2020 NFL Draft Discussion

[C'monchargers]

ANALYTICS TOP 300 RANKINGS (2020 NFL DRAFT)


BY HAYDEN WINKS
March 8, 2020, 6:53 pm ET
Updated On: March 9, 2020, 10:34 am ET
FacebookTwitterRedditEmailPrint Before we get to the rankings, I’ve decided to make the “10 Draft Commandments” after doing all of this research:

1) Understand that all of our prospect rankings suck. Even multi-million dollar GMs are bad.

2) Seriously. Everyone sucks at this. Draft Twitter, anonymous scouts, Belichick. All quite bad!

3) With that out of the way… Try to trade down, unless you’re drafting a quarterback.

4) Adjust for positional value. QB > other positions > RB.

5) Athleticism matters more on defense than it does on offense, especially at EDGE.

6) Younger players are typically better. Adjust college production by age.

7) College production matters more at some positions (QB, RB, WR, EDGE, LB).

8) Counting and market share stats are more important than efficiency stats.

9) TFLs and solo tackles are more predictive than sacks for DT, EDGE, and LB prospects.

10) Players on good college teams are usually better. Small school prospects aren’t discounted.
Analytics Top 300 Rankings (2020 NFL Draft)


If you’re curious about the step-by-step process of this project, I outlined it under the Top 300 Big Board on Page 4, but essentially I’m projecting NFL production using college stats, athleticism, and film. That means this is NOT an analytics-only big board. Instead, it’s a big board that combines film grades and analytics, adjusting for positional value of course. That’s why you’ll see QB Jordan Love ranked highly despite not being an analytics quarterback. If I was only using analytics, Love wouldn’t be a top-five quarterback prospect, but ignoring film (or analytics) is a bad process, as the kids say.



Page 1: QB, RB, WR, TE

Page 2: OT, IOL, DT, EDGE

Page 3: LB, CB, S

Page 4: Top 300 Big Board and Methodology

Quarterbacks


How to read this table: “Justin Herbert, an 82nd percentile quarterback prospect, is the QB4 and No. 5 overall player in the 2020 NFL Draft.”

Overall	Prospect	Pos	Percentile	Notes
1	Joe Burrow	QB1	0.98	Ranked 4th-of-179 in my QB database
2	Tua Tagovailoa	QB2	0.93	2nd in Total QBR in 2018 and 2019
4	Jalen Hurts	QB3	0.91	Efficient 21yo passer with rushing ability
5	Justin Herbert	QB4	0.82	Traits paired with above-avg production
13	Jordan Love	QB5	0.69	Awful numbers, will still be drafted high
24	Anthony Gordon	QB6	0.62	95th percentile in Total EPA, small arm
26	Jake Fromm	QB7	0.61	Below-average stats/traits but smart
32	Jacob Eason	QB8	0.57	Big arm but 1 year of average production
174	Nathan Stanley	QB9	0.26
186	Steven Montez	QB10	0.25
201	Brian Lewerke	QB11	0.23
206	Shea Patterson	QB12	0.22
247	Cole McDonald	QB13	0.15
299	James Morgan	QB14	0.08

For more on the quarterback class, read my early reports and statistical evals here.



Running Backs


How to read this table: “Jonathan Taylor, a 96th percentile running back prospect, is the RB1 and No. 58 overall player in the 2020 NFL Draft.”

Overall	Prospect	Pos	Percentile	Notes
58	Jonathan Taylor	RB1	0.96	3 seasons of 1,975+ RuYDs, 4.39 speed
64	J.K. Dobbins	RB2	0.95	21yo, 161 ToYD/G, most 20+ yard rushes
76	Clyde Edwards-Helaire	RB3	0.89	21yo, 94 RuYD/G, 55 catches, 4.60 forty
84	D'Andre Swift	RB4	0.82	21yo, efficient runner, 73 career catches
98	Cam Akers	RB5	0.76	20yo, 14 TDs/30 catches at FSU, 4.47 forty
128	Zack Moss	RB6	0.63
132	AJ Dillon	RB7	0.62
140	Eno Benjamin	RB8	0.56
179	Antonio Gibson	RB9	0.43
202	Ke'Shawn Vaughn	RB10	0.37
203	Lamical Perine	RB11	0.37
212	Darrynton Evans	RB12	0.34
213	Joshua Kelley	RB13	0.33
230	Anthony McFarland	RB14	0.28
257	Michael Warren	RB15	0.21
287	Scottie Phillips	RB16	0.15
289	LeVante Bellamy	RB17	0.14
292	J.J. Taylor	RB18	0.11
293	DeeJay Dallas	RB19	0.11
294	Salvon Ahmed	RB20	0.11
297	Ben LeMay	RB21	0.10

For more on the running back class, read my early reports and statistical evals here.








The Process

1. Gather data (athleticism, college production, and NFL production) for all drafted players from the 2005 draft class and on.
2. Build two different multiple-linear regression models based on the first step.
   1. Adjusted SPARQ = position-adjusted athleticism
   2. Production Score = age and team-adjusted college production
3. Build the final multiple-linear regression model using Adjusted SPARQ, Production Score, and draft capital that predicts Approximate Value per Game over a prospect’s first four NFL seasons. This means the model is accounting for film as well.
4. Adjust the final model’s projections to further account for positional value. In terms of importance, I have the positions tiered as:
   1. QB
   2. CB, EDGE, OT, WR
   3. DT, LB, S, IOL, TE
   4. RB

Although the way he starts out the article sounded like he is putting out mere info and he doesn't really believe much in it. I believe that is because he is not considering each position has its own importance in different strength and traits. Personally it has a lot to deal with how closely you play to the line of scrimmage.

Anyhow, One name among QB that jumps out to me about his system is Nate Stanley. Would like to hear more on what he thinks.

Also, I agree with the listing for the most part, JK Dobbins is probably my top RB this year, he is up there. Just my personal liking. Then my second favorite value on the list is Cam Akers. His draft position is definitely well worth his talent. Most impressive thing about him has to be his patience in waiting for the open lane. Le'veon-ish.

[chargeroo]

You’ve been pimping him for awhile, sir. Carry on with your bad self and if we draft him I’ll cheer for him along with you. Gotta give these kids a chance, the league needs them lol.

You know, I will say this about Herbert: when he plays, he plays effing fierce. Whatever laid back nature he has leaves the building and he looks very fired up out there. Good thing, IMO.

I'm not advocating we draft him. I'm just saying I think he's better than some do.
I hope they don't use the #6 pick on a QB because there are several position players that be available that would do us a lot more good than a rookie QB will do us. We can get the QB in round 2.

[Topcat]

I'm not advocating we draft him. I'm just saying I think he's better than some do.
I hope they don't use the #6 pick on a QB because there are several position players that be available that would do us a lot more good than a rookie QB will do us. We can get the QB in round 2.

Sounds like Hurts...and if the Bolts see him as an upgrade from T-Mobile and Stick, I'd be willing to pull the trigger...

[chargeroo]

Sounds like Hurts...and if the Bolts see him as an upgrade from T-Mobile and Stick, I'd be willing to pull the trigger...

Hurts does look like the kind of QB our HC would like. I won't be shocked if we pick him. He sure can run an RPO offense and he can hurt teams when he escapes the rush. If he can learn to think pass a bit longer before taking off he could be good.

Who knows? History tells us that knowing ahead of time who will be a great QB is nearly impossible.

[wu-dai clan]

OK coupon shoppers,
how much of a better deal
is it to get Hurts @ #37
than to get James Morgan @ #186 ?

That is the question which needs to be answered.

[chargeroo]

OK coupon shoppers,
how much of a better deal
is it to get Hurts @ #37
than to get James Morgan @ #186 ?

That is the question which needs to be answered.

James Morgan? - Are you looking for third string QB?

[wu-dai clan]

James Morgan? - Are you looking for third string QB?

I will repeat.

In your opinion is Hurts at #37
a better deal,
all things considered,
than Morgan at #186 ?

Part of my inquiry is whether or not
we should roll the dice that Morgan
surprises the hell out of us.

[chargeroo]

I'm the wrong guy to ask because I'm not a fan of Hurts even though I think AL probably is.

As for Morgan, he was pretty decent in the Shrine game. I saw him play one other game during the season and he didn't impress me.

[wu-dai clan]

I'm the wrong guy to ask because I'm not a fan of Hurts even though I think AL probably is.

As for Morgan, he was pretty decent in the Shrine game. I saw him play one other game during the season and he didn't impress me.

I have done a lot of homework.
Thank you for your response.
Morgan has leadership qualities, a quuck release, and some mobility.
The mechanics are sloppy but correctable.
Hurts edge is level of competition.
Otherwise, the description is interchangeable.

[like54ninjas]

I will repeat.

In your opinion is Hurts at #37
a better deal,
all things considered,
than Morgan at #186 ?

Part of my inquiry is whether or not
we should roll the dice that Morgan
surprises the hell out of us.

Steven Montez, Bryce Perkins, Tyler Huntley all fit as day 3 QBs with mobility and upside.
Montez has a very strong arm, leadership attributes, athleticism, speed to execute RPOs but lacks consistency and processing/decision making.

[wu-dai clan]

Steven Montez, Bryce Perkins, Tyler Huntley all fit as day 3 QBs with mobility and upside.
Montez has a very strong arm, leadership attributes, athleticism, speed to execute RPOs but lacks consistency and processing/decision making.

Perkins is a hit for me, ninja.

[sonorajim]

Options if we stay at #6, including those unlikely to remain:
QB : Burrow, Tua
Edge: Young
DB/LB : Okudah, Simmons
OL: Becton, Thomas, Wills, Wirfs
DT: Brown

I'd take my highest remaining guy and be happy. This is a great year to pick at 6 & 37.