## WR Bust Factors: Adjusting for Over/Under Representation

This post has been about a year in the making. BED’s post about ‘safe’ OT was the catalyst to finally post it as the logic is relevant to the discussion of OT as well

I believe that the 50% bust factor for wide receivers drafted in the first round is overstated. I frequently see comments that wide receivers are the riskiest draft picks and the bust factor for wide receivers is 50% (compared to an NFL average bust factor of 28% for all first round picks). I believe that the 50% figure is wrong because it does not adjust for overrepresentation / underrepresentation of wide receivers drafted in the first round. Once the wide receiver bust factor is corrected for overrepresentation it is 39% versus a 28% bust factor for all 1st round draft picks.

From 1998-2007, wide receivers were a disproportionate percentage of first round draft picks. During that period there were 43 wide receivers drafted in the first round (4.3/year) out of a total of 314 total picks (43 / 314 = 14%). However, I estimate that wide receivers account for only 2.5 players of the 22 players on the field or 11% (2.5/22 = 11%)[i]. If wide receivers had been drafted in proportion to their representation on the field there would have been 25 wide receivers drafted in the first round instead of 43. I believe that the higher than average bust factor is due to the extra 18 wide receivers drafted in the first round.

I previously mentioned this hypothesis in a prior post. In the original post on this subject RickT had a common sense suggestion on how to test this hypothesis.

But it seems very testable. Is the bust factor lower if you just look at the first 3-4 WRs selected in a draft? Or, as you allude to, is the bust factor for all WRs selected at a given spot comparable to other players at the same position

RickT

As suggested by RickT I looked at the top three wide receivers drafted in the first round from 1998-2007 (this still overrepresents wide receivers as it produces a set of 28 wide receivers instead of 25). Using the new dataset I recalculated the bust factor. The initial characterization of each player and the calculation of the bust factors for both data sets is summarized in the table below (zero denotes a bust).

 1st Round: Top 3 Bust = 0 1st Round Bust = 0 Calvin Johnson 1 Calvin Johnson 1 Ginn, Jr., Ted 0 Ginn, Jr., Ted 0 Bowe, Dwayne 1 Bowe, Dwayne 1 Meachem, Robert 1 Davis, Craig 0 Gonzalez, Anthony 0 Santonio Holmes 1 Santonio Holmes 1 Braylon Edwards 1 Braylon Edwards 1 Troy Williamson 0 Troy Williamson 0 Mike Williams 0 Mike Williams 0 Matt Jones 0 Mark Clayton 1 Roddy White 1 Larry Fitzgerald 1 Larry Fitzgerald 1 Roy Williams 1 Roy Williams 1 Reggie Williams 0 Reggie Williams 0 Lee Evans 1 Michael Clayton 0 Michael Jenkins 1 Rashaun Woods 0 Charles Rogers 0 Charles Rogers 0 Andre Johnson 1 Andre Johnson 1 Bryant Johnson 1 Bryant Johnson 1 Donté Stallworth 1 Donté Stallworth 1 Ashley Lelie 0 Ashley Lelie 0 Javon Walker 1 Javon Walker 1 Terrell, David 0 Terrell, David 0 Robinson, Koren 1 Robinson, Koren 1 Gardner, Rod 0 Gardner, Rod 0 Moss, Santana 1 Mitchell, Freddie 0 Wayne, Reggie 1 Peter Warrick 0 Peter Warrick 0 Plaxico Burress 1 Plaxico Burress 1 Travis Taylor 1 Travis Taylor 1 Sylvester Morris 0 R. Jay Soward 0 Torry Holt 1 Torry Holt 1 David Boston 1 David Boston 1 Troy Edwards 0 Troy Edwards 0 Kevin Dyson 1 Kevin Dyson 1 Randy Moss 1 Randy Moss 1 Marcus Nash 0 Marcus Nash 0 Total 28 43 Busts 11 19 Bust Factor 39.3% 44.2%

As shown in the table once overrepresentation is adjusted for the bust factor for wide receivers is 39%. So the answer is mixed. 39% is still significantly higher than the 28% average for the NFL but it’s lower than 50%.There are two other factors. First, to the extent there is overrepresentation of wide receivers there must be underrepresentation and other positions, which reduces the bust factor of other positions lowering the average for all positions. In addition, there are several players that I believe are questionable to characterize as busts (Ted Ginn, Mike Williams, Rod Gardner, Peter Warrick, Troy Edwards). With a set of 28 players, each success/bust changes the bust factor by 3.5%. Therefore reclassifying 1 or 2 of those busts has a material impact on the bust factor.

Why.

As I am proposing a new hypothesis I believe it is incumbent upon me to explain why NFL teams are choosing to draft a disproportionate number of wide receivers. I believe NFL teams are drafting a disproportionate number of wide receivers because they believe wide receivers have a disproportionate impact on winning. Let me give an example.

Let's use the standard grading scale of 1 to 100 used by scouts.

100-90: Elite Player

89-80: Outstanding Player

79-70: Good Starter

69-60: Average Starter

Assume that all the players on a team are average starters (i.e. grade of 60). Let's further assume the following: that replacing an average wide receiver (grade equal to 60) with an elite wide receiver (grade equal to 100) increases a team's win total by two wins, and that replacing an average guard (grade equal to 60) with an elite guard (grade equal to 100) increases  a team's win total by one win.

There are a few observations that we can make. First, a team should be indifferent between selecting a wide receiver with a grade of 80 and an offensive guard with a grade of 100. Either player will improve the expected win total by one game.

We can show this mathematically by calculating the change in wins for the change in scouting grade ("SG"). Assume that replacing an average wide receiver with an elite wide receiver (i.e. 60->100) is worth two wins. Then each point change in scouting grade is worth 0.05 wins (2 wins/40 SG = 0.05 wins/SG). Assume that replacing an average offensive guard with an elite player is worth one win or 0.025 wins/SG (1 win/40 SG = 0.025 wins/SG).

We can show that replacing an average wide receiver (SG = 60) with a good wide receiver (SG = 80) will increase wins by one game (SG change of 20 x 0.05 wins/SG = 1 win). Continuing with that logic one must replace an average guard (SG = 60) with an elite guard (SG = 100) to achieve the same expected one win improvement (SG change of 40 x 0.025 wins/SG = 1 win). I believe this passes the smell test. The casual fan recognizes that replacing an average wide receiver with an elite wide receiver (Crayton -> Andre Johnson) is likely to have a much greater impact than replacing an average guard with an elite guard (Kossier -> Hutchinson).

It warrants mentioning that this is the exact logic applied by St. Louis last year. Although the consensus was that Suh was the best college football player last year, St. Louis drafted Bradford because they realized that a quarterback has a greater impact on winning than a defensive tackle.

Furthermore, I assume we can all agree that players with a scouting grade of 80 a more likely to be busts than players with the scouting grade of 100.  We can see that from the team perspective it’s just as logical to draft a WR with a grade of 80 as it is to draft an OG with a grade of 100, even though the WR will be more likely to be a bust.

Conclusion

My initial hypothesis was that the 50% bust factor for WR could be explained by their over-representation in the 1st round. The evidence does not support that conclusion. While over-representation does provide a partial explanation, the fact is that even after adjusting for over-representation the bust factor for WR’s is higher than the average for all picks.

Appendix – OT

Here are the OTs taken over the last 10 years.

 2010 4 Trent Williams, Russell Okung, Anthony Davis, Brian Bulaga 2009 4 2008 7 Jake Long, Ryan Clady, Chris Williams, Gosder Cherilus, Jeff Otah, Sam Baker, Duane Brown 2007 3 Joe Thomas, Levi Brown, Joe Staley 2006 1 D'Bricksaw Ferguson 2005 2 Jamaal Brown, Alex Barron 2004 3 2003 3 2002 4 Mike Williams, Bryant McKinnie, Levi Jones, Marc Colombo 2001 3 Leonard Davis, Kenyatta Walker, Jeff Baukus

Observation: Over the last 3 years OT are being over-represented in the first round. There are 22 starting positions of which 2 are OT. If all positions were of equal value we would expect players to be drafted in proportion to their position. E.g. for OT 2 / 22 … ~ 10% or 3 1st round picks per year. You can see that’s roughly how it was through 2001-2007 … but for 2008-2010 OT has been over-represented every year. We can expect that bust factors for OT will increase as they get pushed up in the draft.

I suspect the same logic is at play with OTs that is at play with WRs. It can be perfectly rational to take an OT even if the bust factors for OTs are increasing because they have a disproportionate impact on wins. Just be aware that bust factors will likely increase as OT are drafted higher.

[i] sports illustrated NFL preview: three or more wide receivers lined up on 49.2% of offensive snaps, there were two wideouts on 38.6%. Therefore, assume one wide receiver on 10% of offensive plays, two wide receivers on 40% of all offensive plays, three wide receivers on 40% of all offensive plays, and four wide receivers on 10% of all offensive plays.  1 x 10% + 2 x 40% + 3 x 40% + 4 x 10% = 2.5.

Another user-created commentary provided by a BTB reader.

## Trending Discussions

forgot?

As part of the new SB Nation launch, prior users will need to choose a permanent username, along with a new password.

I already have a Vox Media account!

### Verify Vox Media account

As part of the new SB Nation launch, prior MT authors will need to choose a new username and password.

We'll email you a reset link.

Try another email?

### Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

### Join Blogging The Boys

You must be a member of Blogging The Boys to participate.

We have our own Community Guidelines at Blogging The Boys. You should read them.

### Join Blogging The Boys

You must be a member of Blogging The Boys to participate.

We have our own Community Guidelines at Blogging The Boys. You should read them.