PDA

View Full Version : Football Positional Value in the Draft


Rain Man
10-26-2009, 12:00 AM
Just for grins, I looked at the last 12 years' 1st round picks and analyzed them by position.

Then I examined how many positions on the field each position occupied, out of 24 positions (including kickers and punters). For example, guards and OLBs each have two starting positions of the 24, whereas QBs and FSs each have one starting position.

Then I divided the proportion of people drafted at each position divided by their proportion of starting spots. For example, quarterbacks represent 8.9 percent of 1st-round picks but would only represent 4.2 percent of picks if all positions were valued equally (i.e., QBs represented 1/24th of all first-round picks).

By dividing these proportions, we can see the draft value of each position on the field. In the index below, a 100 means that the position is valued exactly equally to its propotion of positions on the field. A value of 200 means it's double the value (i.e., produces twice as many first-round picks as expected), and a value of 50 means it's half as valuable (i.e., produces half as many first-round picks as expected).

I then did exactly the same thing for top-ten picks only.

Here we go. Discuss away. They're presented in descending order of the average of the two indices.


Position 1st Round Top 10
QB 214% 380%
HB 246% 260%
CB 159% 200%
WR 159% 190%
DE 152% 160%
OT 133% 150%
DT 137% 133%
OLB 78% 60%
TE 97% 40%
FS 52% 80%
ILB 70% 33%
SS 52% 40%
C 39% 0%
G 26% 0%
K 6% 0%
P 0% 0%
FB 0% 0%

jAZ
10-26-2009, 02:33 AM
Suprised that HB is overvalued and OLB is undervalued. Maybe I overvalue DT and LT and undervalue LT and LJ.

ChiefGator
10-26-2009, 04:50 AM
Interesting.. thanks for all the midnight calculating...

patteeu
10-26-2009, 04:58 AM
Suprised that HB is overvalued and OLB is undervalued. Maybe I overvalue DT and LT and undervalue LT and LJ.

Yeah, it seems like an incredible waste to use high draft picks on RB when that's a position that has one of the shortest careers in the NFL. A team that can plug in late rounder after late rounder and have RB success, like Denver, seems to have a significant leg up on building the rest of it's team over a team that lives off of RBs in the 1st round.

patteeu
10-26-2009, 05:01 AM
My draft philosophy would be to always take one of the following positions with my 1st round pick each year ...

QB
OT
DE
DT

unless I was set at those positions and thought I was on the cusp of Super Bowl contention

Hog Farmer
10-26-2009, 05:14 AM
Very interesting. Good job rain. It would also be interesting to see how it correlates with successful teams vrs teams like us.

cdcox
10-26-2009, 06:47 AM
Believe it or not, this is a repost. My analysis was based on 3 years and 3 rounds of data and used the draft value chart. I've reposted it below, along with the link to your thread that I posted it in.

http://www.chiefsplanet.com/BB/showthread.php?t=167047&highlight=positional

An interesting question, even though I don't agree with a standard scale by position because it doesn't account for how well a player plays that position.

Nevertheless, I have calculated how the NFL values these positions based on draft order (see details below):


HB: 15.5%
FB: 0.3%
QB: 14.2%
WR: 4.8%
WR: 4.8%
TE: 5.4%
LT: 3.3%
LG: 1.1%
C: 1.7%
RG: 1.1%
RT: 3.3%
K: 0.4%
P: 0.1%
KR/PR (one guy): 0%*
FS: 6.9%
SS: 4.6%
LCB: 4.2%
RCB: 4.2%
SOLB: 3.7%
WOLB: 3.7%
MLB: 5.3%
LDE: 3.5%
RDE: 3.5%
DT: 2.2%
NT/DT: 2.2%

The first surprise is RB>QB this is a little misleading. RB burn up faster than QB so you have to draft them more often. The above numbers should be adjusted for average career length of a starter at the position.

The second surprise is Safties > CB. That just shocks me, but probably should not when you consider the number of teams using a cover 2 and how the rules have castrated CB play.

Thrid surprise is MLB/ILB > OLB. Sure the MLB is the most important spot in the 4-3, but a lot of your play makers and pass rushers come from the OLB spot.

Fourth surprise is TE > WR. I think TE are being over-valued in the NFL right now due to the TG and Gates effect.

Methology in next post.

cdcox
10-26-2009, 06:48 AM
Methodology (only for the geeks):

I compiled a list of the players drafted in the first three rounds over the last 3 years. I chose the first 3 rounds because teams are looking for eventual starters in those rounds. Latter rounds are for backups and special team players. I used the last 3 years to get enough data to even out draft-class variations, while being recent enough to capture current trends.

I then assigned a value to each player based on the NFL Draft value chart. This goes from 3000 points in the first round down to 116 points for the end of the 3rd round. I then totaled the number of draft value points for each position over the 3 year period. I then made some additional adjutments based on the fact that you field 1 QB but 2 guards, etc. I also made the following groupings:

MLB = MLB + ILB
DT = DT + NT

There was no way to separate Left/Right Tackle, Guard, DT, DE, OLB, or CB.

The numbers are a little skewed since the above positional alignments are based on standard proset offense and 4-3 defense. Many times an offense lines up with 2 RB or 2 TE or 3 WR instead of a FB. Also there are 3-4 defenses and nickle packages. These would slightly change the value assignments.

cdcox
10-26-2009, 06:54 AM
Here's my best attempt to normalize my analysis to Rain Man's format:
<code>
QB 244.08%
HB 266.43%
CB 144.39%
WR 165.02%
DE 120.32%
OT 113.45%
DT 75.63%
OLB 127.20%
TE 92.82%
FS 118.60%
ILB 91.10%
SS 79.07%
C 29.22%
G 18.91%
K 6.88%
P 1.72%
FB 5.16%
</code>

Mr. Laz
10-26-2009, 09:41 AM
Battle of the Nerds!!!


Rainman Vs cdcox

2 nerds enter/1 nerd leaves


PBJ PBJ

Mile High Mania
10-26-2009, 09:51 AM
<object width="425" height="344"><param name="movie" value="http://www.youtube.com/v/ef6AKriyTx4&hl=en&fs=1&"></param><param name="allowFullScreen" value="true"></param><param name="allowscriptaccess" value="always"></param><embed src="http://www.youtube.com/v/ef6AKriyTx4&hl=en&fs=1&" type="application/x-shockwave-flash" allowscriptaccess="always" allowfullscreen="true" width="425" height="344"></embed></object>


Embedding for this one was disabled... http://www.youtube.com/watch?v=cWk6RgQbPVc

BigChiefFan
10-26-2009, 09:56 AM
This is exactly the point some are trying to make when it comes to drafting a safety in the top 5. Not that anybody will change their mind, though.

Chiefnj2
10-26-2009, 10:17 AM
Positional value is a weak argument. It is skewed by weak teams. It's okay for KC to take a WR with a top 10 pick because Detroit did so (year after year) and because Miami took Ginn.

Dave Lane
10-26-2009, 10:46 AM
Positional value is critical in the top ten picks and particularly in the top 5 picks. You can't take a safety in the top 4-5 picks unless he can use flubber shoes. Seriously there are good safeties available in the 2nd and third rounds. I have no problem taking even 2 of them in the next draft but a top 5 safety would have to be a world beater.

Reaper16
10-26-2009, 11:02 AM
Positional value is critical in the top ten picks and particularly in the top 5 picks. You can't take a safety in the top 4-5 picks unless he can use flubber shoes. Seriously there are good safeties available in the 2nd and third rounds. I have no problem taking even 2 of them in the next draft but a top 5 safety would have to be a world beater.
Berry and Mays are world beaters.

Fat Elvis
10-26-2009, 11:14 AM
Battle of the Nerds!!!


Rainman Vs cdcox

2 nerds enter/1 nerd conceived


PBJ PBJ

FYP

KCUnited
10-26-2009, 11:18 AM
Since we've already got our franchise QB, its time for that running back.

Rain Man
10-26-2009, 02:45 PM
Here's my best attempt to normalize my analysis to Rain Man's format:
<code>
QB 244.08%
HB 266.43%
CB 144.39%
WR 165.02%
DE 120.32%
OT 113.45%
DT 75.63%
OLB 127.20%
TE 92.82%
FS 118.60%
ILB 91.10%
SS 79.07%
C 29.22%
G 18.91%
K 6.88%
P 1.72%
FB 5.16%
</code>


Interesting stuff. So if I understand right, you're using draft values and the top three rounds, whereas I didn't include draft position and just went for the first round.

Overall, we're relatively consistent even with those differences. Unfortunately, it seems that our different methodologies may muddy a comparison. You have higher scores for linebackers, for example, but that may mean that they get taken a lot more often in Rounds 2 and 3, or that they're more commonly taken in the top part of the first round. I wonder if the former is more the case, since my numbers show that they're underrepresented in the top part of the first round compared to the bottom part.

With defensive tackles, your scores are a lot lower than mine and I don't show differences at all in whether they're top-ten picks, so I guess that means that if they don't go in the first round they go much later (i.e., not in the second or third)? That might point out a reality that defensive tackles tend to be viewd in the draft as either monsters or pedestrian.

I like your theory that halfbacks are ranked high because they have to be replaced more often. That makes for a pretty elegant argument that one should roll the dice on them more often in lower rounds than in higher rounds, and also explains their unnatural commonness in the first round. I'd never considered that before. I wonder if there's a psychological thing going there, too, where coaches look at a guy with a zillion yards in college and have an easier time seeing value.

First-round running backs may also be disproportionately popular because they can help a coach turn a weak team around faster, too. Running backs are a little more plug-and-play than some other position, so it might be a phenomenon where new coaches on weak teams have a self-interest in drafting them.

Rain Man
10-26-2009, 02:47 PM
I should point out too that I'm not posting this as an argument against drafting a safety with a top-ten pick. There are teams who deem it worthwhile to do so. However, I think it points out that the safety had better be darn good - he should be as dominant in his position as the top 20% of quarterbacks who are taken in the first round to warrant the pick.

'Hamas' Jenkins
10-26-2009, 02:57 PM
I should point out too that I'm not posting this as an argument against drafting a safety with a top-ten pick. There are teams who deem it worthwhile to do so. However, I think it points out that the safety had better be darn good - he should be as dominant in his position as the top 20% of quarterbacks who are taken in the first round to warrant the pick.

Which is precisely the argument that the Berry/Mays crowd is making. These guys aren't just Ken Hamlin, they're Ed Reed or Troy Polamalu.

ChiefsCountry
10-26-2009, 03:01 PM
Positional value is a weak argument. It is skewed by weak teams.

You say that, but how come the really good teams who draft when they are in the top 5 stay constiant with the value? Baltimore had top 5 picks they took Ogden - LT, Boulware - Pass Rusher, and Jamal Lewis, a RB. Giants took Rivers aka Manning, Colts took Manning and Edge.

Rain Man
10-26-2009, 03:09 PM
Which is precisely the argument that the Berry/Mays crowd is making. These guys aren't just Ken Hamlin, they're Ed Reed or Troy Polamalu.


I know nothing about the Mayberry crowd, but yeah, I think that's the key. They've got to be ultra-top-notch. The nice thing is that top-ten safeties historically seem to be pretty productive for the most part, if not exactly hall of famers - the six I mentioned in another thread over the past ten years are Antrel Rolle, Sean Taylor, Roy Williams, Laron Landry, Donte Whitner, and Michael Huff.

Rolle may have been drafted as a CB, so I'm not sure whether he counts or not, but he's good.

Taylor was good, and the shortened career is obviously not draft-related.

Roy Williams was good, though his star flamed out pretty quickly.

Landry seems like he's good based on reputation, though I don't see enough Redskins games to know much about him. (Was he playing against us?)

Whitner was a controversial top-ten pick and most thought he went too high, so even though he's been a bit of a disappointment it was more the Bills' fault than his.

Huff is a bust, of course, but he's a Raider so what do you expect.

Chiefnj2
10-26-2009, 03:18 PM
You say that, but how come the really good teams who draft when they are in the top 5 stay constiant with the value? Baltimore had top 5 picks they took Ogden - LT, Boulware - Pass Rusher, and Jamal Lewis, a RB. Giants took Rivers aka Manning, Colts took Manning and Edge.

You believe in taking a HB with a top 5 pick?

'Hamas' Jenkins
10-26-2009, 03:25 PM
I know nothing about the Mayberry crowd, but yeah, I think that's the key. They've got to be ultra-top-notch. The nice thing is that top-ten safeties historically seem to be pretty productive for the most part, if not exactly hall of famers - the six I mentioned in another thread over the past ten years are Antrel Rolle, Sean Taylor, Roy Williams, Laron Landry, Donte Whitner, and Michael Huff.

Rolle may have been drafted as a CB, so I'm not sure whether he counts or not, but he's good.

Taylor was good, and the shortened career is obviously not draft-related.

Roy Williams was good, though his star flamed out pretty quickly.

Landry seems like he's good based on reputation, though I don't see enough Redskins games to know much about him. (Was he playing against us?)

Whitner was a controversial top-ten pick and most thought he went too high, so even though he's been a bit of a disappointment it was more the Bills' fault than his.

Huff is a bust, of course, but he's a Raider so what do you expect.

Landry is one of the best safeties in the NFL. He makes that entire defense in Washington much better, because he can patrol the middle of the field by himself. His play allows Chris Horton to freelance near the line and help out in run support.

He's one of those guys whose names you don't hear often, but like a good ref or CB, it's when you don't notice him that he's doing his job the best.

ChiefsCountry
10-26-2009, 03:42 PM
You believe in taking a HB with a top 5 pick?

If they are special, yes, like Adrian Peterson or LT. Also the NFL has changed as well early part of this decade was still more run than pass. Its alot harder today to justify taking a RB unless they are truly great.

chiefzilla1501
10-26-2009, 03:55 PM
These are great. My only counter-point is that this positional value is largely driven by 4-3 defenses. In a 3-4, I'd shift the value far down for an OLB and DE, and shift it up for NT and Safety.

In a 3-4, I think a Safety is far more important to the defense than a shutdown corner. The defense blitzes like crazy, so if you have a safety that blitzes well but can also play on an island, you have a difference maker.

cdcox
10-26-2009, 04:05 PM
Rain Man -- one thing the draft value chart does is put an extremely high weight on early picks. For example, the #7 pick is worth half as much as the #1. Also, I only looked at 3 years of data: 2005, 2006, and 2007. That might highlight recent trends, or at least trends that were active during that period. For example with more teams running the 3-4, there is less demand for DT. Early in your sample period, CB were valued much more strongly than CB, but in recent years, more value was place on S due to defenses like the cover-2 and with better athletes appearing at that position

cdcox
10-26-2009, 04:10 PM
These are great. My only counter-point is that this positional value is largely driven by 4-3 defenses. In a 3-4, I'd shift the value far down for an OLB and DE, and shift it up for NT and Safety.

In a 3-4, I think a Safety is far more important to the defense than a shutdown corner. The defense blitzes like crazy, so if you have a safety that blitzes well but can also play on an island, you have a difference maker.

The premium on NT will be largely offset by the fact that the 4-3 uses 2 DT compared to 1 for the 3-4. Same with ILB except for the reverse: a monster at MLB is essential for a good 4-3 but ILBs aren't that important for a 3-4, but there are two of them.

I expected to see an enrichment in the TE position in my analysis (concentrating on recent years) compared to Rain Man's (spanning a longer time period), since there are so many TE now that were drafted in the first round. However, TE are valued relatively evenly in the two systems..

Buehler445
10-26-2009, 04:20 PM
Interesting stuff fellas, thanks for the good analysis.
Posted via Mobile Device