ROI on Top 10 NFL Draft Picks Since 2000 (1 Viewer)

[Anarchy99]

I was bored and threw some stuff together to get a snapshot on Top 10 NFL draft picks since 2000. I used Weighted Career AV to determine value. Not the greatest of tools, but at least something to use as a comparison. The numbers in some cases could be a little misleading. Recent picks won't have played enough to have high scores. For example, someone like Ja'Maar Chase will fall in the <25 category for now. So that column could mean there are recent players, or it could reflect a really poor draft pick (hard to tell from a summary of the total results).

BASED ON POSITION:
(Position, # of Players, AVG Weighted Career AV, Players with Scores of 50+, Players with Scores of 100+, Players with Scores of <25)

POS    #    AVG    50    100    <25
QB    40    54.60    17    6    9
RB    16    54.31    8    2    1
LB    23    47.83    9    3    8
OL    36    43.58    16    0    11
DL    47    40.96    14    4    16
DB    30    39.63    12    1    8
WR    33    37.39    10    1    14
TE    05    30.00    1    0    2

BASED ON DRAFT PICK:
(Pick#, AVG Weighted Career AV, Players with Scores of 50+, Players with Scores of 100+, Players with Scores of <25)

#    AVG    50    100    <25
Pick 1    62.83    15    4    2
Pick 2    46.52    8    3    7
Pick 3    44.57    7    1    7
Pick 4    46.13    8    2    7
Pick 5    47.83    10    2    4
Pick 6    41.78    8    1    6
Pick 7    40.22    9    1    8
Pick 8    36.83    9    0    9
Pick 9    42.74    7    2    7
Pick 10    34.91    6    1    11

BASED BY TEAM THAT SELECTED THE PLAYER:
(Team, # of Players, AVG Weighted Career AV, Players with Scores of 50+, Players with Scores of 100+, Players with Scores of <25)

LAC    6    71.67    3    2    0
NEP    2    71.00    2    0    0
IND    2    63.50    2    0    0
BAL    4    63.00    2    1    0
CAR    8    61.88    5    2    3
MIN    6    60.17    4    2    1
DAL    5    60.00    4    0    0
ATL    9    59.67    5    2    3
DEN    3    51.00    1    1    1
PHI    4    49.25    2    0    1
CIN    10    48.70    3    2    3
HOU    7    48.43    3    0    2
KCC    6    48.33    3    0    1
TBB    7    48.14    4    0    1
WAS    9    47.00    3    0    1
ARI    11    46.55    6    1    4
PIT    2    46.50    1    0    0
BUF    8    46.25    3    0    1
SFO    10    45.10    5    0    2
CHI    7    43.71    2    1    2
MIA    7    40.86    3    0    3
NYG    8    40.25    1    1    3
LAR    7    39.57    2    0    2
CLE    13    38.69    3    0    4
DET    14    38.29    3    2    6
TEN    7    36.86    3    0    2
LVR    9    35.89    2    0    3
NOS    3    32.33    1    0    1
GBP    3    30.33    1    0    1
SEA    4    29.00    1    0    2
JAX    17    28.88    2    0    10
NYJ    12    26.50    2    0    5

 

[Leroy Hoard]

So best positions are DL-QB-OL and worst are TE-RB-LB?

 

[Anarchy99]

So best positions are DL-QB-OL and worst are TE-RB-LB?

If we believe CareerAV is set accurately (which it probably isn’t), the best ROI has been QB / RB / LB. the first table lists the positions in order.

 

[travdogg]

So best positions are DL-QB-OL and worst are TE-RB-LB?

If we believe CareerAV is set accurately (which it probably isn’t), the best ROI has been QB / RB / LB. the first table lists the positions in order.

Career AV is incredibly unfavorable for TEs, and grossly overrates RBs in my opinion.

For guys still on their 1st contract (like Chase) would an average per year method prove useful, or would that skew results too far into recency bias?

 

[-OZ-]

So best positions are DL-QB-OL and worst are TE-RB-LB?

If we believe CareerAV is set accurately (which it probably isn’t), the best ROI has been QB / RB / LB. the first table lists the positions in order.

Career AV is incredibly unfavorable for TEs, and grossly overrates RBs in my opinion.

For guys still on their 1st contract (like Chase) would an average per year method prove useful, or would that skew results too far into recency bias?

Wouldn’t average per year negate the impact of players who are completely out of the league? Or are you averaging per year since selected?
Meaning, if a player was selected in 2013 and has three seasons of avg 20, then out of the league, is his average 20 from when he played or 6 from when he was drafted?

 

[Hot Sauce Guy]

Not all drafts are equal. Averaging them out seems like not the best methodology. You seem to acknowledge this to some degree, so not harping on you, just sayin.

Maybe a better method would be to go player by player, using pre-draft scouting / ranking / ADP, and seeing what the hit rate was? Probably more difficult to divine. Then again, I was put under general anesthetic today, so I might also be thinking very stupidly right now. lol

I did enjoy the read though. 5 stars. Hopefully you get board/bored again soon.

 

[travdogg]

So best positions are DL-QB-OL and worst are TE-RB-LB?

If we believe CareerAV is set accurately (which it probably isn’t), the best ROI has been QB / RB / LB. the first table lists the positions in order.

Career AV is incredibly unfavorable for TEs, and grossly overrates RBs in my opinion.

For guys still on their 1st contract (like Chase) would an average per year method prove useful, or would that skew results too far into recency bias?

Wouldn’t average per year negate the impact of players who are completely out of the league? Or are you averaging per year since selected?
Meaning, if a player was selected in 2013 and has three seasons of avg 20, then out of the league, is his average 20 from when he played or 6 from when he was drafted?

Like average AP per year played. For example, Russell Wilson has 167 AV in 11 seasons, which is roughly 15 per season. At the same time, Matt Ryan has 210 in 15 seasons, which is an average of 14. So who is better? Or another example, Justin Herbert has 45 in 3 seasons for an average of 15, how can we rate that pick? Seems very unfair to call it equal to a 45 of a guy who played for a decade, but also seems unfair to just assume he'll be as good as Russell Wilson.

 

[Anarchy99]

So best positions are DL-QB-OL and worst are TE-RB-LB?

If we believe CareerAV is set accurately (which it probably isn’t), the best ROI has been QB / RB / LB. the first table lists the positions in order.

Career AV is incredibly unfavorable for TEs, and grossly overrates RBs in my opinion.

For guys still on their 1st contract (like Chase) would an average per year method prove useful, or would that skew results too far into recency bias?

Wouldn’t average per year negate the impact of players who are completely out of the league? Or are you averaging per year since selected?
Meaning, if a player was selected in 2013 and has three seasons of avg 20, then out of the league, is his average 20 from when he played or 6 from when he was drafted?

Like average AP per year played. For example, Russell Wilson has 167 AV in 11 seasons, which is roughly 15 per season. At the same time, Matt Ryan has 210 in 15 seasons, which is an average of 14. So who is better? Or another example, Justin Herbert has 45 in 3 seasons for an average of 15, how can we rate that pick? Seems very unfair to call it equal to a 45 of a guy who played for a decade, but also seems unfair to just assume he'll be as good as Russell Wilson.

But it works out for all players across all years. CareerAV gives credit for longevity over peak productivity in fewer years. I was trying to use a simple comparison. Eventually it will all even out when the players are done playing. Ten years from now, the current players will look Ryan or Rivers. Right now, it’s a snapshot of a point in time (that will evolve and change).

 

[Anarchy99]

By draft class . . .
Draft Year, Average AV of Top 10, AV of 50+, AV of 100+, AV of <25

2000        55.8    5    1    0
2001        63.6    6    2    2
2002        53.3    5    1    2
2003        63.3    6    3    2
2004        63.7    4    2    1
2005        41.8    3    0    2
2006        45.8    5    0    1
2007        45.6    4    1    4
2008        44.5    3    1    3
2009        39.0    2    1    3
2010        57.3    6    1    0
2011        69.9    7    4    2
2012        49.9    5    0    2
2013        29.5    2    0    6
2014        47.3    5    0    2
2015        42.4    6    0    2
2016        54.5    6    0    0
2017        41.2    3    0    2
2018        37.6    2    0    1
2019        31.0    1    0    4
2020        21.9    0    0    8
2021        15.8    0    0    10
2022        07.7    0    0    10

2013 . . . oof.

 

[Ilov80s]

2013 . . . oof.

Just an awful draft. I think the Lions got one of the best top 10 players and that was Ziggy Ansah who could never stay healthy.

 

[Bracie Smathers]

AVG Career AV = If we believe Career AV is set accurately (which it probably isn’t),

Could you share how the scores are determined and why they are broken out in quarter increments, basically the context of the numbers.

 

[Anarchy99]

AVG Career AV = If we believe Career AV is set accurately (which it probably isn’t),

Could you share how the scores are determined and why they are broken out in quarter increments, basically the context of the numbers.

HERE is a breakdown of how CareerAV is determined. What the secret sauce is is a bit of a mystery. Generally speaking, the longer a player plays, the higher his score will be. There have been some discussions over the years about how useful such a metric is, and which positions tend to score higher than others. There is no good way to compare players from different eras, and it's not really intended for that purpose (although people use it that way).

 

[Anarchy99]

Here's a look at each QB draft class. Whether people realize it or not, most QB's don't amount to much. I will get more in-depth about how QB's performed based on actual draft slot, round, and by their draft position (ie, the 3rd QB drafted in that particular draft) later.

QUARTERBACK DRAFT CLASS PERFORMANCE
Year, # of QB Selected, Total Weighted Career AV, 50 Points, 100 Points, <25 Points, 0 Points

    #    Total    50    100    <25    0
2022    09    22    0    0    9    2
2021    10    94    0    0    10    3
2020    13    151    0    0    10    5
2019    11    133    1    0    9    2
2018    13    222    2    0    9    5
2017    10    195    2    0    7    2
2016    15    261    3    0    11    5
2015    07    141    2    0    5    1
2014    14    247    1    0    10    3
2013    11    59    0    0    10    6
2012    11    471    4    1    5    3
2011    12    359    2    1    8    2
2010    13    81    0    0    12    7
2009    11    191    1    1    8    4
2008    13    282    2    1    10    7
2007    11    66    0    0    11    3
2006    13    197    1    0    11    4
2005    14    513    3    1    8    4
2004    17    504    4    3    13    7
2003    13    210    1    1    11    4
2002    15    201    1    0    11    6
2001    11    312    2    1    9    2
2000    12    319    3    1    9    6

- The 2013, 2010, and 2007 draft classes were terrible.
- Of 279 total QB selected, 216 scored <25 points (77%) and 93 (33%) scored 0 (meaning they never played or did nothing in their limited time on the field).
- Only 11 of 23 seasons produced two QB with scores of 50+ points (obviously recent classes will get more guys added, but there aren't many years where 3 guys end up scoring that high).

 

[SoBeDad]

Just curious, @Anarchy99 , what software are you using?

 

[Anarchy99]

Just curious, @Anarchy99 , what software are you using?

You give me credit for being way more high tech then I actually am. I am somewhere above using my fingers and toes but just below using an abacus. I cut and pasted the draft data into Excel and then sorted / organized from there. I am sure there were better ways to import the data, but that probably required a subscription.

 

[Tau837]

I appreciate the effort, and there is possibly no better way to do this. But AV is seriously flawed IMO. Aggregating some past comments about it:

• Assumption #1: The offensive line is exactly as good as the offense. This is obviously flawed, and Doug Drinen, AV creator admits it.
• Assumption #2: The offensive line is equally important in the running game as it is in the passing game. Drinen admitted this may be flawed, but he wasn't convinced of that. I think it is.
• Assumption #3: The ratio of pass-thrower importance to pass-catcher importance is constant from team to team. This is obviously flawed, and Drinen admitted this.
• All Pro and Pro Bowl selections are used for OL and defensive players to generate AV... but not for offensive skill players. This amounts to another assumption, that the offensive skill position scoring correlates to All Pro levels and Pro Bowls. Sometimes that is valid, but not always.
• Team AV points are based on points per drive scored/allowed. No elements for yards gained/allowed, points scored/allowed total (not per drive) or any other offensive or defensive metrics.
• Points are mostly based on yards, ignoring receptions, TDs, first downs, sacks taken, and offensive turnovers. This results in situations like this one for the 2018 Chargers:
  • Mike Williams - rushing: 7/28/1 with 2 first downs; receiving: 66 targets, 43/664/10 with 42 first downs; 0 fumbles = 6 AV
  • Tyrell Williams - rushing: 2/15/0 with 1 first down; receiving: 65 targets, 41/653/5 with 34 first downs; 2 fumbles = 6 AV
  • Mike was clearly much more positively impactful, yet their AV scores were the same.
• Credit for efficiency is limited. For example, runners only get an adjustment (up or down) for YPC if they have 200+ carries in a season.
• Drinen basically admits that he doesn't know how to positionally divide up OL or defensive credit, which is not surprising since that is a complex problem. But his answer is to assign fixed positional values. So a tackle gets 20% more credit than a guard or center and 2/3 of defensive points go to the front 7 and 1/3 to the secondary. These choices are fine for generalizing positional value but could vary for individual situations.
• For defense, a given team's total defense points to be divided up among the defensive players is based exclusively on defensive points allowed per drive.
  • 1/3 of those points are assigned to defensive backs and 2/3 to the rest of the position groups, independent of how good those groups are on any given defense.
    • This disproportionately hurts a player like Ronde Barber, who played with another HOF defensive back.
  • This methodology is also dated, since teams play 5+ DBs more often than they play just 4, so this splitting of the points shorts the DB group nowadays.
  • Defensive backs are given zero credit for tackles
    • Tackles was a big strength for Ronde Barber, and he gets zero credit for it.
  • Defensive players get points for interceptions but not for passes defended.
    • PD was a big strength for Ronde Barber, and he gets zero credit for it.
  • Defensive players get points for fumble recoveries, but not for forced fumbles.
  • Defensive players get points for sacks, but not for other pressures (QB hits, hurries).
• AV accounts only for regular season games; it ignores postseason games.

I get that AV is an approximation tool, and that's fine. Just saying that it has flaws.

 

[Snotbubbles]

To me, the problem with drafting a RB in the Top 10 isn't that I'm concerned they will bust. The problem is their shelf life. When I draft a guy in the top 10, I want him around and playing at a high level for 10+ years. You don't see too many 33 year old stud running backs but offensive linemen and quarterbacks can play at a high level well into their 30s.