Jump to content


ZWK

Member Since 14 Jul 2008
Offline Last Active Today, 04:14 AM
*****

Posts I've Made

In Topic: ZWK's 2015 Prospect Analysis

Yesterday, 11:01 PM

Here are links to clips of every pass that Mariota threw 15+ yards downfield against Arizona or Oregon State in 2014 (available thanks to Draft Breakdown), sorted by the length of the throw.

 

36 yards incomplete Arizona 1:55

34 yards incomplete Oregon State 0:22

33 yards incomplete Arizona 0:39

29 yards incomplete Arizona 1:32

29 yards complete Oregon State 3:15

29 yards incomplete Oregon State 7:06

28 yards complete Arizona 9:38

27 yards incomplete Arizona 4:31

26 yards complete Oregon State 10:00

25 yards incomplete Arizona 7:33

23 yards incomplete Oregon State 5:22

21 yards complete Oregon State 9:07

20 yards incomplete Arizona 6:21

20 yards complete Arizona 7:45

19 yards incomplete Arizona 8:52

19 yards incomplete Oregon State 7:39

19 yards complete Oregon State 0:41

18 yards incomplete Arizona 0:58

18 yards incomplete Arizona 1:20

17 yards complete Arizona 7:23

16 yards complete Oregon State 1:29

15 yards complete Oregon State 4:21

15 yards complete Oregon State 8:28


In Topic: ZWK's 2015 Prospect Analysis

Yesterday, 09:31 PM

Cross-posted to the Brett Hundley thread:

 

Here are links to clips of every pass that Hundley threw 15+ yards downfield, in the four 2014 games that are on Draft Breakdown (Arizona, Utah, USC, Virginia), sorted by the length of the throw.

 

42 yards incomplete Arizona 3:21

41 yards incomplete USC 2:21

40 yards complete Arizona 6:04

40 yards incomplete Arizona 6:51

32 yards complete Virginia 0:01

31 yards intercepted Utah 4:27

31 yards incomplete USC 3:36

29 yards complete Utah 7:04

28 yards incomplete Virginia 5:24

25 yards incomplete Utah 6:54

25 yards complete USC 6:17

24 yards complete Utah 5:31

24 yards incomplete USC 4:52

23 yards complete Utah 1:45

19 yards incomplete USC 7:53

18 yards incomplete Virginia 3:33

18 yards complete USC 0:58

17 yards complete Utah 4:50

17 yards incomplete USC 7:33

16 yards complete Virginia 1:43

16 yards incomplete Virginia 7:22

16 yards incomplete Arizona 0:48

15 yards complete Virginia 5:11

15 yards incomplete Utah 0:53

15 yards complete USC 5:34

15 yards incomplete Arizona 0:33

15 yards complete Arizona 1:36

 

There are some nice throws in there (like the 29 yard completion against Utah at 7:04), but on the whole I'm not that impressed with him as a downfield passer.


In Topic: Dynasty: UCLA QB Brett Hundley

Yesterday, 09:30 PM

Here are links to clips of every pass that Hundley threw 15+ yards downfield, in the four 2014 games that are on Draft Breakdown (Arizona, Utah, USC, Virginia), sorted by the length of the throw.

 

42 yards incomplete Arizona 3:21

41 yards incomplete USC 2:21

40 yards complete Arizona 6:04

40 yards incomplete Arizona 6:51

32 yards complete Virginia 0:01

31 yards intercepted Utah 4:27

31 yards incomplete USC 3:36

29 yards complete Utah 7:04

28 yards incomplete Virginia 5:24

25 yards incomplete Utah 6:54

25 yards complete USC 6:17

24 yards complete Utah 5:31

24 yards incomplete USC 4:52

23 yards complete Utah 1:45

19 yards incomplete USC 7:53

18 yards incomplete Virginia 3:33

18 yards complete USC 0:58

17 yards complete Utah 4:50

17 yards incomplete USC 7:33

16 yards complete Virginia 1:43

16 yards incomplete Virginia 7:22

16 yards incomplete Arizona 0:48

15 yards complete Virginia 5:11

15 yards incomplete Utah 0:53

15 yards complete USC 5:34

15 yards incomplete Arizona 0:33

15 yards complete Arizona 1:36

 

There are some nice throws in there (like the 29 yard completion against Utah at 7:04), but on the whole I'm not that impressed with him as a downfield passer.


In Topic: ZWK's 2015 Prospect Analysis

Yesterday, 04:30 PM

Running Back update (previous post on RBs here)

 

ZWK, that is sweet. Are you going to color code your RB's as well?

I've added color coding to the 2014 season / 2015 draft class in my RB stats spreadsheet & RB elusiveness spreadsheet. Green = good, red = bad, whiter = averager, and other than that the precise colors don't mean anything (the scale is actually set automatically by Google sheets, relative to other players that year on that statistic).

 

I haven't updated any of the numbers in the spreadsheets, despite some new information (like pro days from Tevin Coleman & David Cobb), because I'm currently using a borrowed laptop while mine is in the shop so I can't easily re-run my formulas.

 

College Football Focus has put out their top 4 RB prospect ranking, along with a couple articles on their RB stats including their elusiveness rating and RB receiving & pass blocking. They have Gordon & Gurley 1-2, and then Abdullah & Ajayi a tier down at 3-4. I have the same top 4, and the same tiering, but I put Gurley ahead of Gordon and Ajayi ahead of Abdullah (based on my impressions of them - I believe the raw calculations by my formulas would agree exactly with CFF on the order). Gurley does extremely well by their stats - they have him first in elusiveness, by a lot - and I'm inclined to put him a mini-tier ahead of Gordon, neck-and-neck with Cooper for the overall #1 rookie spot. CFF has Gurley behind Gordon in their ranking primarily because of his ACL risk, but I think I'm more worried about the fact that Gordon doesn't have any elite physical attributes according to his combine numbers (size, speed, jumps, etc.). Gurley's positive buzz also seems to be on a different level from Gordon's.

 

CFF's pass blocking numbers match Abdullah's reputation as a poor pass blocker. That, plus his fumbling and lack of size (205 at the combine, 198 at the senior bowl), raise enough questions about his ability to carve out a big role for him to rank behind Ajayi in my eyes.

 

Tevin Coleman's pro day 40 is being reported as a 4.40 by nfldraftscout, which I think is enough to move him up to RB5 for me, just ahead of Duke Johnson. Coleman also does well in CFF's elusiveness stats, but I think that's less impressive than it looks at first glance because 1) his yards after contact per carry was padded by some long runs with early contact, which don't actually demonstrate that much elusiveness and 2) CFF says that a big chunk of his broken tackles came against weak competition. Duke Johnson actually ranks ahead of him in CFF's elusiveness rating against Power 5 opponents.

 

David Cobb's pro day 40 is being reported as a 4.73 by nfldraftscout, which is slow enough to be very concerning (especially because I had similar concerns watching him play). Some other sources are reporting it as being around 4.6, but it's hard to know what to make of those reports. There are also a couple of small pieces of bad news about Cobb from CFF, with him not making their top 10 in elusiveness rating and also ranking in their bottom 10 at pass blocking. I'm inclined to drop him down a tier below Coleman, Duke, and David Johnson, to RB8. 

 

I haven't been that impressed by what I've seen of David Johnson's running, so he's pretty clearly behind Coleman & Duke Johnson, but he still slots in at #7 thanks to his size & speed & receiving.

 

Cameron Artis-Payne & TJ Yeldon don't look that good by my numbers, but they both made CFF's top 5 in elusiveness rating against Power 5 conferences. I currently see them as a cut above the rest of the field, rounding out my top 10 RBs.

 

Javorius Allen is one guy who I've been down on for a long time, and CFF just gave me one more reason to be down on him: apparently he's a bad pass blocker.


In Topic: ZWK's 2015 Prospect Analysis

17 April 2015 - 11:35 PM

Football Outsiders and College Football Focus have both published their WR prospect rankings. They produce two of the four stats-based prospect rankings that I pay attention to (along with wdcrob's and my own), and have caused me to make some updates to the order that I'd draft WR prospects in.

 

Some of the more notable players: They both have Amari Cooper as the clear #1 (as do my numbers) and put Nelson Agholor in the top 5 (my numbers have him 10th). They are both pretty high on Jaelen Strong (8th & 4th), and to a lesser extent on Tyler LockettDevin Smith, and DeAndre Smelter. The most striking difference is on White & Parker. Football Outsiders are major outliers on White & Parker: they rate both of them as below average prospects, in part because their formula penalizes prospects who stayed in college after they were NFL draft eligible (they also don't pro-rate stats at all, so Parker is treated as if he had 43/855/5 in 13 games rather than in 6 games). CFF has them at 2 & 3, with Parker ahead of White (as do I).

 

Taking their numbers into account (as well as other recent buzz), I'm inclined adjust my rankings from a month ago to add a mini tier break among WRs after Cooper, move Agholor up (probably ahead of Lockett), and put Smelter on of my list of sleepers (probably ahead of Austin Hill & John Harris, who didn't get combine invites).

 

On methodology:

 

Football Outsiders' WR ratings, which they call Playmaker Score, come from a formula which is based on a regression of analysis of historical data. For WRs drafted since 1996, they found which variables (out of the ones that they have access to) predicted NFL success, and with what weights. This is a lot more rigorous than my approach, which is a clear advantage - they are systematically testing their guesses about which variables matter using historical data, and throwing out the variables that haven't been predictive. On the other hand, their approach has the disadvantage of forcing them to leave out any variable which we don't have 20 years of data on, so they can't include numbers like yards per target and drop rate (which I do include). There is also a risk of overfitting - when there are several similar variables that they could include, they choose the one which has historically been most predictive, whereas my approach is to include all of them averaged together somehow. For most players, it won't matter much difference (e.g. TDs per team passing attempt and percent of team's passing TDs are highly correlated), but for a few players it can make a big difference. (I think that this is basically what happened with their decision to not pro-rate data: the small sample of players who played a partial season have happened to do badly in the NFL, so the formula is more predictive of historical NFL success if you don't pro-rate.)

 

College Football Focus mostly bases their rankings on 2014 performance. They are most known for what they call "grades", which they create by watching every play for every game of every player, marking each play down as "good" "bad" or "neutral" according to some consistent standard which they don't publicize, and then counting up a player's total number of "good" plays and subtracting the number of "bad" plays. (They also count some plays as doubly or triply good/bad, and recenter their numbers to average 0). Their grades reward consistency and avoid rewarding players too much for easy/fluky plays which can inflate their standard stats, but I believe that they don't correct all that much for differences due to things like strength of schedule or scheme. CFF also track a bunch of other stats - yards per route run and drop rate are great stats to know (although unfortunately they don't share them for every player). Their prospect rankings seem to be based on their (stats-informed) opinions, rather than being calculated directly from their stats, and I don't know how much weight they gave to things like size, speed, and age.

 

The WR numbers that I've put together are still in this spreadsheet, and I've added color-coding (green=good, red=bad, whiter=averager) to make it easier to see at a glance.