Welcome to the forum 👋, Visitor

To access the forum content and all our services, you must register or log in to the forum. Becoming a member of the forum is completely free.

  • Welcome to The Platinum Board! We are a Nebraska Cornhuskers news source and community. Please click "Log In" or "Register" above to gain access to the forums.

An unbiased look at our ranking and progress

kenyanfeline

Pussy Patrol
Insider
Elite Member
Messages
1,197
Likes
11,974
ESPN FPI

This is probably TLDR for many. Sorry in advance.

I am biased and so are you. I’m typically an optimistic sunshine pumper. Some of you are pessimistic taters. Hopefully, if we look at ESPN FPI, we can find some middle ground by looking at some unbiased numbers. The old saying “the truth is always somewhere in between” is something people on both sides need to remember.

ESPN FPI is an algorithm that “represents how many points above or below average a team is.” It’s thought to be one of, if not the most accurate algorithm for predicting and evaluating college football games.

1. First I decided to simply take a look at FPI to see what it tells us about how good of a team we are, who we beat/lost to, and the progress we have/haven’t made since 2017. Keep in mind this is only one metric and is what computers say about us. Your eyes may tell you something else, but I’m leaving that out of it.

Many on here and nearly everyone over on Tater Island says we are below average and one of the worst P5 teams in the country. Kansas level.

But here’s the deal. We are ranked 45 nationally with an FPI of 4.9. That means it predicts that if we play the most average team in the nation, we would be projected to win by 5 points. I was surprised by this based on all of the chatter

There are 9 non-P5 teams ranked ahead of us.

There are 36 P5 teams ranked ahead of us.

Out of 65 P5 schools, that puts us pretty much right in the middle.

Bottom Line: We are average. We are not good. We are not bad. We are average.

2. Second, I took a look at who we played and the outcomes of those games through the scope of FPI. Remember, we were loo ranked 41st. I bolded all teams that were within +/- 20 of us in FPI ranking, as those are the teams we should be competing with.

OpponentFPI RankOutcome
OSU2L by 35
NW23L by 8
PSU21W by 7
ILL79L by 18
Cockeye11L by 6
Purdue 51W by 10
Minn41L by 7
Rutgers70W by 7

In the games I bolded (+/-20 FPI ranking), we went 2-3 and had a +/- of -0.8 ppg. If you include Illinois and Rutgers (outside of +/-20 FPI ranking), we end up 3-4 with a +/- of -2.1 ppg.

To me, this table also says we are average. We lost to teams ranked ahead of us by 35, 8, 7, and 6 points, respectively. We beat teams ranked behind us by 10 and 7 points. We beat PSU, who was ranked 24 spots ahead of us, by 7. And we lost to a terrible Illinois team ranked 28 spots behind us by 18.

The Illinois game seems to be an anomaly. Every other game seems to indicate that we are average. We kept games close against all other teams not named OSU.

Bottom Line: We did not underperform outside of one game. We did just about exactly how we should have done. If you take a step back and rinse the bad taste of the Illinois game out of your mouth, we were decidedly average this season.

3. Now, a lot of people have strong feelings regarding if we are making progress or not. So finally, I thought I would track our FPI from 2017 to 2020.

YearFPI (pts +/- average)Rank
2017 (MR)-2.170
2018 (SF)+3.157
2019 (SF)+4.547
2020 (SF)+4.941

Now, the two things you are looking at on this chart are FPI and FPI rank. Again, FPI is a predictor of if we played the most average team in the country, what we would win or lose by. FPI rank is simply our ranking out of all FBS teams.

Since Mike Riley’s departure in 2017, our FPI has increased every year, and so has our ranking. Has it increased by as much we had hoped? No. But it has still increased every year.

Bottom line: We ARE making progress. As much as we wanted? Again, no. But progress is being made nonetheless.


4. Finally, from a non-objective point of view, I will talk from a biased perspective about how close I think we are.

Think about how many games we lost this year by 1-2 scores. If you are like me, you can probably point to 2-3 plays in each of those games that lost it for us.

I hear talk all the time about how we are 1 good QB from winning this game. We are 1 good WR from winning that game. We are a healthy running back from winning another game. We are 2 penalties or 1 turnover from winning another, et

That notion is absolutely correct. We are in fact one or two plays a game from being 5-2 with losses to OSU and Illinois. But that isn’t how life works and we can’t get those plays or games back.

When this team turns the corner, it isn’t going to be because of one player. It isn’t going to be from playing mistake-free football. It’s going to be about several different parts of our program that have held us back getting just a little bit better, all at the same time.

That isn’t going to be fixed in the transfer portal or in the recruiting rankings. That is going to be fixed in the off-season with dedication and reps. AM or whoever the QB is next year is going to get a little bit better at ball security. Our running backs are going to get a little more consistent. Our young WR’s are going to learn the playbook a little better and get on the field a little more. Our pass rush is going to home 1-2 times more per game. And Frost is going to get a little better at managing games and play calling.

All of those things have been improving. Again, maybe not as much as we want, but improving nonetheless. We’ve had setbacks like Mo Washington and JD leaving, AM regressing, etc. But despite all of that, we HAVE improved.

We’re all tired of hearing Frost say we are “close” to turning the corner. I know it’s hard to see from our perspective. But take a step back, look at the whole picture. I believe him that we are close to turning the corner, and you should too. It’s going to be a lot of fun when we do and I hope you are all still here to witness it.
 
Last edited:
Also, my understanding of FPI is that it does some weighting of preseason expectations when it models and, while I don't do a lot of Bayesian statistics personally, I'm pretty sure that the priors you assume can pretty heavily skew what your model spits out. Plus this ignores the fact that you almost always want to report some measures of errors. Are these fpi numbers the mean of the posterior distribution? What's variance?

Here is the fpi explainer from espn fwiw: https://www.espn.com/blog/statsinfo/post/_/id/122612/an-inside-look-at-college-fpi
 
Last edited:
ESPN FPI

This is probably TLDR for many. Sorry in advance.

I am biased and so are you. I’m typically an optimistic sunshine pumper. Some of you are pessimistic taters. Hopefully, if we look at ESPN FPI, we can find some middle ground by looking at some unbiased numbers. The old saying “the truth is always somewhere in between” is something people on both sides need to remember.

ESPN FPI is an algorithm that “represents how many points above or below average a team is.” It’s thought to be one of, if not the most accurate algorithm for predicting and evaluating college football games.

1. First I decided to simply take a look at FPI to see what it tells us about how good of a team we are, who we beat/lost to, and the progress we have/haven’t made since 2017. Keep in mind this is only one metric and is what computers say about us. Your eyes may tell you something else, but I’m leaving that out of it.

Many on here and nearly everyone over on Tater Island says we are below average and one of the worst P5 teams in the country. Kansas level.

But here’s the deal. We are ranked 45 nationally with an FPI of 4.9. That means it predicts that if we play the most average team in the nation, we would be projected to win by 5 points. I was surprised by this based on all of the chatter

There are 9 non-P5 teams ranked ahead of us.

There are 36 P5 teams ranked ahead of us.

Out of 65 P5 schools, that puts us pretty much right in the middle.

Bottom Line: We are average. We are not good. We are not bad. We are average.

2. Second, I took a look at who we played and the outcomes of those games through the scope of FPI. Remember, we were loo ranked 41st. I bolded all teams that were within +/- 20 of us in FPI ranking, as those are the teams we should be competing with.

OpponentFPI RankOutcome
OSU2L by 35
NW23L by 8
PSU21W by 7
ILL79L by 18
Cockeyes11L by 6
Purdue 51W by 10
Minn41L by 7
Rutgers70W by 7

In the games I bolded (+/-20 FPI ranking), we went 2-3 and had a +/- of -0.8 ppg. If you include Illinois and Rutgers (outside of +/-20 FPI ranking), we end up 3-4 with a +/- of -2.1 ppg.

To me, this table also says we are average. We lost to teams ranked ahead of us by 35, 8, 7, and 6 points, respectively. We beat teams ranked behind us by 10 and 7 points. We beat PSU, who was ranked 24 spots ahead of us, by 7. And we lost to a terrible Illinois team ranked 28 spots behind us by 18.

The Illinois game seems to be an anomaly. Every other game seems to indicate that we are average. We kept games close against all other teams not named OSU.

Bottom Line: We did not underperform outside of one game. We did just about exactly how we should have done. If you take a step back and rinse the bad taste of the Illinois game out of your mouth, we were decidedly average this season.

3. Now, a lot of people have strong feelings regarding if we are making progress or not. So finally, I thought I would track our FPI from 2017 to 2020.

YearFPI (pts +/- average)Rank
2017 (MR)-2.170
2018 (SF)+3.157
2019 (SF)+4.547
2020 (SF)+4.941

Now, the two things you are looking at on this chart are FPI and FPI rank. Again, FPI is a predictor of if we played the most average team in the country, what we would win or lose by. FPI rank is simply our ranking out of all FBS teams.

Since Mike Riley’s departure in 2017, our FPI has increased every year, and so has our ranking. Has it increased by as much we had hoped? No. But it has still increased every year.

Bottom line: We ARE making progress. As much as we wanted? Again, no. But progress is being made nonetheless.


4. Finally, from a non-objective point of view, I will talk from a biased perspective about how close I think we are.

Think about how many games we lost this year by 1-2 scores. If you are like me, you can probably point to 2-3 plays in each of those games that lost it for us.

I hear talk all the time about how we are 1 good QB from winning this game. We are 1 good WR from winning that game. We are a healthy running back from winning another game. We are 2 penalties or 1 turnover from winning another, et

That notion is absolutely correct. We are in fact one or two plays a game from being 5-2 with losses to OSU and Illinois. But that isn’t how life works and we can’t get those plays or games back.

When this team turns the corner, it isn’t going to be because of one player. It isn’t going to be from playing mistake-free football. It’s going to be about several different parts of our program that have held us back getting just a little bit better, all at the same time.

That isn’t going to be fixed in the transfer portal or in the recruiting rankings. That is going to be fixed in the off-season with dedication and reps. AM or whoever the QB is next year is going to get a little bit better at ball security. Our running backs are going to get a little more consistent. Our young WR’s are going to learn the playbook a little better and get on the field a little more. Our pass rush is going to home 1-2 times more per game. And Frost is going to get a little better at managing games and play calling.

All of those things have been improving. Again, maybe not as much as we want, but improving nonetheless. We’ve had setbacks like Mo Washington and JD leaving, AM regressing, etc. But despite all of that, we HAVE improved.

We’re all tired of hearing Frost say we are “close” to turning the corner. I know it’s hard to see from our perspective. But take a step back, look at the whole picture. I believe him that we are close to turning the corner, and you should too. It’s going to be a lot of fun when we do and I hope you are all still here to witness it.
Good stuff! Thanks
 
Of course, most genocides have been the result of adhering to an ethos that all sides of an issue have validity. So there's that...

Now, to the substance...

FPI is pretty universally dragged as a model, but I appreciate this approach and desire to use objective analysis, OP. All models have inherent bias though, and some do a lot better at minimizing it than others.

I think the point that your analysis misses is that:
1. Saying we are close and
2. Saying we'll never get over the hump with Frost due to the absurd ways his teams find to shit all over themselves

are not incongruous thoughts. These things can both be true and really, that's what's so damn frustrating.

It also points to the fact that for all our warts, our name still does hold some sway with regards to acquiring talent. Even if we assume for a second that FPI is a decent metric (highly debatable), we still track below our supposed talent level. Of course, recruiting rankings are arguably (definitely) worse wrt bias, bullshit, and variance, but we do generally have more talent on paper than most of our opponents.

The ultimate question of is that talent being properly developed and then put in position to succeed is sort of open at this time, but that door is quickly closing with each season. I still can't get beyond that sequence the led to Wandale's first TD. If it were an aberration, fine, no problem. It was expected though...and that's a major major problem.

We can be an average football team with below average results. In fact, that's exactly what we are. At some point, the guy in charge has to be accountable for that. Should be from Day 1. Most certainly should be at the end of Year 3.
 
But if my eyes tell me one thing and yours tell you another it can be a useful tool to find some middle ground.

Which was my point.
Oh most definitely, which is why models are developed and used. Still have to be super careful when using them to make inferences. Garbage in, garbage out. You might think a model is getting you closer to some middle truth, but not all situations have one and most models are based on finding such an assumed reality.
 
But if my eyes tell me one thing and yours tell you another it can be a useful tool to find some middle ground.

Which was my point.
On a separate related point, which is foundational to your argument, you've set up a false dichotomy and made the assumption that the biases that lead to these two potential outcomes are equivalent AND that there is a middle truth. That's a pretty flawed assumption in my (yes, biased) view and also a biased analysis in its own right without any consideration for the FPI model biases and assumptions.
 
The unbiased algorithm that is calculated by a computer not capable of feelings and subjective thought says we are average, so I’m gonna say we are average.
The algorithm doesn’t get the benefit of actually watching the product in action. Real people who do get to see the games, who aren’t blinded by fanaticism, see the product for what it is - below average.

Hopefully that changes, but I have my doubts Frost will be the one to get it done. Three years of being consistently out coached with players repeating the same mistakes over and over has led me to this opinion.
 
The algorithm doesn’t get the benefit of actually watching the product in action. Real people who do get to see the games, who aren’t blinded by fanaticism, see the product for what it is - below average.

Hopefully that changes, but I have my doubts Frost will be the one to get it done. Three years of being consistently out coached with players repeating the same mistakes over and over has led me to this opinion.
I agree. Wonder what the algorithm thinks about continual mistakes by this staff like having only 10 defensive players on the field for a 2 Pt Conversion?
 
I agree. Wonder what the algorithm thinks about continual mistakes by this staff like having only 10 defensive players on the field for a 2 Pt Conversion?
Timeouts after kick offs. Throwing the ball 30 times against a team that can’t stop the run. Inability to get your athletic WRs on the field. Special teams as a whole. Etc. etc.
 

Theme customization system

You can customize some areas of the forum theme from this menu.

  • Wide/Narrow view

    You can control a structure that you can use to use your theme wide or narrow.

    Grid view forum list

    You can control the layout of the forum list in a grid or ordinary listing style structure.

    Picture grid mode

    You can control the structure where you can open/close images in the grid forum list.

    Close sidebar

    You can get rid of the crowded view in the forum by closing the sidebar.

    Fixed sidebar

    You can make it more useful and easier to access by pinning the sidebar.

    Close radius

    You can use the radius at the corners of the blocks according to your taste by closing/opening it.

  • Color combinations cannot be used

    Color combinations are not available to you, this area may be restricted by administrators. Please contact the administrator for more information.

    Backgrounds are not available!

    Background checks are not available to you, this area may be restricted by administrators. Please contact the administrator for more information.

Back