FiveThirtyEight: This Bills Team Worse Than Most Bills Teams this Century

[MJS]

Shows how bad the AFC was this year.

[Casey D]

30 minutes ago, Rob's House said:

The numbers are skewed by 2 bad games, one of which we took a chance and started a rookie instead of our starting QB.

 

Between the Saints and Chargers we were -67 points. 

 

I'd add that the 2 Patriots games were a lot closer than the final scores would indicate. Of course, the Jets game was not nearly as close as the score indicated, so ...

 

In any case, the Jauron teams were good at consistently losing close games, and Fitz would always bring us back only to consistently turn the ball over at the end.

 

If you look at this team overall throughout the season, they're clearly better than all but 2 or 3 Bills teams this century.

To elaborate a bit more, the Bills were absolutely dreadful for that three game stretch with the Jets, NO and Chargers.  -80 in three games.  Before those three games 5-2, after 4-2 and overall + 23, with both second half losses to NE.  I don't know why they were so bad for those three games, but it skews all the statistical analysis.  Run a +23 over 13 games and I'll bet 9-4 is not so outlandish.  The two losses to NE were -41.  So 9-2 with a +64 in 11 games. 

 

Basically for all but those three horrid games, the Bills played consistently well, and won close games.  Their season was similar to KC, who was 5-0, 1-6, and 4-0. Why, who knows.  Unless the team from those three horrid games shows up, they will give Jax a good game.

Edited January 3, 2018 by Casey D

[Mat68]

55 minutes ago, Buffalo716 said:

 

The national media is calling us a Joke. From Bill Simmons to Colin cowherd they are giving us no chance

 

there are WNY transplants all over the country who love the Bills like shooter mcgavin(yes I know it's not his real name lol)

 

but it's definitely the minority... nobody in the national media thinks we are gonna win

Colin Cowherd is picking Buffalo to beat Jacksonville because his theory is the better qb wins in the playoffs.

[Hapless Bills Fan]

1 hour ago, 26CornerBlitz said:

This year’s Bills team finished 9-7, but based on Football Outsiders’ estimated wins and Pythagorean wins metrics,¹ it only earned 6.8 and 6.3 wins, respectively. Put another way, an NFL team that gets outscored by 57 points on the year (as the Bills did) mathematically ought to win about six or seven games.

 

But it’s not just about wins. Pick a metric of team strength, and the Bills are typically at or below the median: 21st in scoring differential, 15th in FiveThirtyEight’s Elo ratings, 23rd in Simple Rating System,² 21st in Defense-adjusted Value Over Average.³

 

These Bills don’t stack up well against the 17 iterations that came before them, either:

 

OK so how does Football Outsiders estimated wins and Pythagorean win metrics stack up against actual results?

[Buffalo716]

9 minutes ago, Mat68 said:

Colin Cowherd is picking Buffalo to beat Jacksonville because his theory is the better qb wins in the playoffs.

 

He picked buffalo on the record?

 

he never has anything good to say about the Bills 

[syhuang]

1 hour ago, 26CornerBlitz said:

This year’s Bills team finished 9-7, but based on Football Outsiders’ estimated wins and Pythagorean wins metrics,¹ it only earned 6.8 and 6.3 wins, respectively. Put another way, an NFL team that gets outscored by 57 points on the year (as the Bills did) mathematically ought to win about six or seven games.

 

But it’s not just about wins. Pick a metric of team strength, and the Bills are typically at or below the median: 21st in scoring differential, 15th in FiveThirtyEight’s Elo ratings, 23rd in Simple Rating System,² 21st in Defense-adjusted Value Over Average.³

 

These Bills don’t stack up well against the 17 iterations that came before them, either:

 

 

First, if people are not familiar with "estimated wins" and "Pythagorean wins", they may be misled into thinking the above bold statement means using two different formulas resulting in similar results that Bills should only have 6~7 wins. In fact, "estimated wins" is just an adjustment from "Pythagorean wins". In other words, "estimated wins" would usually be fairly close to "Pythagorean wins" after Football outsiders adjust "Pythagorean wins" by DVOA(s), or the so called "forest index". The basic idea of "estimated wins" is to take strength of opponents into account.

 

Second, with "estimated wins" based on "Pythagorean wins", the next question is what "Pythagorean wins" is? It's a formula created by Bill James for baseball advanced statistics. Only two factors are used, points for and points against. The following is the formula (after adjusted from Bill James' original formula):

 

 

 

How good this formula is is up to debate. However, one obvious flaw in the formula is the sample size. Using points for and points against from 162 games and 16 games is totally different. With larger sample size like 162 games, few good/bad games and injuries have much smaller impact on points for and points against. The same can't be said to a 16-game sample size.

 

Pretty much all Bills fans know this team won several close games this season and lost two games by large margins (10-47 against Saints and 24-54 against Chargers). Points for and points against are greatly affected by these games. Again, how good Pythagorean win formula is is debatable, but obviously it wouldn't work in football as well as baseball simply due to sample size.

 

[Billsfan1972]

1 hour ago, C.Biscuit97 said:

We won a 4 way tie at 9-7 beating mostly horrible teams.  If we weren't Bills fans, this was a pretty boring, nothing special team.

 

But the Bills finally caught a break and they get to play a meaningful game for the first time in 18 years.  And anything can happen.  But this was not a good football team. But they don't beat themselves and in a league loaded with crap teams, that can be enough. 

Bang on....  And I was forever stating 9-7 gets the Bills in the playoffs vs. a very beatable Jacksonville......  BTW 8-7-1 would not have.....

Edited January 3, 2018 by Billsfan1972

[Chilly]

There were 256 losses and 256 wins in the NFL this year (no ties). 

 

Of the losses, here's how they break down:

 

1-9 points - 50.4%

10-19 points - 27.3%

20-29 points - 18%

30-39 points - 3.5%

40+ points - 0.8%

 

Bills losses looked like this:

 

1-9 Points (2 losses, 29% of losses):

Panthers (6 Points)

Bengals (4 Points)

 

10-19 Points (1 loss, 14% of losses):

Jets (13 Points)

 

20-29 Points (2 losses, 29% of losses):

Patriots (20 Points)

Patriots (21 Points)

 

30-39 Points (3 losses, 43% of losses):

Sants (37 Points)

Chargers (30 Points)

 

When the Bills were bad, they were really bad, and way worse than the NFL on average.  I'm sure those losses that got away from them are causing the metrics to look really bad.  Especially because DVOA takes into account game situations, and the later in the game it is, I'm sure it penalizes you for it.

 

I do think this likely means that the Bills generally have times when they get into matchup issues and it kills them.  But I like the matchup vs the Jags.  Call me crazy, but I think it's going to be a very defensive game, and because of that, I'll take the QB that doesn't act like a moron.

 

 

 

[Punching Bag]

2 hours ago, 26CornerBlitz said:

This year’s Bills team finished 9-7, but based on Football Outsiders’ estimated wins and Pythagorean wins metrics,¹ it only earned 6.8 and 6.3 wins, respectively. Put another way, an NFL team that gets outscored by 57 points on the year (as the Bills did) mathematically ought to win about six or seven games.

 

As a Honors Math Major I'd say I'd like to see the proof for their faulty math so I could correct the formula.

In math if they results do not consistently match results then there is something wrong with formula.

[RyanC883]

1 hour ago, Fadingpain said:

Funny, the computer says we should have won the 6 or 7 games most predicted, including me at 6 games.

 

 A perfect storm of factors came together this year to get us in that has not happened in recent years, despite having an arguably better team in those years.

 

I get that people don't care and we are just happy to have the monkey off our back...

 

But we still have a way's to go.  For a first year effort from a new GM and coaching staff, you'd have to say things are going well.

 

 

 

exactly.  Things are going well, much improvement still needed (QB, pass rush top the list).  But, a perfect storm of factors has kept us out a bunch of times, so I'll take Dalton to Boyd getting us in this year! Most Playoff teams (esp. wildcard) need some "luck."   And we also beat playoff teams this year.  

[Jasovon]

After years of moral victories and winning offseasons or leading the league in metrics, i'm happy just to make it to the dance.

[Rico]

All the Bills teams from 2000 to 2016 have one thing in common: they were all failures.

 

This is just more fake news, sad!

[WhoTom]

That's why they play the game on the field instead of in a computer.

 

 

 

7 minutes ago, Limeaid said:

 

As a Honors Math Major I'd say I'd like to see the proof for their faulty math so I could correct the formula.

In math if they results do not consistently match results then there is something wrong with formula.

 

This is just one data point, which could simply be a statistical outlier. That doesn't mean their formula is incorrect.

 

 

[Gray Beard]

It’s amazing how many people diss the idea of the “eyeball” impression versus statistics. Until the statistics don’t match the “eyeball“ results. 

[QCity]

2 hours ago, C.Biscuit97 said:

We won a 4 way tie at 9-7 beating mostly horrible teams.  If we weren't Bills fans, this was a pretty boring, nothing special team.

 

But the Bills finally caught a break and they get to play a meaningful game for the first time in 18 years.  And anything can happen.  But this was not a good football team. But they don't beat themselves and in a league loaded with crap teams, that can be enough. 

 

An accurate assessment. 

[yungmack]

2 hours ago, klos63 said:

there are no atheists in foxholes

Many foxes are atheists.

[H2o]

3 hours ago, 26CornerBlitz said:

This year’s Bills team finished 9-7, but based on Football Outsiders’ estimated wins and Pythagorean wins metrics,¹ it only earned 6.8 and 6.3 wins, respectively. Put another way, an NFL team that gets outscored by 57 points on the year (as the Bills did) mathematically ought to win about six or seven games.

 

But it’s not just about wins. Pick a metric of team strength, and the Bills are typically at or below the median: 21st in scoring differential, 15th in FiveThirtyEight’s Elo ratings, 23rd in Simple Rating System,² 21st in Defense-adjusted Value Over Average.³

 

These Bills don’t stack up well against the 17 iterations that came before them, either:

The numbers look this way due to the Saints, Chargers, Jets, and Pats games we laid an egg in. Douche 38 is just scraping for something to write about.

[26CornerBlitz]

ESPN Insider Required:

 

Edited January 3, 2018 by 26CornerBlitz

[BuffaloBill]

The most probable reason for this team doing well is because of the turnover differential.

[TheBrownBear]

Honestly, we are like a crappy version of the Pats.  Just plod along, minimize mistakes and allow the other team to eventually slip up and give you the game.  Our talent, on the whole, is pretty pitiful.  McDermott deserves tons of credit for what he was able to get out of this team.