What's new
Fantasy Football - Footballguys Forums

Welcome to Our Forums. Once you've registered and logged in, you're primed to talk football, among other topics, with the sharpest and most experienced fantasy players on the internet.

comparing static VBD vs ADP: overrated and underrated players (1 Viewer)

moleculo

moleculo

Footballguy
TL;DR> TE's and QB's are underrated, RB's are overrated.

using free projections from around the web (FFToday, Fantasy Sharks, ESPN, CBS).  I have generated a set of VBD rankings assuming PPR and 4 points per passing TD.  I was curious to see how this VBD ranking compares to average draft position (using ADP from fantasyfootballcalculator.com, but the general trend seems to hold).  Comparing player to player and seeing the number of spots isn't a very good way to do this because a difference of a few spots in round 1 is much more significant than a few spots in round 10.

Instead, I use the draft trade chart to determine value in both VBD rankings and ADP rankings, and compared the difference.  Here's how it works: Cam Newton, for example, has a VBD  value of 78 points.  That places him 18th overall, worth 900 points.  In ADP, Cam is being drafted at 3.11: good for 550 points.  comparing the two, Newton is undervalued by 350 points.

I do this for all players, and I have a good sense of who is underrated and who is overrated (per my projections and ADP).

the top 10 underrated players:

  1. Rob Gronkowsk NE
  2. Jordan Reed WAS
  3. Alshon Jeffery CHI
  4. Keenan Allen SD
  5. Cam Newton CAR
  6. Russel Wilson SEA
  7. Travis Kelce KC
  8. Antoni Gates SD
  9. Zach Ertz PHI
  10. Tom Brady NE
the top 10 overrated players:

  1. Adrian Peterson MIN
  2. Ezekiel Elliott DAL
  3. Dion Lewis NE
  4. Allen Robinson JAC
  5. Le'Veon Bell PIT
  6. Thomas Rawls SEA
  7. Eddie Lacy GB
  8. A.J. Green CIN
  9. Lamar Miller HOU
  10. Josh Gordon CLE
Discussion:

a clear pattern emerges: TE dominates the underrated list and RB dominates the overrated list.  I only displayed the top 10 of each list, but if I extend out to 20, 5 of the top 20 are QB's and 8 are TE's.  There are no TE's or QB's anywhere on the overrated side.  Conversely, 16 RB's are overrated (none underrated).  That being said, I'm not sure what to do with that info.  If your league has an early run on RB's and you miss out, that puts you in a bad spot if you bulk up on QB's and TE's.  It's important to keep ADP in mind and try to stay ahead of the runs, IMO.

a few points to note:

  • For Brady, I prorated his projections so he should be drafted as if he were playing 16 games.
  • ADP data is an average of drafts, many of which happened before Dion Lewis injury news was out there.  Also, LeVeon Bell's suspension.
  • I can get these to shift around by playing with the relative baselines.  The above uses worst starter as VBD baseline.  I will continue to tweak to see if I can find a better balance.
Conclusions: 

If you believe in VBD and last-starter baseline, you should go into the draft with the understanding that nearly all TE's are underrated and many QB's are.  You should be aware that nearly all RB's are being drafted earlier than they should be. That being said, the clear trend of most RB's being overvalued and TE's being undervalued leads me to think that worst-starter may not be an appropriate baseline.  to be continued...

 
Last edited by a moderator:
If I set the following as baselines, everything balances out (i.e. just as many QB's overvalued as undervalued, just as many RB's overvalued as undervalued, etc):

QB6

RB34

WR36

TE6

I'm not completely sure what to do with this info, but it's interesting I think.  This is an exercise in making VBD baselines match up roughly with ADP.

One thing to note: changing the TE baseline from TE12 to TE6 changed Gronkowski from the most undervalued player to the 4th most overvalued.  Again, I'm not completely sure what that means.

 
It would make sense that QBs and TEs be in the category of most undervalued because their ADP is being suppressed by supply and demand. In most leagues you are only starting 1 QB and 1 TE where in most leagues you start 2-4 RBs and 2-4 WRs. So even if you take Gronk and he gets hurt, it's much more likely you can find that TE13 on the WW than it would be to find RB25 if Peterson is hurt or WR37 if AJ Green is hurt. 

 
There are a couple problems with VBD that don't allow it to be used as the end all be all strategy to dominate FF.

1.  Predictability by position.  The reason no one looks at kickers and defenses in VBD terms is that they are not predictable year to year.  QB's are also not very predictable.  For instance Rivers, Palmer and Newton all went for less than 1% of auction budget last year and Rodgers and Luck when for 25% and 21%.  

2.  VBD doesn't mean anything unless you take into account the cost of acquiring it.  It looks like you are starting to look at that, but you have to take it a step further.  Your baseline is not best starter but should be more like best free or near free to acquire during the draft.  For instance this year, winston, Fitzpatrick and Stafford are crazy cheap and I have them projected for similar if not more production than Big Ben, Palmer, and Bortles. So my base line for QB is like QB6 or 7.  This lowers the top tier guys to like 4/5th round value.  This is the same for TE although not as drastic.  I have Gronk as a late second round value. 

3.  Week to Week variance of the position.  I did some research on this last year and I don't have my data for some reason, but the gist was that the high variance of the top RB's and WR's tends to win you more weeks if you have several of them vs one of them and a top QB and or TE.  The standard deviation of the the top 24 RB's and WR's is higher than that of the QB's and by having 4-5 of the top RB/WR's you have a better chance of very high weekly scores than if you have fewer and better QB's and TE's.  I ran a couple Monte Carlo sims with various make ups of players to prove it to my self.  I started looking at this due to an article on numberfire published a couple years ago that showed that over the last 5 seasons in several leagues the winning team each week would have still won the game 70% of the time with out the QB's score.  For TE's its even higher.  I went back and looked at my main league for the last couple years and had very similar results.  

Using year end projections is an issue as well.  I think if you took projections on a PPG basis you would find that RB's are not as overvalued.  But most projections take into consideration that injuries will happen which effects RB's more.  As long as FF is played as a series of weekly contests, PPG will be more important than year end projections.  

From my numbers I show the most undervalued players are RB2's and the most overvalued are QB's and TE's.  WR's seem to be mostly fairly valued with a few exceptions that are likely just my flag plant guys (Allen, Fitz, Jones are going to cheap and Hopkins, Jeffery, Nelson, Hilton to high).  

I think it is interesting to see where the market is in relation to vbd and thanks for starting the conversation.  

 
abbottjamesr said:
There are a couple problems with VBD that don't allow it to be used as the end all be all strategy to dominate FF.

1.  Predictability by position.  The reason no one looks at kickers and defenses in VBD terms is that they are not predictable year to year.  QB's are also not very predictable.  For instance Rivers, Palmer and Newton all went for less than 1% of auction budget last year and Rodgers and Luck when for 25% and 21%.  

2.  VBD doesn't mean anything unless you take into account the cost of acquiring it.  It looks like you are starting to look at that, but you have to take it a step further.  Your baseline is not best starter but should be more like best free or near free to acquire during the draft.  For instance this year, winston, Fitzpatrick and Stafford are crazy cheap and I have them projected for similar if not more production than Big Ben, Palmer, and Bortles. So my base line for QB is like QB6 or 7.  This lowers the top tier guys to like 4/5th round value.  This is the same for TE although not as drastic.  I have Gronk as a late second round value. 

3.  Week to Week variance of the position.  I did some research on this last year and I don't have my data for some reason, but the gist was that the high variance of the top RB's and WR's tends to win you more weeks if you have several of them vs one of them and a top QB and or TE.  The standard deviation of the the top 24 RB's and WR's is higher than that of the QB's and by having 4-5 of the top RB/WR's you have a better chance of very high weekly scores than if you have fewer and better QB's and TE's.  I ran a couple Monte Carlo sims with various make ups of players to prove it to my self.  I started looking at this due to an article on numberfire published a couple years ago that showed that over the last 5 seasons in several leagues the winning team each week would have still won the game 70% of the time with out the QB's score.  For TE's its even higher.  I went back and looked at my main league for the last couple years and had very similar results.  

Using year end projections is an issue as well.  I think if you took projections on a PPG basis you would find that RB's are not as overvalued.  But most projections take into consideration that injuries will happen which effects RB's more.  As long as FF is played as a series of weekly contests, PPG will be more important than year end projections.  

From my numbers I show the most undervalued players are RB2's and the most overvalued are QB's and TE's.  WR's seem to be mostly fairly valued with a few exceptions that are likely just my flag plant guys (Allen, Fitz, Jones are going to cheap and Hopkins, Jeffery, Nelson, Hilton to high).  

I think it is interesting to see where the market is in relation to vbd and thanks for starting the conversation.  
Click to expand...
Good stuff.  I should say that I actually use a dynamic VBD to draft with, and I use ADP to generate the moving baseline as the draft progresses.  That being said, I do like to find static VBD pre-draft for analysis like this, but also to print up a 1page sheet to have just in case my computer craps out during the draft.

To your more specific points:

1. I'm not sure I buy that.  There is turnover at the high end of all positions.  Defenses (and especially kickers) are drafted late because the best one isn't all that much better than the 15th one.

2. It sounds like you have adjusted your QB baseline up, which is what I found I had to do to make my overrated/underrated lists even out (i.e. as many QBs overrated as underrated, relative to ADP).

3. Agreed, but I don't really have the ability to quantify this.  I will say I'm not comfortable projecting injury or games missed.

 
moleculo said:
Good stuff.  I should say that I actually use a dynamic VBD to draft with, and I use ADP to generate the moving baseline as the draft progresses.  That being said, I do like to find static VBD pre-draft for analysis like this, but also to print up a 1page sheet to have just in case my computer craps out during the draft.

To your more specific points:

1. I'm not sure I buy that.  There is turnover at the high end of all positions.  Defenses (and especially kickers) are drafted late because the best one isn't all that much better than the 15th one.

2. It sounds like you have adjusted your QB baseline up, which is what I found I had to do to make my overrated/underrated lists even out (i.e. as many QBs overrated as underrated, relative to ADP).

3. Agreed, but I don't really have the ability to quantify this.  I will say I'm not comfortable projecting injury or games missed.
Click to expand...
That is not the whole story though.  The Denver DST last year would had 71 points of VBD over the 12th ranked DST, would have been worth RB7 or WR8 in VBD.  The problem is they weren't projected to score as well.  

VBD is great to get a gauge of where the market is inefficient but it won't win you your league.  Getting a couple extra points of VBD is great but its not the difference maker.  The key is to get massive VBD for near nothing.  This comes from players that have break out years that the world wasn't expecting and you being the first to jump.  Josh Gordon 2013, Freeman 2015, Cam 2015, Luck 2014, etc...  The good thing for us is that the experts are not willing to go nuts with there rankings on guys that they think will break out but, if you listen you can hear who the experts think it might be and adjust your projections up so that they are an appealing looking value by VBD.   I have a list of guys that I think are going to destroy their ADPs in production, my projections aren't pie in the sky, but they reflect my optimism and as such they show as great values in my board.  

My guys this year
QB: Wilson, Fitzpatrick
RB: Miller, Charles, McCoy, Anderson, Mathews, Crowell
WR: Allen, Fitzgerald, Jones, Hurns, Gordon
TE: V. Green, M. Bennett, Gates

 
Last edited by a moderator:
You must log in or register to reply here.

Similar threads

A
What's Your Process?
Replies
2
Views
618
acarey50
A
FFCollusion
Juggernaut RB Theory
Replies
3
Views
682
Da Franchise
D
M
Looking at perception of 2022 class vs 2023 class at similar juncture
Replies
21
Views
875
menobrown
M

Users who are viewing this thread

Total: 2 (members: 0, guests: 2)
Top