VBD Baseline Questions (1 Viewer)

[Seven Costanza]

Hey guys! I know there are various methods for determining VBD baselines, but I prefer to use the replacement value method and basically look at my stat projections and see where the players start to look like "just a guy."

Below is a range of my projections by position that I feel like my baseline ought to fall in, I would love to hear some opinions on where to set my baseline.

At QB, my 14th - 21st QBs are as follows:

14 Russell Wilson

15 Cam Newton

16 Ben Roethlisberger

17 Andy Dalton

18 Alex Smith

19 Ryan Tannehill

20 Joe Flacco

21 Carson Palmer

I am thinking of using Alex Smith as my baseline for QB....thoughts?

At RB, my 30th - 40th RBs are as follows:

30 Trent Richardson

31 Danny Woodhead

32 Lamar Miller

33 Maurice Jones-Drew

34 Steven Jackson

35 DeAngelo Williams

36 Darren Sproles

37 Stevan Ridley

38 Bernard Pierce

39 Devonta Freeman

40 Mark Ingram

I am really torn on who to use as my baseline for RB...I was thinking Trent Richardson, but I have a hard time putting Sproles and Ridley below the baseline as "just a guy"...thoughts?

At WR, my 51st - 63rd WR are as follows:

51 Doug Baldwin

52 Riley Cooper

53 Rod Streater

54 Danny Amendola

55 Jarrett Boykin

56 Kenny Stills

57 Marqise Lee

58 Jordan Matthews

59 Harry Douglas

60 Hakeem Nicks

61 Andrew Hawkins

62 Wes Welker

Similar to RB I am torn here, i feel like at the Baldwin, Cooper, Streater range they start to look like just guys, but I struggle to put Hawkins and Welker below the baseline....thoughts?

At TE, my 5th - 19th TE are as follows:

5 Greg Olsen

6 Jason Witten

7 Dennis Pitta

8 Kyle Rudolph

9 Jordan Reed

10 Vernon Davis

11 Zach Ertz

12 Martellus Bennett

13 Antonio Gates

14 Charles Clay

15 Heath Miller

16 Tyler Eifert

17 Delanie Walker

18 Dwayne Allen

19 Ladarius Green

Personally I feel like after Jordan Cameron at 4 the TEs start to look like just a guy with Greg Olsen, but using Olsen as my baseline really depresses the X values of the top 4 TEs so I am not sure that is fair. So then I considered using Ladarius Green as my baseline, that way I would include some of the upside younger guys like Ertz, Eifert, and Allen above the baseline....thoughts?

I know this is a long post, but I really appreciate any input....aside from the stat projections, setting the baseline is the most important part of the VBD approach in my opinion!!

 

[Seven Costanza]

Bump...anybody?

 

[moleculo]

I don't think anyone has proposed the JAG baseline. I like the concept, but where the JAG guy starts will be dependent on your personal preferences much more than scoring systems. to me, a JAG analysis would break down after the 2nd or 3rd round. Everyone looks about the same to me after the elite guys are gone.

Maybe the way to do it is look at your projected scoring, look at tiers, and set the JAG baseline at the end of a tier where it makes sense (tier just before last starter, maybe).

Here's my projected QB scoring:

Peyton Manning DEN 513
Drew Brees NO 481
Aaron Rodgers GB 478
Matthe Stafford DET 438
Nick Foles PHI 415
Andrew Luck IND 410
Matt Ryan ATL 409
Tony Romo DAL 403
Tom Brady NE 401
Colin Kaepernic SF 401
Robert Griffin I WAS 394
Jay Cutler CHI 390
Cam Newton CAR 390
Philip Rivers SD 388
Ben Roethlisb PIT 380
Russel Wilson SEA 378
Andy Dalton CIN 363
Carson Palmer ARI 363
Ryan Tannehill MIA 353
Alex Smith KC 340
Joe Flacco BAL 337
Eli Manning NYG 334
Sam Bradford STL 327
Josh McCown TB 303
EJ Manuel BUF 282

using last starter as baseline, that would be cutler @ 390, but is 390 really all that different than Romo at 403? That's less than 1 point per game, well within the accuracy of a prediction so maybe Romo is just a guy. After all, there are 7 guys within a 15 point band - is there really any difference between Romo and Rivers (QB14)?

 

[Modog814]

It's hard to say because my rankings don't align with yours. Personally I see Big Ben as just a guy, so I could draw the line there, but I have Tannehill, Palmer, Dalton and Smith rated higher.

Not to mention "just a guy" is a vague concept. Just a guy compared to who? I mean yeah every TE is pretty much just a guy compared to Graham/Gronk/Thomas. So if that's your criteria, then how compare every QB after Peyton/Brees/Rodgers isn't also "just a guy"?

I think you need to better define what your criteria is... and maybe then people could help you out.

 

[CalBear]

I've moved away from setting baselines based on players; instead, I look at past years for the year-ending point total where I consider replacement value to be, and using that point value as the baseline. The problem with choosing an individual player to set baselines for a position is that then you're basing the baseline number on someone's projection for that person. Projections for deep players don't tend to be very good.

So, when I look at last year's RB stats, in my main league (slightly unconventional scoring), the "just a guys" start around 77 points (RB34, Jacquizz Rodgers), but there are still some good players down below who missed time or showed late value (Vereen, Montee Ball, McFadden). After Marcel Reece (RB43, 63 points), pretty much everyone is just a guy. So I'd set the baseline somewhere around 63 points.

Here's where the difference comes in. Based on Dodds' projections, RB34 will score 97 points, and RB43 will score 69. I'm pretty sure that won't happen. So while Devonta Freeman (Dodds' RB43) might be a good place to draw the baseline, using 69 points will set the baseline too high.

 

[Seven Costanza]

Man I never knew that determing a baseline could be so hard!!

 

[Seven Costanza]

I guess by JAG I mean something similar to your replacement value concept....for example, based on my projections I would think the replacement value at RB would be somewhere between Danny Woodhead and Bernard Pierce...no?

 

[Modog814]

If I were you, the first thing I'd do is find the natural breaking points. That is, go down each of your positional lists and ask yourself, would you be equally happy getting Wilson or Newton? If no, then you at the least have found a tier break, if not your baseline. If the answer was yes, repeat with Newton and Roethlisberger, and keep going until you find the answer to be no.

 

[ZWK]

What about using "best available in the 12th round" for a snake draft, or "best available for $1" for an auction?

 

[GregR]

What about using "best available in the 12th round" for a snake draft, or "best available for $1" for an auction?

I do like looking at things from an auction perspective because the fact you had to put a dollar figure to every player helps give you something concrete to test your VBD results against.

To me the big issue that's already brought up is the projections for the players in that area. I wouldn't say the problem is so much that the projections there aren't good as that what is being projected isn't what we base our concept of value on for those players. We don't draft those players based on them scoring 60 points as a backup, we draft/auction Grimes based on what he's worth if Foster goes down compared to what Jerrick McKinnon is worth if Adrian Peterson goes down. The projection doesn't really attempt to capture that.

I've felt for awhile that to do things justice, we need two baselines and sets of projections, one for starters and key backups where what we think they will actually score expresses most of their value, and then another baseline and set of projections for the players we draft for potential and security. And they need to be consistent enough with each other where 1 VBD point of backup value is still equal to 1 point of starter value so you can get at things like auction salaries.

 

[GregR]

One other thing I've never liked is how much the values depend on the baseline player when using a specific player. This was illustrated very well this season, when I was doing auction values for my 2 QB league. I moved the baseline for QB by 1 spot and my values changed drastically. It was because the difference between Matt Schaub and the player ahead of him at the time was like 40 fantasy points, so that was a huge change then to all the other QB's value.

From that standpoint, I do think there may be some real merit to looking at historical results and using hard numbers instead of choosing players from this year's crop and their projection.

 

[Biabreakable]

I wrote an article about this here which was published August 12th. This is focused on dynasty, but would be applicable to redraft as well.

The focus of this article is on 3 baselines which I called-

1. Core level player baseline.

2. Worst starter baseline.

3. Worst roster spot//Replacement level baseline.

Using a 3 year average for leagues of 12 teams at 20 roster spots for standard and PPR leagues that start 1 QB 2RB 3WR 1TE and 1 Flex (RB/WRTE).1K 1DST

I think it is important to fit your baselines specifically to your league scoring system and starting requirements. But the baselines in this article covers the most common formats.

The 3 year average at each position of these categories is-

Core player baseline-

Standard leagues:

Quarterback 12 302 points or 18.9 points per game

Running back 24 134 points or 8.6 points per game ( last 2 seasons 132 or 8.27 )

Wide receiver 24 130 points or 8.1 points per game

Tight end 12 95 points or 5.9 points per game

Points for receptions leagues:

Quarterback 12 302 points or 18.9 points per game

Running back 24 167 or 10.4 points per game

Wide receiver 24 198 or 12.3 points per game

Tight end 12 155 or 9.7 points per game


Worst starter baseline-

Standard leagues:

Quarterback 24 212 points or 13.2 points per game

Running back 36 98 points or 6.1 points per game

Wide receiver 48 96 points or 6 points per game

Tight end 24 68 points or 4.3 points per game


Points for receptions leagues:

Quarterback 24 212 points or 13.2 points per game

Running back 36 167 points or 10.4 points per game

Wide receiver 48 145 points or 9 points per game

Tight end 24 106 points or 6.7 points per game


Worst roster spot or replacement level baseline-

Replacement level value is 36 points or 2.25 points per game in standard leagues .

Replacement level value is 49 points or 3 points per game in point for receptions leagues.


So using these baselines as a guide you then match whatever projections you are using to determine which category each player is in.

 

[CalBear]

One other thing I've never liked is how much the values depend on the baseline player when using a specific player. This was illustrated very well this season, when I was doing auction values for my 2 QB league. I moved the baseline for QB by 1 spot and my values changed drastically. It was because the difference between Matt Schaub and the player ahead of him at the time was like 40 fantasy points, so that was a huge change then to all the other QB's value.

From that standpoint, I do think there may be some real merit to looking at historical results and using hard numbers instead of choosing players from this year's crop and their projection.

My main league is an auction league, and that's exactly why I went to the point-based baseline. Even switching from one staff member to another for projections can throw the values way out of whack. A difference of 40 points in the QB baseline could mean a difference of 10% in the total allocation of dollars to the QB position relative to other positions, which is a pretty big deal in an auction, especially because the valuation curve isn't linear. A 10% difference in positional value is probably a 15% difference in value for the top players; it could be the difference between Manning being worth $30 and $35.