Schefter on Roquan Smith

[dll2000]

I

23 minutes ago, vegas492 said:

You very well could be right.

But as far as PFF goes...they apply their metrics the same across the board.  Pretty clearly they have Smith as your third best ILB...and the margin isn't close.

Odds are the Bears staff saw some of the same things.

Look, we all know that Roquan is physically a better player.  There's a lot more upside there.  But he just wasn't getting the job done, or at least he wasn't doing his job as well as others were.

People are talking about mental health, car wreck...etc.  When the answer is probably something simple.  But, we shall see.

Either way, I hope the man is okay.  And if he isn't, I hope he gets some help.  He's too young and way too talented to let something derail his career.

PFF does their stuff fast and takes short cuts resulting in a lot of bad data.

I haven't looked at Smith, but I can tell you I have looked at other players and they can be way off up or down.

To properly judge a player it takes time watching a play - at least one time at speed and then at very least a few times in slow motion to get an idea of how someone did on each play.  

You first have to identify the play and what responsibility of said player within that play was.  Then you have to focus on the player to see how they did in that context.  Then you have to add context of what was happening with everyone else which takes at least another slow motion viewing - usually two or three.  

There is no conceivable way PFF does above given how fast they pump out info and how many players they do it on.  They have a system that allows them to do it fast, but short cuts make for bad data.

From listening to them it seems they someone or perhaps someones watch each play one time and give a player a numerical grade immediately ranging from -2 to +2.  0 is you did what was "expected."  2 is really good - 2 is really bad.   

I think, I know based on some of their grades, they are missing a bunch of context in exchange for efficiency and speed.   Human bias also factors in a lot more than they would admit as well.

 

 

 

 

[dll2000]

Just now, dll2000 said:

I

PFF does their stuff fast and takes short cuts resulting in a lot of bad data.

I haven't looked at Smith, but I can tell you I have looked at other players and they can be way off up or down.

To properly judge a player it takes time watching a play - at least one time at speed and then at very least a few times in slow motion to get an idea of how someone did on each play.  

You first have to identify the play and what responsibility of said player within that play was.  Then you have to focus on the player to see how they did in that context.  Then you have to add context of what was happening with everyone else which takes at least another slow motion viewing - usually two or three.  

There is no conceivable way PFF does above given how fast they pump out info and how many players they do it on.  They have a system that allows them to do it fast, but short cuts make for bad data.

From listening to them it seems they someone or perhaps someones watch each play one time and give a player a numerical grade immediately ranging from -2 to +2.  0 is you did what was "expected."  2 is really good - 2 is really bad.   

I think, I know based on some of their grades, they are missing a bunch of context in exchange for efficiency and speed.   Human bias also factors in a lot more than they would admit as well.

 

 

 

 

In short PFF puts out real time grades.  They have grades ready to go as soon as a game is over.  That is simply impossible unless we have very advanced programs that can do it.  Humans cannot do that.  

In my view that tells me their data cannot be as good as it needs to be and therefore cannot be trusted.  

 

[GBHalas]

Just going on the metrics maybe this is what exactly the Bears are saying it is...something is going on in his personal life that has his mind off of football.  Maybe they sat him down and he told them what was going on and they gave him some time in order to get his young life in order.  You have to think this is the first time he has been on his own, and maybe like everyone else in society, life got a little hard and he needed a time to straighten it out.  

[dll2000]

19 minutes ago, GBHalas said:

Just going on the metrics maybe this is what exactly the Bears are saying it is...something is going on in his personal life that has his mind off of football.  Maybe they sat him down and he told them what was going on and they gave him some time in order to get his young life in order.  You have to think this is the first time he has been on his own, and maybe like everyone else in society, life got a little hard and he needed a time to straighten it out.  

I am 90% sure it is one of two things.  I refuse to say publicly though.

 

[malagabears]

https://chicago.suntimes.com/bears/2019/10/2/20895335/bears-roquan-smith-full-go-practice-raiders-personal-issue-vikings

Suntimes is now reporting he will be a full practice today. That hopefully means he will be ready to go either this week or next.

[vegas492]

2 hours ago, dll2000 said:

I

PFF does their stuff fast and takes short cuts resulting in a lot of bad data.

I haven't looked at Smith, but I can tell you I have looked at other players and they can be way off up or down.

To properly judge a player it takes time watching a play - at least one time at speed and then at very least a few times in slow motion to get an idea of how someone did on each play.  

You first have to identify the play and what responsibility of said player within that play was.  Then you have to focus on the player to see how they did in that context.  Then you have to add context of what was happening with everyone else which takes at least another slow motion viewing - usually two or three.  

There is no conceivable way PFF does above given how fast they pump out info and how many players they do it on.  They have a system that allows them to do it fast, but short cuts make for bad data.

From listening to them it seems they someone or perhaps someones watch each play one time and give a player a numerical grade immediately ranging from -2 to +2.  0 is you did what was "expected."  2 is really good - 2 is really bad.   

I think, I know based on some of their grades, they are missing a bunch of context in exchange for efficiency and speed.   Human bias also factors in a lot more than they would admit as well.

 

 

 

 

500+ employees.  Many of them are former players who know what they are looking at and looking for.

Collinsworth himself is an owner and looking for the best analytics that his team can produce.

They have the manpower to do it their way and do it efficiently.

Not really defending them here, as I really don't care.  But, their metric is applied the same way to the same players.  Rant against the metric all you want, but at least it is consistent.

I haven't watched Roquan enough to know if his grade was deserved.

I signed up for PFF this year and find that when I look at my Packers, I agree with their scoring.  It matches my "eyeball" test.

BTW...Ha Ha and Jackson are kicking GB's safeties pretty good.

HHCD is #5 score of 85.2

Jackson is #21.  Score of 69.5.

Amos is #23.  69.1

Savage is #26.  68.2.

Can't believe you guys found a way to make HHCD not stink.  Terrible for us.  Worse for the 'Skins.  Gotta think that pairing him with Jackson really helped.  And there's always the Mack effect.

[dll2000]

45 minutes ago, malagabears said:

https://chicago.suntimes.com/bears/2019/10/2/20895335/bears-roquan-smith-full-go-practice-raiders-personal-issue-vikings

Suntimes is now reporting he will be a full practice today. That hopefully means he will be ready to go either this week or next.

If he plays people will forget about it in time or leave it alone.

If he keeps sitting out entirely different story.

 

[dll2000]

1 minute ago, vegas492 said:

500+ employees.  Many of them are former players who know what they are looking at and looking for.

Collinsworth himself is an owner and looking for the best analytics that his team can produce.

They have the manpower to do it their way and do it efficiently.

Not really defending them here, as I really don't care.  But, their metric is applied the same way to the same players.  Rant against the metric all you want, but at least it is consistent.

I haven't watched Roquan enough to know if his grade was deserved.

I signed up for PFF this year and find that when I look at my Packers, I agree with their scoring.  It matches my "eyeball" test.

BTW...Ha Ha and Jackson are kicking GB's safeties pretty good.

HHCD is #5 score of 85.2

Jackson is #21.  Score of 69.5.

Amos is #23.  69.1

Savage is #26.  68.2.

Can't believe you guys found a way to make HHCD not stink.  Terrible for us.  Worse for the 'Skins.  Gotta think that pairing him with Jackson really helped.  And there's always the Mack effect.

All I am saying is there are no short cuts to quality film review.  Even Belichick needs to look at a play at least twice.

 

[vegas492]

1 minute ago, dll2000 said:

All I am saying is there are no short cuts to quality film review.  Even Belichick needs to look at a play at least twice.

 

90 NFL and Division One Teams subscribe to their analytics.

PFF employs guys who know the positions they are watching.  And they employ a lot of them. 

I could link you to articles that explain this, but I doubt you care.  

I find it interesting, but not interesting enough to get into the weeds.

When I see that 90 NFL teams/Division One Collegiate teams use their metrics, that tells me it is a valid system and has some merit.

[dll2000]

3 minutes ago, vegas492 said:

500+ employees.  Many of them are former players who know what they are looking at and looking for.

Collinsworth himself is an owner and looking for the best analytics that his team can produce.

They have the manpower to do it their way and do it efficiently.

Not really defending them here, as I really don't care.  But, their metric is applied the same way to the same players.  Rant against the metric all you want, but at least it is consistent.

I haven't watched Roquan enough to know if his grade was deserved.

I signed up for PFF this year and find that when I look at my Packers, I agree with their scoring.  It matches my "eyeball" test.

BTW...Ha Ha and Jackson are kicking GB's safeties pretty good.

HHCD is #5 score of 85.2

Jackson is #21.  Score of 69.5.

Amos is #23.  69.1

Savage is #26.  68.2.

Can't believe you guys found a way to make HHCD not stink.  Terrible for us.  Worse for the 'Skins.  Gotta think that pairing him with Jackson really helped.  And there's always the Mack effect.

Believe me that if Roquan or anyone was stinking up the joint this board would not be shy about saying so.

They may make excuses for a player, but no one is going to say he is playing good when he isn't.

 

 

[vegas492]

3 minutes ago, dll2000 said:

Believe me that if Roquan or anyone was stinking up the joint this board would not be shy about saying so.

They may make excuses for a player, but no one is going to say he is playing good when he isn't.

 

 

It's not that his play wasn't "good".  It just wasn't as good as the other guys that were in there.  

And in that sample size, it could have been as small as 10 plays where Roquan just wasn't there.

In your defense, the way it flies around at all levels, that would have been very easy for the fans eye to miss, as the play was still being made.

If you didn't have so many "Eff up the Play" guys on defense, it would stand out more.  BTW...I liked Roberson-Ryan last year.  What in God's name has gotten into him this year?  Kid has All Pro written all over him.

[dll2000]

1 minute ago, vegas492 said:

90 NFL and Division One Teams subscribe to their analytics.

PFF employs guys who know the positions they are watching.  And they employ a lot of them. 

I could link you to articles that explain this, but I doubt you care.  

I find it interesting, but not interesting enough to get into the weeds.

When I see that 90 NFL teams/Division One Collegiate teams use their metrics, that tells me it is a valid system and has some merit.

It doesn't cost a lot (to them) to have a nice statistical database to use as a starting point and second check on your own assessments and scouting.  Of course they would subscribe.  

It can be a valuable tool.  Not saying all their stuff is garbage.  I listen to their podcast.  

I am just saying their assessments and reviews have a human opinion element and isn't repeatable science or a mathematical certainty.   I also think they are not doing it as carefully and slowly as they should for better quality.

 

 

[dll2000]

I can give you an example.  Oops gotta go.

Work.

[vegas492]

9 minutes ago, dll2000 said:

I can give you an example.  Oops gotta go.

Work.

No need.  You won't change my mind about it.

I'm going to like any metric that shows Bradbury as the 32nd ranked center in the league!

[WindyCity]

1 hour ago, vegas492 said:

90 NFL and Division One Teams subscribe to their analytics.

PFF employs guys who know the positions they are watching.  And they employ a lot of them. 

I could link you to articles that explain this, but I doubt you care.  

I find it interesting, but not interesting enough to get into the weeds.

When I see that 90 NFL teams/Division One Collegiate teams use their metrics, that tells me it is a valid system and has some merit.

It is the advance stats that they subscribe for, not the grades.

The advanced stats are very nice to have. But the grades are super subjective and can never be that accurate without knowing the play call and the responsibility.