2021 CHORUS MOCK DRAFT V4: The "Last" One [Archived] • Page 19

Campbell is such a Bengals corner.

 

i agree — that write up makes it sound like I like him less than I do

 

what lmfao

im already starting to hate RAS. Is fucking everybody an outlier? What is this data set based on and why is it so uncommon for a player’s numbers to just be par for the course. How the hell is 6’3” considered that incredibly short for a TE when there are several great 6’3” TEs as @PeacefulOrca pointed out?

like, I already posted and then deleted a rant about this cause it’s not that serious but Surtain being a 9.96 is stupid. He didn’t score above 9.73 in any category and his athleticism numbers were just, above average. That is some 2+2 = 5 shit for those numbers to scale like that, makes no actual algorithmic sense. I mean he has enough athleticism to keep up with fast receivers but he’s not goddamn Champ Bailey / Deion Sanders not even close. Hell, he’s not even Byron Jones athlete-wise.

 

Oh good, more of this

 

It’s like RAS graphs would be an upside down parabola, the opposite of a normal distribution, where actually very few players are within one standard deviation of the mean and most players are edge cases. It makes no sense.

 

bad data analytics are bad

 

things you don't understand or agree with are not automatically bad

 

A scale that grades above average athletes as a 9.96, and grades 6’3” lower than a 2 at a position where 6’4” is an ideal height, is stupid.... because it does not use numbers the way they’re supposed to be used - as a consistent frame of reference

 

They have even less of an excuse than PFF because no part of this data collection is subjective. So it really is just bad regression analysis / statistics

 

Imagine if both

21/35, 200 yards, 2 TD 1 INT

and

29/32, 340 yards, 3 TD 0 INT

Was basically the same QB rating.
But then

20/36, 180 yards, 2 TD 2 INT

Was suddenly a horrible QB rating, no in between.


That’s basically what RAS is doing

 

I don't understand his formulas, I'm not going to pretend to, and you shouldn't either. Agree or disagree with the result but this new thing where you're coming after entire systems of people doing an exceptional amount of work is real weird and dumb.

 

It is quite literally and obviously not.

 

Also, I could completely see that scenario being true anyway. Say QB C threw two 50 yard screens for TDs. QB A's INT was a dropped pass that got knocked back up. Not hard to imagine

 

Or maybe, making up formulas that actually work for these sorts of things is really hard and just because someone puts a lot of effort into something does not mean it effectively or accurately scales the data

 

Sure.

 

he’s still 27/30 for 240 yards and 2 TD without them.

even if you factor this in, QB A still wasn’t nearly as efficient.

 

QB C, not B.

 

Oh shit, my bad.

 

All I’m saying is it takes legit mathematicians to figure this stuff out.

For example, they learned that the amplitude of a sound wave can not increase by standard units, because it takes a much higher amplitude to create a slightly louder sound, and this increases exponentially. This is how logarithms were invented, because the idea is that they scale each +1 decibel increase by a weird ratio multiplied by a magnitude of 10. (this ratio is represented as “e” on a scientific calculator) As good as I am at math, I still couldn’t explain to you exactly how logarithms work with a formula, but that is the scale that is widely used for decibels, Hertz, and many other physics / science phenomena that increase exponentially.

the way to factor in a lot of things and create a composite, aggregate statistic or metric to measure all those stats, is extremely difficult. One of those high level math things where there’s only one or two right answers and a million wrong ones. Breaking new ground in statistics is very high level thinking.

 

Until they do and Randall still disagrees haha

 

his overall RAS is 4.55, which is pretty close to average.

And he’s technically 6’2” and 1/2, so all those guys have him beat if they’re an even 6’3”. It is not a surprising statement to me if you told me 17 of 20 draft eligible TEs in the past 10 years are 6’3” or above, and that’s basically all that is saying

And I don’t really care about how close the raw score is to 10, especially in a year where the drill times are going to be artificially fast. I generally do like it more as a comparative metric and a relative metric - all else equal, I would like the larger, faster, quicker prospect. It is a red flag that guys like Paris Ford and Richard LeCounte have RASes around 1, but I would not say a guy with a 9.75 and similar draft pedigree is necessarily better than a guy with an 8.25 as a tiebreaker - I would just say they are both sufficiently athletic to play the position. But I might say that comparing the 9.75 to a guy around 6 with similar draft pedigree and taking into account whether the metrics the player is lacking at matter for the position you’re drafting

 

Exactly lol Talking about numbers being used incorrectly when he's not even using the thing he's criticizing correctly.

 

Then my conclusion is that it values size way too much because it gave Patrick Surtain II soooooo much credit for “OMG A 6’2” CORNER WHO ISNT SLOW OR STIFF” which is good, but it’s not once in a generation. Give him a 9/10, sure, but 9.96? 8th most athletic cornerback of all time? What the fuck. Literally Jaycee Horn outperformed him THIS YEAR. Meanwhile being 6’3” is enough to knock a very athletic Brevin Jordan down to 4.55

Idk man. I can’t draw reliable conclusions from RAS numbers because I really disagree with the way it’s weighted.

 

Again, you're using it wrong and then criticizing it for it lol. Why are you looking all time? It's literally supposed to compare to the class.

 

he lists “CB #8 out of 1728” at the top of his RAS profile