NU and Football Outsiders FEI Ratings

[ColoradoHusk]

Mike'l Severe sent out some tweets this morning regarding Nebraska's Fremeau Efficiency Index (FMI) according to Football Outsiders. If you aren't familiar with Football Outsiders, they are basically a website that does in-depth and innovative statistical analysis for NFL and college football.

Here is a link to what I am referencing in this post: http://www.footballoutsiders.com/stats/fei

Anyway, according to the FEI, Nebraska is the 34th most efficient team in CFB. Given the FEI rating, NU would have expected to win 8.5 FBS regular season games. With NU going to finish at 5 or 6 FBS wins in the regular season, how do you view this season? Was NU just unlucky to lose games against BYU, Illinois, etc.? Does Riley deserve credit for having an "efficient" team, but deserves criticism for "losing games that NU should have won". Does this efficiency rating show that NU isn't as "untalented" as many people have thought?

There are a few more interesting nuggets from the link. Among teams in the Big Ten West, NU is ranked above every team except Iowa. When garbage time is excluded, NU ranked 50th in efficiency ratings (column GE). NU's strength of schedule ranks 92nd in this measure. NU's offense efficiency is ranked 28th, while the defense efficiency is ranked 80th.

Since I am kind of a stat nerd, I thought these measure were pretty interesting. People can use stats to "prove" whatever theory they have, but I think these efficiency rankings show what we have seen for most of the season. From my point of view, the team has looked good at times, and I have expected more wins than we have seen this year. IMO, NU should have easily beaten Illinois and Purdue. The games against NW, BYU, and Wisconsin were "toss-ups" and it is very disappointing that NU lost all three. The Michigan State win is a very good win. A win over Iowa wouldn't be as huge an upset as the CFB Playoff rankings would make it appear. Finally, Miami appears to have completely fallen off after beating NU in week 3.

 

[NUance]

Interesting site and stats. I agree that stats can be used to prove/disprove most anything. I'm not sure what the FEI rating proves or disproves.

The crux of this season is that we lost five heartbreakers (plus that fiasco loss to the Boilers). We can point to various factors for the losses: injuries, poor clock management, boneheaded plays by players, coaching errors, dumb penalties, etc. For whatever reason we've looked better these past couple of games. But the season is already pretty much in the toilet. So let's just hope all of these things improve next year. And let's kick Iowa's a$$!

 

[BigRedBuster]

Not surprising and I think typical for what can happen during a coaching change.

 

[Sargon]

Here's the stats that explain NU's 5-6

NU efficiency RANK (128 is worst)

First 59 minutes: 28

Last minute and OT: 128

(estimates from Sargon's efficiency ranking system, correlation to observed outcomes .98)

 

[ColoradoHusk]

Not surprising and I think typical for what can happen during a coaching change.

I am not sure what you mean by "what can happen during a coaching change"? Are you saying it's typical for teams to lose a bunch of close games during coaching changes?

 

[NUinID]

You see I am the complete opposite, I look very little at statistics when determining how good the season has gone. The Illinois game was a poor offensive game plan that was poorly executed. The defense played well enough to gain a victory.

The Purdue game was the low point of the season. NU went into the game expecting to win and played poorly. The turnovers are the big story of the game, but defensively they didn't play well at all.

Those are two games that NU should have won because they were clearly the more talented team. Those were truly poor effort gaems. BYU was a fluke win by them. Wisconsin and Miami were toss up games.

The NW game and games like are a big reason why I don't look at stats too closely. If you look at only the statistics in the game and not the final score you would probably determine that NU should have won, but they didn't. Statistics rarely tell the whole story.

 

[Sargon]

Maybe he's saying now that Oregon State finally upgraded their coaching staff they are 2-9 heading to 2-10

At least that's what I thought he was saying

 

[Sargon]

You see I am the complete opposite, I look very little at statistics when determining how good the season has gone. The Illinois game was a poor offensive game plan that was poorly executed. The defense played well enough to gain a victory.

The Purdue game was the low point of the season. NU went into the game expecting to win and played poorly. The turnovers are the big story of the game, but defensively they didn't play well at all.

Those are two games that NU should have won because they were clearly the more talented team. Those were truly poor effort gaems. BYU was a fluke win by them. Wisconsin and Miami were toss up games.

The NW game and games like are a big reason why I don't look at stats too closely. If you look at only the statistics in the game and not the final score you would probably determine that NU should have won, but they didn't. Statistics rarely tell the whole story.

Be interesting to watch the trolls try and troll this.

 

[Enhance]

Statistics are just all part of the puzzle - a measuring stick, in some ways. They don't tell the whole story.

But, often times, they do show an accurate representation. If you're Top 25 in defense and offense, chances are you're a very good team. If you've put up good numbers offensively, but your defense hasn't played well, chances are you have an average record (see Nebraska).

We can talk about last second losses and a handful of plays here or there that could've meant victories, but over the course of our 6 losses, we didn't put together complete enough efforts to win those games.

 

[Pedro Guerrero]

But, often times, they do show an accurate representation. If you're Top 25 in defense and offense, chances are you're a very good team. If you've put up good numbers offensively, but your defense hasn't played well, chances are you have an average record (see Nebraska).

Baylor, Okie State and Florida stink on one side of the ball and are really good on the other yet they have far from an average record. that being said they are far from the norm though.

 

[Enhance]

But, often times, they do show an accurate representation. If you're Top 25 in defense and offense, chances are you're a very good team. If you've put up good numbers offensively, but your defense hasn't played well, chances are you have an average record (see Nebraska).

Baylor, Okie State and Florida stink on one side of the ball and are really good on the other yet they have far from an average record. that being said they are far from the norm though.

Yeah, that's why I said "chances." Those teams, particularly Baylor, tend to skew defensive statistics because they run so many plays.

I guess my overall point though is statistics tend to tell you where a team is at. If you looked at our numbers objectively and didn't know they were Nebraska's, and someone asked you to guess what record that team had, you'd probably say somewhere around 5-8 wins.

 

[ColoradoHusk]

I hear a lot of mention regarding rote statistics, but what the football outsiders guys do is massage the rote statistics to put out efficiency ratings. Baylor and their fast paced offense is going to create a higher YPG produced on offense and allowed on defense. That may not mean their offense or defense is better or worse than others. I like it when "efficiency" measures are done, because the stats are normalized, and made into an apples to apples comparison.

 

[NUinID]

I hear a lot of mention regarding rote statistics, but what the football outsiders guys do is massage the rote statistics to put out efficiency ratings. Baylor and their fast paced offense is going to create a higher YPG produced on offense and allowed on defense. That may not mean their offense or defense is better or worse than others. I like it when "efficiency" measures are done, because the stats are normalized, and made into an apples to apples comparison.

Massage sounds a lot like manipulate.

Anyway I get how you are really into that stat end of the game. Lots of people really like the stuff. In baseball I get the stat end of it because it is a 162 game season. But to me a 12 game season is a little short to draw huge statitical information. Do you play fantacy football, or do you gamble on games? People that do a lot of that tend to look at statistic a lot more than some do. I do neither. Instead of looking at stats I watch a lot of games, because I love the game of football.

 

[ColoradoHusk]

I hear a lot of mention regarding rote statistics, but what the football outsiders guys do is massage the rote statistics to put out efficiency ratings. Baylor and their fast paced offense is going to create a higher YPG produced on offense and allowed on defense. That may not mean their offense or defense is better or worse than others. I like it when "efficiency" measures are done, because the stats are normalized, and made into an apples to apples comparison.

Massage sounds a lot like manipulate.

Anyway I get how you are really into that stat end of the game. Lots of people really like the stuff. In baseball I get the stat end of it because it is a 162 game season. But to me a 12 game season is a little short to draw huge statitical information. Do you play fantacy football, or do you gamble on games? People that do a lot of that tend to look at statistic a lot more than some do. I do neither. Instead of looking at stats I watch a lot of games, because I love the game of football.

Yes, a 12 game season is a small sample size, but each game has 60-100 offensive snaps from each team. The number of snaps creates a large enough sample size to analyze the data efficiently.

I do play fantasy football, but am not super into it. I don't do Draft Kings or Fan Duel. I used to gamble in college, but not anymore. I am into the stat trends because I like math. I also watch games on Saturday. I like having statistical information as another tool to analyze information.

One of the complaints of the CFB Playoff committee is that they are too old, and prefer old-school football, so teams in the Big 12 get penalized. The non-statistical minds may think that certain teams play better defense just because they give up a fewer points per game. In fact, defenses may be "stronger" in a certain conference just because the pace of play is slower. For example, a team like Oklahoma is ranked 25th in scoring defense by PPG, but football outsiders rank their defensive efficiency #1. On the other hand, Wisconsin is #1 in scoring defense in PPG allowed, but #14 in defensive efficiency. Yes, Wisconsin's D is good, but their PPG benefits from style of play from their offense, and the poor offenses that Wisconsin plays in the Big Ten.

 

[NUinID]

I hear a lot of mention regarding rote statistics, but what the football outsiders guys do is massage the rote statistics to put out efficiency ratings. Baylor and their fast paced offense is going to create a higher YPG produced on offense and allowed on defense. That may not mean their offense or defense is better or worse than others. I like it when "efficiency" measures are done, because the stats are normalized, and made into an apples to apples comparison.

Massage sounds a lot like manipulate.

Anyway I get how you are really into that stat end of the game. Lots of people really like the stuff. In baseball I get the stat end of it because it is a 162 game season. But to me a 12 game season is a little short to draw huge statitical information. Do you play fantacy football, or do you gamble on games? People that do a lot of that tend to look at statistic a lot more than some do. I do neither. Instead of looking at stats I watch a lot of games, because I love the game of football.

Yes, a 12 game season is a small sample size, but each game has 60-100 offensive snaps from each team. The number of snaps creates a large enough sample size to analyze the data efficiently.

I do play fantasy football, but am not super into it. I don't do Draft Kings or Fan Duel. I used to gamble in college, but not anymore. I am into the stat trends because I like math. I also watch games on Saturday. I like having statistical information as another tool to analyze information.

One of the complaints of the CFB Playoff committee is that they are too old, and prefer old-school football, so teams in the Big 12 get penalized. The non-statistical minds may think that certain teams play better defense just because they give up a fewer points per game. In fact, defenses may be "stronger" in a certain conference just because the pace of play is slower. For example, a team like Oklahoma is ranked 25th in scoring defense by PPG, but football outsiders rank their defensive efficiency #1. On the other hand, Wisconsin is #1 in scoring defense in PPG allowed, but #14 in defensive efficiency. Yes, Wisconsin's D is good, but their PPG benefits from style of play from their offense, and the poor offenses that Wisconsin plays in the Big Ten.

Ok, I understand the statistical end of the game, but in the end it is just one small thing to look at in a game. Statistiaclly speaking there is no reason why OSU should not have been able to run or pass against an MSU defense that Nebraska almost put 500 yard on. Yet they had less than 150. Statistically speaking Purdue should not have had the kind of game they did against Nebraska, yet they did.

Football is an emotional game that has players playing better " in the moment" than they should be able to statistically. It happens all the time. That is why I guess I don't like too many statistics in football, it takes the emotion out of the game. For someone like you that is exactly why you like the statistic part. Just different opinions I guess.