Player Rankings + RAS

[deathstar]

I was listening to the Pack A Day podcast yesterday and the creator of RAS (https://relativeathleticscores.com/) was on. He spoke about how the Packers are one of the most predictive teams in the NFL with regards to selecting athletes. We spoke about it in one of the threads as well: in the last two drafts we've selected two players ranked below 7.7 RAS. So with that in mind I created a spreadsheet this morning that combines rankings of players from different sources (currently only TDN and PFF pre-combine) and their RAS. I averaged the rankings of the big boards then averaged that ranking with their RAS to give players a score. The lower the score the higher they rank on our board. I used TDN's top 100 as a template and threw out anyone with an RAS below 8 to save time. I also have not listed anyone who has not worked out.

UPDATE: I've included all players and merely highlighted ones that I believe are off our board. I've included 3 big boards, CBS, drafttek, and nflbigboard.com. If anyone has more big boards that are in tables please send me the links so I can include them in the average.

Current Issues: improve legibility, would like to have position ranks as well as overall ranks.

[Bulbouswhore]

Here’s a score worth noting

[JaireAlex]

Great list. Very much the players I was thinking we would draft. 

 

Now just drop character concerns like Tillery and I can see this as being close to the GB board.

 

Wonder  if safety Harris (BC) makes the marks. I have about seven safeties I like for GB in this draft so far.

Thanks for the work.

[SaginawReunion83]

15 minutes ago, JaireAlex said:

Great list. Very much the players I was thinking we would draft. 

 

Now just drop character concerns like Tillery and I can see this as being close to the GB board.

 

Wonder  if safety Harris (BC) makes the marks. I have about seven safeties I like for GB in this draft so far.

Thanks for the work.

 

I don't have the good looking full picture as listed above...but here is his ranking. Makes him the highest rated by the RAS scale for SS in this class.

 

SS

Will Harris

Boston College

2019COMBINE

9.62

[AlexGreen#20]

This is going to sound like a stupid question but, why do we care about RAS?

The Packers don't use it internally as far as we know. Is there a published formula.

[SaginawReunion83]

4 minutes ago, AlexGreen#20 said:

This is going to sound like a stupid question but, why do we care about RAS?

The Packers don't use it internally as far as we know. Is there a published formula.

It should be ignored. All it shows to me is who tested out the best at their combines. For example a guy like Brown from oklahoma will have a much lower score due to scoring so low in height and weight. It's just showing who is the ideal athlete with the perfect size.

[Packerraymond]

7 minutes ago, AlexGreen#20 said:

This is going to sound like a stupid question but, why do we care about RAS?

The Packers don't use it internally as far as we know. Is there a published formula.

Why should we care how good an athlete a football player is? Why wouldn't we?

We don't know anything the Packers do internally, but we do know that since about 2016 we've moved into a draft philosophy that's relied heavily on top end athletic numbers. I don't need to know how exactly it's calculated it's pretty obvious that a baseline has been calculated for each position (probably a rolling average of all known data in that field for the position) that scores you a 5 and from there each standard deviation out from that baseline you are moves you up and down that scale from 0-10. 

[Packerraymond]

8 minutes ago, SaginawReunion83 said:

It should be ignored. All it shows to me is who tested out the best at their combines. For example a guy like Brown from oklahoma will have a much lower score due to scoring so low in height and weight. It's just showing who is the ideal athlete with the perfect size.

Why would you ever want to ignore any metric or data provided to you? 

[BrettFavre004]

Jonah Williams was disappointing, around like 5.5 as a tackle.

I asked Kent was Wiliams would test as a guard, that is where I like him better anyways.

 

And as for the Packers, they are one of the most consistent RAS drafters of all the teams.  I am not implying they are using this or a similar metric, but there is enough data points to show by now we definitely giving higher grades to players who have elite athletic traits. Things like good jumps often times you can see on the field from a DT or a pulling guard with a good first step, etc. etc.

[deathstar]

10 minutes ago, Packerraymond said:

Why should we care how good an athlete a football player is? Why wouldn't we?

We don't know anything the Packers do internally, but we do know that since about 2016 we've moved into a draft philosophy that's relied heavily on top end athletic numbers. I don't need to know how exactly it's calculated it's pretty obvious that a baseline has been calculated for each position (probably a rolling average of all known data in that field for the position) that scores you a 5 and from there each standard deviation out from that baseline you are moves you up and down that scale from 0-10. 

That's pretty much it. It compares everyone to everyone else, with data going back to the 80s. 

I doubt Green Bay uses RAS, but there is correlation among high RAS scores and the Packers' selections.

[deathstar]

1 hour ago, Bulbouswhore said:

Here’s a score worth noting

He could definitely be a target later in the draft. I'm pretty sure no one expected him to test that well.

[AlexGreen#20]

20 minutes ago, Packerraymond said:

Why should we care how good an athlete a football player is? Why wouldn't we?

We don't know anything the Packers do internally, but we do know that since about 2016 we've moved into a draft philosophy that's relied heavily on top end athletic numbers. I don't need to know how exactly it's calculated it's pretty obvious that a baseline has been calculated for each position (probably a rolling average of all known data in that field for the position) that scores you a 5 and from there each standard deviation out from that baseline you are moves you up and down that scale from 0-10. 

I get that, but how is the overall number calced?

Surely there's a better way of running this than doing standard deviations?

[Packerraymond]

2 minutes ago, AlexGreen#20 said:

I get that, but how is the overall number calced?

Surely there's a better way of running this than doing standard deviations?

Do good athlete's score highly in the formula? Yes. Do bad athletes score poorly? Yes.

Why change what's not broken?

[deathstar]

5 minutes ago, AlexGreen#20 said:

I get that, but how is the overall number calced?

Surely there's a better way of running this than doing standard deviations?

https://www.prideofdetroit.com/2016/5/16/11678686/relative-athletic-scores-what-they-are-and-why-they-work

This goes into more specifics. It takes every 40 run by every player at that position and scores it 1-10. If you score a 9.8 you're in the 98% of 40s run by that position.

[Joe]

27 minutes ago, BrettFavre004 said:

Jonah Williams was disappointing, around like 5.5 as a tackle.

I soured on him at the combine. I think his testing numbers make him tumble down the board. Possibly 20-25 range rather than the 5-15 range he had been previously pegged. It may not show up on the Joe Schmoe sites, but it HAS to be on teams' boards; especially considering the knock on 'Bama players is that they're topped off when entering the league.