Fun With Numbers: Size Vs Speed

Interested in discussing college prospects? Or even High School Prospects? This is the forum.
Amarok
Practice Squad
Practice Squad
Posts: 168
Joined: Wed Jul 24, 2013 2:38 pm

Re: Fun With Numbers: Size Vs Speed

Postby Amarok » Thu Apr 17, 2014 8:33 am

Patmos wrote:While I have not done anything nearly as extensive, I have researched the influence of metrics in an attempt to build a "player profile" for the top WR. I found that the broad jump and vertical are great indicators, along with size/weight/speed. For example, the only top 10 WR with a vertical less than 36" last year was AJ Green at 34.5." All of the top 10 had a broad jump greater than 10' (120"). With the exception of DeSean Jackson, Dez was the only WR shorter than 6'3" (he's 6'2"). Nobody was slower than a 4.55 forty.

As a result, my player profile is:

1- 6'2"+ / 210+ (ideal is 6'3" / 220+)
2- 4.55 forty or better
3- 10' broad jump / 36" vertical

If I look at a WR in the 6'-6'1" range, he needs to be 4.45 forty or better (but I factor less production due to less touchdowns).

I personally believe that combining the metrics into a single metric diminishes the accuracy, I prefer to see if the player meets all the profile requirements individually. If he does, I like him. I will adjust for vertical or height, but I eliminate anyone who doesn't hit the other metrics.

I would certainly be curious on your findings regarding the vertical and broad jump.

This is very interesting, and I think you're spot on. I think the metric I've come up with may be a better substitute for Points 1 and 2 you mentioned because it takes into account how fast a player is for their size, which to me is much more applicable than just arbitrarily deciding where the line is for speed/height.

From what I see on my metric, anything over 190 is good, and anything over 200 is elite. Using your third factor (broad/vertical jump), I think you get a pretty damn good idea of a player's explosiveness and physical ability.

Thoughts?
14 Team Dynasty League with 21 man rosters
.5 PPR, 4 pts/Passing TD, Scored w/ Fractionals
Start 1 QB/2 RB/3 WR/1TE/2 Flex/1 K/1 Def

QB: M. Stafford; E. Manning; R. Griffin III
RB: L. McCoy; A. Abdullah; A. Ellington; R. Mathews; C. Michael;
WR: C. Johnson; D. Thomas; Antonio Brown; J. Gordon; P. Harvin; D. Adams; D. Moncrief; J. Strong
TE: J. Thomas; L. Green; E. Ebron
K: C. Sturgis
DEF: Jacksonville

Amarok
Practice Squad
Practice Squad
Posts: 168
Joined: Wed Jul 24, 2013 2:38 pm

Re: Fun With Numbers: Size Vs Speed

Postby Amarok » Thu Apr 17, 2014 10:04 am

2014 rookie WRs w/ 10' broad jump, 36" vertical, and speed/size score over 190:

Davante Adams
Albert Wilson
Allen Robinson
Bennie Fowler
Brandin Cooks
Bruce Ellington
Donte Moncrief
Jeff Janis
Martavis Bryant
Mike Campanaro
Odell Beckham Jr.
14 Team Dynasty League with 21 man rosters
.5 PPR, 4 pts/Passing TD, Scored w/ Fractionals
Start 1 QB/2 RB/3 WR/1TE/2 Flex/1 K/1 Def

QB: M. Stafford; E. Manning; R. Griffin III
RB: L. McCoy; A. Abdullah; A. Ellington; R. Mathews; C. Michael;
WR: C. Johnson; D. Thomas; Antonio Brown; J. Gordon; P. Harvin; D. Adams; D. Moncrief; J. Strong
TE: J. Thomas; L. Green; E. Ebron
K: C. Sturgis
DEF: Jacksonville

User avatar
Butt Liqueur
Pro Bowler
Pro Bowler
Posts: 1082
Joined: Thu Jun 13, 2013 9:15 am

Re: Fun With Numbers: Size Vs Speed

Postby Butt Liqueur » Thu Apr 17, 2014 10:25 am

Amarok wrote:2014 rookie WRs w/ 10' broad jump, 36" vertical, and speed/size score over 190:

Davante Adams
Albert Wilson
Allen Robinson
Bennie Fowler
Brandin Cooks
Bruce Ellington
Donte Moncrief
Jeff Janis
Martavis Bryant
Mike Campanaro
Odell Beckham Jr.
Robinson and Ellington are my two WR targets of this draft. Nice to see them on here

16 Team, PPR
(1 QB, 1 RB, 2 WR, 1 TE, 3 FLEX)
QB: J. Cutler, M. Mariota, M. Cassel, C. Henne, A. McCarron, D. Fales
RB: J. Bell, F. Gore, R. Bush, D. Williams, R. Helu, L. Dunbar, J. Allen, K. Williams, C. Artis-Payne
WR: J. Jones (ATL), J. Nelson, A. Hawkins, C. Johnson (MIN), A. Robinson, R. Randle, K. Thompkins, B. Ellington, H. Douglas, D. Harris, M. Campanaro, J. Hardy, J. Saunders, D. Moore, I. Blakeney, T. Lippett
TE: C. Clay, A. Quarless, A. Sefarian-Jenkins, J. Cumberland, W. Saxton

Draft Picks: 2016 2 3rds, 4th, 5th, 6th, 7th

User avatar
Patmos
Role Player
Role Player
Posts: 357
Joined: Wed May 02, 2012 5:24 pm

Re: Fun With Numbers: Size Vs Speed

Postby Patmos » Thu Apr 17, 2014 2:17 pm

Amarok wrote:2014 rookie WRs w/ 10' broad jump, 36" vertical, and speed/size score over 190:

Davante Adams
Albert Wilson
Allen Robinson
Bennie Fowler
Brandin Cooks
Bruce Ellington
Donte Moncrief
Jeff Janis
Martavis Bryant
Mike Campanaro
Odell Beckham Jr.
Did Cody Latimer miss because he did not perform at the combine? He smoked it at his pro day and, if I understand your metrics, he is definitely on the list.
aka Indy Speedway in the Lead Dog Championship League

Amarok
Practice Squad
Practice Squad
Posts: 168
Joined: Wed Jul 24, 2013 2:38 pm

Re: Fun With Numbers: Size Vs Speed

Postby Amarok » Thu Apr 17, 2014 2:24 pm

Patmos wrote:
Amarok wrote:2014 rookie WRs w/ 10' broad jump, 36" vertical, and speed/size score over 190:

Davante Adams
Albert Wilson
Allen Robinson
Bennie Fowler
Brandin Cooks
Bruce Ellington
Donte Moncrief
Jeff Janis
Martavis Bryant
Mike Campanaro
Odell Beckham Jr.
Did Cody Latimer miss because he did not perform at the combine? He smoked it at his pro day and, if I understand your metrics, he is definitely on the list.
Latimer did not have enough data to run the test, unfortunately. I would venture a guess that you're correct though.
14 Team Dynasty League with 21 man rosters
.5 PPR, 4 pts/Passing TD, Scored w/ Fractionals
Start 1 QB/2 RB/3 WR/1TE/2 Flex/1 K/1 Def

QB: M. Stafford; E. Manning; R. Griffin III
RB: L. McCoy; A. Abdullah; A. Ellington; R. Mathews; C. Michael;
WR: C. Johnson; D. Thomas; Antonio Brown; J. Gordon; P. Harvin; D. Adams; D. Moncrief; J. Strong
TE: J. Thomas; L. Green; E. Ebron
K: C. Sturgis
DEF: Jacksonville

BwK
Role Player
Role Player
Posts: 285
Joined: Sun Jul 11, 2010 4:33 pm

Re: Fun With Numbers: Size Vs Speed

Postby BwK » Thu Apr 17, 2014 7:45 pm

Patmos wrote:While I have not done anything nearly as extensive, I have researched the influence of metrics in an attempt to build a "player profile" for the top WR. I found that the broad jump and vertical are great indicators, along with size/weight/speed. For example, the only top 10 WR with a vertical less than 36" last year was AJ Green at 34.5." All of the top 10 had a broad jump greater than 10' (120"). With the exception of DeSean Jackson, Dez was the only WR shorter than 6'3" (he's 6'2"). Nobody was slower than a 4.55 forty.

As a result, my player profile is:

1- 6'2"+ / 210+ (ideal is 6'3" / 220+)
2- 4.55 forty or better
3- 10' broad jump / 36" vertical

If I look at a WR in the 6'-6'1" range, he needs to be 4.45 forty or better (but I factor less production due to less touchdowns).

I personally believe that combining the metrics into a single metric diminishes the accuracy, I prefer to see if the player meets all the profile requirements individually. If he does, I like him. I will adjust for vertical or height, but I eliminate anyone who doesn't hit the other metrics.

I would certainly be curious on your findings regarding the vertical and broad jump.
But you are only confirming the ones who have made it correct? I am sure there are many prospects who reach your criteria that busted. Then ultimately you can never predict success and only confirm?

I should also add that I assume 90% of the top 10 wr are the first wr off the board in most dynasty drafts so if you dont get the top wr in the class you should not be expecting a top 10 wr. Too many owners set their expectations too high. Of course the top wr's in the league perform the best in the drills at the combine. If you arent getting the #1 or #2 wr in the draft you should not be expecting much. If you happen to get lucky I am not sure that it is because of your metric as many will score fairly well and not make it. situation and desire are two criteria the metric doesnt account for. If your criteria is based on talent just trade your picks for the top player at each position and I will promise you hit 90% of the time...no metric needed. Now if you want to get a wr#2 then you need to expand your metric outside the top 10 nfl wr.
10 team non-ppr no FA pickups
QB- Ryan, Romo, Winston
RB-Foster, GIO, Mathews, L. Murray, JStew, Duke
WR-Green, Sammy, Benjamin, B. Coleman, Fitzgerald, Lee, Latimer, Algholar, C. Coleman, Doctson
TE-Graham, Green
K-Bailey, Prater
DEF-Rams

Csl312
Pro Bowler
Pro Bowler
Posts: 1023
Joined: Sun May 06, 2012 4:01 pm

Re: Fun With Numbers: Size Vs Speed

Postby Csl312 » Wed Apr 23, 2014 9:00 pm

So I did a little bit of crunching the numbers here. The first thing I wanted to look at is can we see a difference between the size/speed score (which I will just call "score" for simplicity) for players selected in different rounds.

Here are the results of that:
Round Mean Score Standard Deviation
1 192.8 8.89
2 185.1 10.42
3 183.7 8.70

So on first glance the round 1 guys look to be on average a little better than the round 2 or 3 guys. Round 2 shows the most variability here (I would guess this is where a decent number of speedy but not as tall receivers are taken but that is just a hunch).

I then did a student's t-test to compare the scores as grouped by round selected in the NFL draft
This got me a p-value of 0.012 comparing round 1 v 2, 0.00068 for comparing round 1 v 3, and 0.579 for round 2 v 3. That result indicates that there is a "significant" difference between round 1 and either 2 or 3, but basically no difference between the scores for players selected in the second or third rounds from one another. Take this for what you will, it seems the 1v3 comparison is probably pretty valid for a cheap statistical test. (Note I am by no means a statistician).

So that is interesting, but not super informative. So I wanted to look at how these guys performed on the field and compare to this score. The results were not too great though. I calculated each players average fantasy points per game using their career receiving and rushing stats (in a 10yds/pt, 1PPR, 6pt/TD scoring setup) and their total games played. I did not include return stats, partly because I did not want to record yet another number, but also because I wanted to see how this translates to an offensive player's performance and not include special teams.

Plotting average fantasy points per game vs. the score for each player gave me an idea if these two variables were related. The results were pretty scattered. I tried to put a linear fit (because it is simple and the data looked linear if anything to me) to the resulting spread and here are the results:

For all players on this list
Slope = 0.4933 R-squared value = 0.0466
For only round 1+2 players
Slope = 0.4844 R-squared value = 0.0431
For only round 1 players
Slope = 0.691 R-squared value = 0.1363

Slope here is not super meaningful, but I was hoping it would be much steeper. That doesn't matter too much though because the R-squared values give an indication as to how well that trend line fits the data - if the line passes through each data point the value would be 1. None of these numbers inspire much confidence that the linear fit is representative of any trend in the data. Taking out players who have 16 games or less experience did not improve anything either.

I tried to upload the excel file I was working with, but the board won't allow any spreadsheet formats to be uploaded (don't have any idea what you can attach then). For now I just put a jpeg with the three plots, attached.
Plots from excel of data.
Plots from excel of data.
WR_data_plots.JPG (51.61 KiB) Viewed 325 times
Hope you guys find this helpful, anyone have any ideas where to go from here?


Who is online

Users browsing this forum: No registered users and 34 guests