Football Record Chances Using FPI

[sleepy dawg]

If you look at FPI given by ESPN, it gives you the chances of winning each game. It also gives the expected win/loss total based on the projections, which is 8.4 wins to 3.7 losses, given us the likely scenario of a 8 or 9 win season (excluding post season). I wanted to know more about our ceiling and floor though, so I ran these numbers through 100 seasons of this. Obviously there's a lot of assumptions to be made here, so don't treat this like an exact science.

Here's the breakdown out of 100 seasons at these projected numbers (http://www.espn.com/college-football/team/fpi?id=344&year=2018).



Record
# of occurrences


12-0
1


11-1
5


10-2
10


9-3
25


8-4
32


7-5
16


6-6
8


5-7
3




https://image.ibb.co/gUTtEJ/chart.png

[msstate7]

I've thought 9-3/10-2 since our bowl win. Sticking with that. If we go 8-4 without major injuries, Moorhead and this staff screwed up imo

[MetEdDawg]

Here's the question. How much is prior performance taken into this? I would be interested to know the weight on how much overachieving/underachieving previous win total expectations. I have to feel that as a betting person if someone showed me the talent of this team and I had to say this team either goes 7-5 or 9-3, I'm taking 9-3.

Better chances we lose to Bama, Auburn, and LSU, or better chances we lose to Bama, Auburn, LSU, Kansas State, and TAMU? I think the former is much more likely.

I'll also add that I'm shocked at the amount of emphasis being placed on the head coaching change. It happened very early in the process and everyone said home run hire. Now everyone is saying well subtract a win because new head coach new scheme new staff. I wonder if the media realizes how much worse our previous staff, at least on paper, was to this staff in terms of experience and cache.

[TUSK]

I've thought 9-3/10-2 since our bowl win. Sticking with that. If we go 8-4 without major injuries, Moorhead and this staff screwed up imo

This stretch is y'all's whole season, IMO:

9/29 Florida
10/6 Auburn
10/13 BYE
10/20 @ LSU
10/27 A&M

I think those 4 games may be the difference between 11-1 and 7-5...

[msstate7]

This stretch is y'all's whole season, IMO:

9/29 Florida
10/6 Auburn
10/13 BYE
10/20 @ LSU
10/27 A&M

I think those 4 games may be the difference between 11-1 and 7-5...

I think we split LSU/auburn, and lose to bama. Won't be surprised if we lose all 3 of those games though

[Lumpy Chucklelips]

Regarding coaching change affect....all one needs to do is look at our bowl game. We were practically without a coach and the team with the 2nd string QB defeated Louisville and a Heisman Trophy winner. I almost wish we had practiced and played without a head coach against OM......notice I said "practiced and played". I know we played without one...

[Lumpy Chucklelips]

This stretch is y'all's whole season, IMO:

9/29 Florida
10/6 Auburn
10/13 BYE
10/20 @ LSU
10/27 A&M

I think those 4 games may be the difference between 11-1 and 7-5...


I know I'll catch flack for this, but I've seen it too many times over my 55 years.....Not just with MSU, but with a lot of college teams. But I'm worried about Kentucky the week before Florida. We better not be looking past them to Florida or we'll be playing Florida for more than just wanting to beat our previous coaches. I don't care who you are....you look past an SEC team and you can have your ass handed to you. We need to be focused and take care of business that week.

[sleepy dawg]

Here's the question. How much is prior performance taken into this? I would be interested to know the weight on how much overachieving/underachieving previous win total expectations. I have to feel that as a betting person if someone showed me the talent of this team and I had to say this team either goes 7-5 or 9-3, I'm taking 9-3.

Better chances we lose to Bama, Auburn, and LSU, or better chances we lose to Bama, Auburn, LSU, Kansas State, and TAMU? I think the former is much more likely.

I'll also add that I'm shocked at the amount of emphasis being placed on the head coaching change. It happened very early in the process and everyone said home run hire. Now everyone is saying well subtract a win because new head coach new scheme new staff. I wonder if the media realizes how much worse our previous staff, at least on paper, was to this staff in terms of experience and cache.

Well even based on the numbers above, there's a good bit higher chance of 9-3 than 7-5, but both seem like pretty realistic chances.

[bostondawg]

How'd you do this? Is this a Monte Carlo simulation?

[MetEdDawg]

Well even based on the numbers above, there's a good bit higher chance of 9-3 than 7-5, but both seem like pretty realistic chances.

Honestly I don't think 7-5 is very realistic considering the talent we have returning. On paper we have a better team than last year. Last year's team went 9-4. To me the metrics have to be based on our entire history, not just the last decade. I have a hard time believing that 16% of the time the team as its currently constructed would go 7-5

[sleepy dawg]

How'd you do this? Is this a Monte Carlo simulation?

I wrote an application that simulated the season 100 times where our chances to win were based solely on FPI percentages. Then I used a random number to determine if that game was won or not.

For example, FPI says we have a 99.7% chance to beat Stephen F. Austin. That means if we played them 1000 times, we would expect to lose about 3 times of those 1000. So if the random number for that iteration was 0-996, then it was a win. If the number was 998-1000, then we lost.

[sleepy dawg]

Honestly I don't think 7-5 is very realistic considering the talent we have returning. On paper we have a better team than last year. Last year's team went 9-4. To me the metrics have to be based on our entire history, not just the last decade. I have a hard time believing that 16% of the time the team as its currently constructed would go 7-5

I can't say exactly what the metrics are based on. I don't think ESPN releases how they come up with those exactly, but it certainly isn't based on our entire history or even the last 10 years. If anything its mostly based on what we did last year and what we have coming back this year. I think if everything goes according to plan, then you're right, but predicting sports is near impossible. Things don't always go according to plan. Not only that, you are only focusing on how good we are, not how good our opponents may be. FPI certainly isn't going to be right, no system is, but it's probably as good of a preseason prediction model there is which is why I used it.

[Joe Schmedlap]

Peter f'n Sirmon ��


Regarding coaching change affect....all one needs to do is look at our bowl game. We were practically without a coach and the team with the 2nd string QB defeated Louisville and a Heisman Trophy winner. I almost wish we had practiced and played without a head coach against OM......notice I said "practiced and played". I know we played without one...