This might be why Bills fan's are concerned despite the Bills 5-2 record. There will be math.

[GoBills808]

70% success rate picking winners is amazing tbh

 

the data i have says NFL favorites win outright 66.7% rate and the models i follow don't approach 68%

[ChronicAndKnuckles]

31 minutes ago, dave mcbride said:

More than solid - that healthy Ravens team was a top-five team in the NFL. They were decimated by injuries in subsequent games to the point where it became tragicomic.

We have been so lucky with Allen *knock on wood, sign of the cross* health wise. Compare that to the Ravens and Bengals or virtually any team with significant QB injuries. Durability always needs to be taken into account when summing up the value of a player. Would you rather have Joe Burrow or Jared Goff? At this point, I am picking Goff bc I know he’s going to be there suited up on Sunday. Edit: just an example, I am making a point.

Edited yesterday at 05:24 PM by ChronicAndKnuckles

[billsfan89]

The Bills just have to worry about beating KC. Get a harder conference game off your schedule and into the win column. After that just take it one week at a time. There's nothing to honestly worry about longer term because you can only project so far out. The Bills just need to beat KC do that and then you can see how it all shakes out. 

[harryS]

40 minutes ago, GunnerBill said:

There might be math, but the grammar is terrible.

Plus, since 0 Bills fans knew about this metric before opening the thread, it can't be the reason why we're concerned.

[transient]

1 hour ago, thenorthremembers said:

Just kind of a moving forward thing because I see it over and over again, Our GMs last name is BEANE with an E at the end.

 

If you're going to roast him at the very least spell his name right.  

Roasted beans...

[milfandcookies]

All this math makes my head hurt 

[thenorthremembers]

1 hour ago, Chaos said:

who are you talking too?

Applicable to anyone who thinks our GM is named Brandon Bean.

[Dukestreetking]

Many thanks for this. I'm a little bit familiar w the model; very interesting.

 

But why is KC listed as 6-2?

[Simon]

2 minutes ago, Dukestreetking said:

Many thanks for this. I'm a little bit familiar w the model; very interesting.

 

But why is KC listed as 6-2?

 

Something, something, refs...

[D. L. Hot-Flamethrower]

The Giants are juggernaut who knew!!!!

[WNYFAN1]

45 minutes ago, ChronicAndKnuckles said:

We have been so lucky with Allen *knock on wood, sign of the cross* health wise. Compare that to the Ravens and Bengals or virtually any team with significant QB injuries. Durability always needs to be taken into account when summing up the value of a player. Would you rather have Joe Burrow or Jared Goff? At this point, I am picking Goff bc I know he’s going to be there suited up on Sunday. Edit: just an example, I am making a point.

Insanely lucky

[ganesh]

1 hour ago, ProcessTruster said:

Bills will be fine. 

Roster starting to get healthy at the right time. 

Most of the pieces are back and the rest are on the way. 

Losing Oliver is not that big a deal.  

Ed Oliver can take over. a game at a critical time.  His loss will be felt by this team

[The Wiz]

[Captain_Quint]

54 minutes ago, billsfan89 said:

The Bills just have to worry about beating KC. Get a harder conference game off your schedule and into the win column. After that just take it one week at a time. There's nothing to honestly worry about longer term because you can only project so far out. The Bills just need to beat KC do that and then you can see how it all shakes out. 

It's possible that we catch them sleeping the week before their bye, like Atlanta did to us. 

 

And I agree, if we take care of business this week, the outlook improves drastically. 

[hondo in seattle]

8 hours ago, Chaos said:

Its not likely many people are familar with graph network analysis or page rank.  But these are standard statisitical systems that are often applied to sports as ranking or predictive model.   The base formula is standard and I have not altered it in any way.  The results below are what are generated  by the formulas.  I have validate the specific math between the Eagle, Chiers, Bills and Ravens, because they do not seem intuitive. But based on this statistical model, they are accurate.  Here is the math:

The table is the results of this statistic applied to the 2025 NFL season so far.  As a side note. The Chargers at -460 are a compelling, if boring value vs. the Titans this weekend.  And the Bronco's seem tremendously underpriced as an outright underdog in a game they should be a significant favorite at  with the Texans.  Past is not prologue.  But this statistic has a pretty good track record. With tweaks (which I have not done) this can be made to identify the winners accurately (outright winners, not against the spread) about 70% of the time. 
 

essentially the issue is we do not have a single win against a team that is not a bottom dweller, and lost to two mediocre teams. 

 

 

Thanks for posting but...

 

When I was in high school, I created a formula using point and yardage differentials to predict games (outright winners, not against the spread) with an accuracy a little north of 70%.  About the same as this more complicated model.  Knowledgeable fans not using any statistical model can do about the same.   

 

Another good system for ranking teams is ELO, which has the Bills ranked #5.  As already mentioned, DVOA has us as 9th.  

 

I just think the sample sizes are too small for these statistical models to work very well.  A chess Grand Master's ELO rank is determined by hundreds of chess matches.  The Bills current ELO (or DVOA, Page Rank, or whatever) score is determined by 7 idiosyncratic games.   

 

Additionally, injuries, adjustments in scheme, bad bounces, inexplicable ref calls, and all that stuff will create weekly variability in football.  In chess, for example, you don't lose your rook to injury or get your knight back from suspension.  And yet, even in chess, sometimes the player with the lower ELO score wins.  

 

Edited 20 hours ago by hondo in seattle

[billsfan89]

Just now, Captain_Quint said:

It's possible that we catch them sleeping the week before their bye, like Atlanta did to us. 

 

And I agree, if we take care of business this week, the outlook improves drastically. 

 

It is always a big mental boost for this team to beat KC in the regular season. The team really took off in 2024 after they beat KC (only losing 1 more game the rest of the season other than the second meaningless Pats game) and the comeback from 6-6 to win out in 2023 started with a win against KC as well. If the Bills can get the win against KC I think it will make a big impact on the team mentally. KC of course hasn't had their bye and played on Monday night (thus a 1.5 day rest advantage and KC has to travel). Hoping that this team can get it into gear against KC because that's usually a good sign.

[Chaos]

35 minutes ago, Dukestreetking said:

Many thanks for this. I'm a little bit familiar w the model; very interesting.

 

But why is KC listed as 6-2?

That is most likely a mistake. 

[BritBill]

Lads, they’ve done it on Excel. What more evidence do you need? 

1 hour ago, GunnerBill said:

There might be math, but the grammar is terrible.

Math? You’re dead to me. 

[Mark80]

2 hours ago, Chaos said:

Its not likely many people are familar with graph network analysis or page rank.  But these are standard statisitical systems that are often applied to sports as ranking or predictive model.   The base formula is standard and I have not altered it in any way.  The results below are what are generated  by the formulas.  I have validate the specific math between the Eagle, Chiers, Bills and Ravens, because they do not seem intuitive. But based on this statistical model, they are accurate.  Here is the math:

The table is the results of this statistic applied to the 2025 NFL season so far.  As a side note. The Chargers at -460 are a compelling, if boring value vs. the Titans this weekend.  And the Bronco's seem tremendously underpriced as an outright underdog in a game they should be a significant favorite at  with the Texans.  Past is not prologue.  But this statistic has a pretty good track record. With tweaks (which I have not done) this can be made to identify the winners accurately (outright winners, not against the spread) about 70% of the time. 
 

essentially the issue is we do not have a single win against a team that is not a bottom dweller, and lost to two mediocre teams. 

 

Bwhahahahahaha, the Giants ranked 5th in the league in anything completely ruins any credibility to this.  Picking outright winners 70% of the time is not that impressive.  And any true statistician knows that a sample size of 7 or 8 is in no way statistically significant.

[finn]

1 hour ago, thenorthremembers said:

Just kind of a moving forward thing because I see it over and over again, Our GMs last name is BEANE with an E at the end.

 

If you're going to roast him at the very least spell his name right.  

Me, I hate comma splices. If you're doing to complain about spelling, at least get your punctuation right. 😇