the U-M loss, PSU trailed Minnesota at home by 2 scores, midway through the 3rd quarter, on a rainy, cloudy dark late afternoon. The home crowd was grumbling, the team was sleep-walking.
But PSU hits a 80-yard-pass on 3rd-and-10 out of nowhere. The crowd gets back into it, it goes to OT, PSU pulls it out. From there, a home win over Maryland. A game where they played OSU very well (not the statistical blow-out some people make it out to be) - the ball bounces one way on a blocked FG, and PSU gets the upset win. NOW they got momentum and are believing in themselves! They just keep on winning and end up in Indianapolis and then Pasadena.
That "magic" of 2016 was awesome, but it will be impossible to exactly replicate in 2017. We'll see how it goes .......
But we're not debating last season, we're debating 2017. And believing that Penn State was overrated and lucky last year is important because it makes it less likely their success will continue this year.
April 18th, 2017 at 11:39 AM ^
Where do you come up with our mismatch in their secondary? We have barely broken in our outside receivers and tight ends.
April 18th, 2017 at 11:44 AM ^
April 18th, 2017 at 12:37 PM ^
Isn't their top rated db out allready. Talking up crybaby lamont wade. After fifteen fade routes in a row to Tarik Black, they will understand why 5'9" is not good for a db. You covered enough mantis. We have no business losing to a team we beat by thirty nine points. I have allready debated, but I feel as far as starting talent we are tops in the big ten. When Peters is given his chance he will quiet talk about Qb's like Barett and Mcsorely, who will never sniff the nfl because they just don't rate. Am I the only one who Franklin just annoys the hell out of?.
April 18th, 2017 at 11:29 AM ^
Harbaugh already beat them in Happy Valley ... even with the penalties being 22 - 3 against us.
Go Blue!
April 18th, 2017 at 12:28 PM ^
To this day, that is the only road game I've been to. To put it lightly, it was pretty disappointing.
in 2015 at home by OSU, but that outlier didn't stop us from being optimistic about The Game the following year because our performance in all our other games in 2015 indicated we were a lot better than that our performance in that game. Indeed we outplayed them at their place in 2016. Same as PSU last year. Single game outcomes can be highly variable and aren't a great indicator of overall quality. We also lost a ton of players from the team that detroyed them in 2016.
PSU won the freaking Big Ten last year and returns almost everyone and is a really hard place to play. We're the underdogs in that game based on all currrenlty available information.
Our team that wrecked them last year has mostly moved on.
April 18th, 2017 at 11:46 PM ^
April 18th, 2017 at 10:54 AM ^
I'll take the over on this one.
April 18th, 2017 at 10:54 AM ^
April 18th, 2017 at 10:54 AM ^
My advanced analytics predict: these numbers will change a lot.
April 18th, 2017 at 10:55 AM ^
I really don't see four losses. There aren't really more than four loseable games in UF, PSU, UW, OSU. MSU is a total mess, Indiana is now DeBord, Purdue is the other crossover, etc. There is no "night game at Kinnick" obvious trap situation on the schedule this season.
If we lose all of those four that's a dissapointment, period, however I can't imagine it happening. There is too much talent and three of four "loseable" games are in the back half of the schedule which is advantageous to a young team. Think we'll surprise to the positive.
April 18th, 2017 at 11:02 AM ^
If you go by chalk and assuming Zaire doesn't go to UF then I think UM will be favored in 9 games. It's time to see some of Stanford Jim Harbaugh where he was able to punch USC in the face despite being a massive underdog. We will need some of those "surprises" this year - last year I think we were favored in every game but the last.
April 18th, 2017 at 11:07 AM ^
If we beat PSU we'll no doubt be favored going into Madison as well. National opinions on this team are going to vary wildly week by week, starting right away.
April 18th, 2017 at 11:18 AM ^
FPI has Michigan favored heavily in 8 games with Florida being close to a coin flip.
The other three: at Penn State, at Wisconsin, and Ohio State aren't too rosy.
Which, if we're being honest feels about right. Going to need to see them perform well on the road for once before I feel like they can roll into Happy Valley or Camp Randall and pull out a win.
For those curious, the preseason FPI (in August) had Michigan at 10.2-2.2 last season.
April 18th, 2017 at 11:50 AM ^
I still think Penn State is just enormously overrated. I am just not understanding the hype. McSorley arm punted his way through the 2nd half of last season. They were not overly impressive
you guys keep talking about arm punts like it's a bad thing. If you have dynamic, play-making WRs who can out-jump smaller DBs, it's actually a good idea to throw it up to them and let them make a play. Unfortunately, for as good as Darboh and Chesson were, they were not the type of WR where you could make a living arm-punting all day. Godwin and Hamilton were good WRs. Gesicki, even though he couldn't block, was a threat in the passing game (48-679.) And you had to play them honestly because of Barkley. We shut them down because we had arguably the #1 or #2 defense in the country. Will 4 new secondary guys be able to defend the arm punt like Lewis, Stribling, Wilson, and Thomas did? Time will tell.
April 18th, 2017 at 11:52 AM ^
I'm surprised that UF is a favorite, albeit a small one, against UM in that opening game. I feel pretty good about UM's chances to win that game. I'd put it at about 63.7265% for UM.
OSU will be good, not doubt about it, but 17.8% for UM at home? That seems crazy low. Survey says: 41.4823% chance of winning that one.
Correct EV wins for 2017: 9.2.
April 18th, 2017 at 12:33 PM ^
How if the fuck is Rutgers, at home, ONLY a 95.4% chance of a win?
Seriously?
That should be, like 110% chance of us winning, if not 201% chance.
And don't get me started on the Sparty game.....that one should be 3-90% chance of us winning.
April 18th, 2017 at 12:57 PM ^
Made the bold prediction yesterday that we would be favored in every game. So I can clear it up for some who don't understand betting lines change as the season goes on. I think we will be favored over Florida by gametime, but we can't control this one. I feel we will be undefeated heading into the Penn State game and they will have at least one loss. Same goes for Wisconsin and if we beat them and both us and osu are undefeated I think we get the nod at home. Let's see fpi numbers after the Florida Game.
Frustrating to see shit like that vs OSU at home in 3rd year of Harbaugh. Sub 20%.
April 18th, 2017 at 11:49 AM ^
We have a good shot at UF and PSU. We have holes, but so do they.
Wisconsin looks extremely tough. Fortunately, it's the game we can most afford to lose.
It's likely still to come down to Michigan and Ohio State for the Big Ten East. And all the refs will be from Michigan this time.
the worst case-number of wins, not the expected number. I could see a trip-up at Maryland (I guess?) but I'm grasping to see another losable game in there. And it'd be unlikely to loss all four of the tough ones, even though we'll probably be underdogs in all of them except maybe UF.
I'd say 9 is the most likely but 10 is probably more likely than 8.
April 18th, 2017 at 10:58 AM ^
I dont see 4 losses anywhere on that schedule. We'll know very early one way or the other...
April 18th, 2017 at 11:06 AM ^
MSU at 59 and 5.2-6.8
April 18th, 2017 at 11:09 AM ^
Somewhere between blehhh and blehhhhh
April 18th, 2017 at 11:09 AM ^
April 18th, 2017 at 11:14 AM ^
April 18th, 2017 at 11:14 AM ^
Florida game will set the tone and the Penn State game will determine if we have a great or good season.
Beat Florida.
April 18th, 2017 at 11:15 AM ^
As a statistician, let me tell you some things about statistical models used to predict future events...
They are almost always "backward looking," meaning that they are trying to predict the future by how things have gone in the past. This can be effective on the average, but often fails mightily for individual cases, as anyone who plays the stock market (or knows about "ecological fallacies") can tell you.
Indeed, the model they are using (I am making an assumption here, but am probably right) is most likely a linear model. That means it has some measure of error, probably in the neighborhood of 25-50%, in fact. This is NOT an error in how well they predict the future (that error is likely much larger), but rather an error in how well past data fits the "predicted win line" they estimate.
This is important because they are probably predicting a win total for us of 8.3 +/-1.5 wins. In other words, their prediction has a floor and a ceiling, and that ceiling may be something much closer to what we actually expect, like 9.8 wins. You never want to talk about predictive error in models, because most people don't understand it, and most people would take that to mean the prediction is bunk. But in this case, it should be remembered, especially because I assume (as do most) that Harbaugh will overperform most years.
Not sure why I've written all this, nor if anyone cares, but here it is...
I appreciate it fauxmo, but god I hated stats class, but not as much as finance. Word of advice. Do not take these classes when your on oil based medicine. LOL.
April 18th, 2017 at 11:17 AM ^
April 18th, 2017 at 11:35 AM ^
Can't unsee last year's slaughter. And, I think they are the media darlings this year. FL is a different story -- they seem pretty legit, but I feel that game is a real toss-up
April 18th, 2017 at 11:49 AM ^
year the final polls did unsee 49-10.
Weird. But it did happen.
April 18th, 2017 at 11:55 AM ^
their offense is putrid. Even if Malik Zaire transfers there, it is currently less than 5 months till the game and he's not on campus. Remember how Rudock was against Utah? Zaire isn't just going to learn the offense and be so awesome in Game 1 that it overcomes how offensively flaccid Florida has been recently. It's going to be tough for them to score points.
The counter-point to that is that this is Game 1 for us with essentially a new secondary. This is as bust-prone as that secondary will be all season.
If PSU was in Ann Arbor, I'd expect to kill them again. Happy Valley is tough to win at though
April 18th, 2017 at 11:57 AM ^
If the PSU game was in Ann Arbor and not Happy Valley, I would not even be concerned.
But it's not. They play on a different level in white out night games.
OSU despite all of its success over the last decade has always struggled there. We won in 2006 with a historically good defense, but it was still a struggle. That was our only white out night game win at PSU.
Anybody who does not think that home night games against a rival don't provide an advantage is crazy. Penn State, generally mediocre, is a prime example.
we lost to Iowa in a night road game last year. PSU will be a lot more difficult.
April 18th, 2017 at 11:38 AM ^
Nope. They'll probably be solid or good, but I don't see them as juggernauts either.
April 18th, 2017 at 11:17 AM ^
not hard to find 3-4 losses? where?
April 18th, 2017 at 11:17 AM ^
I will bet all of my 401k with who ever put Michigan at 8 wins. I will take the over for sure. Michigan will have a top five defense and the offense will be better especially in the run game. If we start Peters I will give the idiots odds.
April 18th, 2017 at 11:26 AM ^
How big is your 401K? I just read that the average 40-something American has only a few thousand dollars in retirement savings, so if you are younger and average, that means you are willing to bet NOTHING on this outcome... :-D
April 18th, 2017 at 11:19 AM ^
A tie with Rutgers?
April 18th, 2017 at 11:40 AM ^
A tie with Rutgers is actually worth 1.4 losses.
April 18th, 2017 at 11:20 AM ^
Am I a homer?. Yes I am, but this is the fastest Michigan team I have ever seen and I have been loyal for almost fifty years. Go Blue.
April 18th, 2017 at 11:25 AM ^
...ESPN loses between 8.3% and 83% of it's viewership this year.
April 18th, 2017 at 11:26 AM ^
I am a pessimist and I would still take the over. There is no way they are going to lose all 4 big games (Florida, Wisconsin, PSU, OSU), especially with Harbaugh running the show.