Todays Athletic article - Reality Check on Bills

[DCOrange]

3 hours ago, CincyBillsFan said:

According to the article Allen is one of the worst in the NFL at this.  Yet by your description:

 

*  The Miami game INT did not lead to any points.  Sure it put them in scoring position but does outcome play a role here?  And how bad is a pick 6 viewed in this model? They lead directly to 7 points and I've seen a lot of those this season but none from Allen.

 

*  The Houston INT also did not lead to points

 

*  The TN INT had the worst outcome with the Titans scoring a TD off it. BTW, this was the only Allen INT through the first 9 games that led directly to a TD. 

 

*  The INT's against Jacksonville were the next worst as they likley took 3 points off the board and gave the Jaguars 3 points.

 

*  The INT against the Jets was not relevant as the score was 38 -3 at the time.  Allen clearly got greedy trying to feed Diggs.

 

You know what this looks like to me? Normal QB play in the NFL with all it's ups and downs.  And if the above put Allen near the bottom of the pack with this stat then it's a meaningless stat because Allen's "incidents" don't look like much considering they cover NINE games.

 

Given the enormous amount of uncontrolled variation on every single NFL play the attempt to mathematically quantify performance is doomed.  It's a foolish exercise and IMO results in deeply misleading takes on what happened in a game.

 

 

 

 

 

 

It doesn’t matter if they actually lead to points for this metric. It matters if they lead to expected points (i.e. how likely a team is to score as a result of the INT). A pick 6 on first down at the goal line would be the worst possible play in this system since it’s essentially a guaranteed 14 point swing. 
 

So again, pretty much every Allen INT either took points off the board for us or handed points to the other team (the fact that opponents routinely shot themselves in the foot after the fact doesn’t matter). That is not normal QB play as evidenced by the fact that he’s nearly the worst in the league in this regard. You can argue it’s just bad luck and we’re talking about 10 or so INTs so it’s probably not a meaningful sample size to extrapolate from, but nobody is extrapolating anyways. They’re just saying he’s been nearly the worst in the league in terms of EPA subtracted from his INTs and saying if that improves, it’ll help the team. 

Edited November 24, 2021 by DCOrange

[CincyBillsFan]

5 hours ago, DCOrange said:

It doesn’t matter if they actually lead to points for this metric. It matters if they lead to expected points (i.e. how likely a team is to score as a result of the INT). A pick 6 on first down at the goal line would be the worst possible play in this system since it’s essentially a guaranteed 14 point swing. 
 

So again, pretty much every Allen INT either took points off the board for us or handed points to the other team (the fact that opponents routinely shot themselves in the foot after the fact doesn’t matter). That is not normal QB play as evidenced by the fact that he’s nearly the worst in the league in this regard. You can argue it’s just bad luck and we’re talking about 10 or so INTs so it’s probably not a meaningful sample size to extrapolate from, but nobody is extrapolating anyways. They’re just saying he’s been nearly the worst in the league in terms of EPA subtracted from his INTs and saying if that improves, it’ll help the team. 

I don't accept the accuracy or relevance of this measure. 

 

You're right about one thing - the sample size is ridiculously small here as Allen only had 6 INT's through the first 9 games (the period you said was covered by the article).  Only SIX.

 

 

 

[Nextmanup]

On 11/23/2021 at 10:26 AM, dorquemada said:

To the OP's point though, the Bills did get some bounces last year, and we're not getting them this year.  Add to that the one dimensional O, and it's a recipe for losing close games.  If there was only some solution!

The "O" is no different than it was last year. 

 

And it's not the problem--the one dimensionality.

 

The problem is execution/coaching/focus/preparedness on both sides of the ball.