VOR (Value Over Replacement) Drafting - Fantasy Football Cafe 2014 Fantasy Football Cafe


Return to Football Talk

VOR (Value Over Replacement) Drafting

Moderator: Football Moderators

VOR (Value Over Replacement) Drafting

Postby Wood Chipper » Sat Jul 21, 2012 2:08 pm

Has anyone else tried VOR drafting? I just started using it recently in mocks and wow. My team has started to look a bit better.
Wooden Icicles PPR (Horse Traders Dynasty)
Wood Chipper
Special Teams Staff
Special Teams Staff

User avatar
CafeholicCafe Writer
Posts: 330
Joined: 17 Jul 2012
Home Cafe: Football
Location: Knoxville, TN

Re: VOR (Value Over Replacement) Drafting

Postby Firemoth » Mon Jul 23, 2012 8:29 am

I'm afraid I would need more of a definition to discuss this. Wikipedia lists Value Over Replacement as a baseball statistic.

Could you describe a Value Over Replacement approach, or how it is different than this top google hit for "Value Based Drafting": http://fantasystrategies.com/content/dr ... d-drafting ?
Firemoth
Special Teams Staff
Special Teams Staff


Posts: 116
(Past Year: 5)
Joined: 25 Jul 2005
Home Cafe: Football

Re: VOR (Value Over Replacement) Drafting

Postby Wood Chipper » Mon Jul 23, 2012 6:07 pm

"An Introduction to Value Over Replacement (VOR)

Being an industrial engineer, I am a numbers and efficiency-driven person. In the end, fantasy football is going to come down to how good your projections are. Therefore, the point is to get the most out of your projections every year. So blanket strategies of going RB/RB in the first two round of a draft, or any combination of two positions for that matter, just don’t cut it for me.

So this project started out as a just a way to determine any given player’s value over the first non-starter for different league types. That value I called VOR (Value Over Replacement), which is difference in the fantasy points per game (FPG) projected for that player over the first projected non-starter at his position. For example, in a 12-team league that starts 2 RBs, the VOR of projected RB #4 would be the difference between his projected FPG over the projected FPG of RB #25 (by definition, the VOR of the first non-starter at any given position is 0.0). The numbers prove to be a very useful way to value players early in your draft, and to determine the proper time to select a player.

But as with anything in fantasy football, it’s not as simple as plugging in numbers and just going on cruise control. After figuring out the values, it became pretty obvious that just drafting based on VOR isn’t the best way to go, because everyone else in the draft isn’t drafting that way. If you’re drafting solely on a player’s VOR, you may draft a player in the second round when you could have had a different but somewhat equivalent player a round or two later, and instead drafted a position with greater positional drop-off in a higher spot.

What do I mean? For example, in a standard scoring 12-team league, starting two RBs and three WRs, I drafted #1 overall. My pick in the late 2nd round came to me after taking Foster #1. On the board, the top two VORs are Roddy White at 3.375 VOR and Steve Smith at 3.313 VOR, with RBs Michael Turner at 3.013 and Steven Jackson at 3. But instead of selecting the top two VOR players without thinking, it’s important to look forward to my next two picks, where Dwayne Bowe at 2.906 and Jeremy Maclin at 2.063 are projected to be there. By contrast, at RB, Willis McGahee at 0.688 and BenJarvus Green-Ellis at 0.563 are all that is projected to be there. According to the numbers, the drop-off between those RBs and those projected to be available a few rounds later is minimal. In the earlier rounds, there’s no reason to take a back up over a starter (I can only start two RBs), so I take White and Turner, the best overall pick based on VOR, and the best RB at a position with a huge drop-off. The numbers are helpful in letting me know I can get good WRs in the next round who are comparable to Smith, but the RB drop-off would have been huge, so I needed to shore up that position."

http://www.fantasyguru.com/football/sub ... -12vor.php
Wooden Icicles PPR (Horse Traders Dynasty)
Wood Chipper
Special Teams Staff
Special Teams Staff

User avatar
CafeholicCafe Writer
Posts: 330
Joined: 17 Jul 2012
Home Cafe: Football
Location: Knoxville, TN

Re: VOR (Value Over Replacement) Drafting

Postby Firemoth » Mon Aug 06, 2012 7:13 pm

Since you went through the trouble to explain, I'm obliged to be troubled to reply, albeit late. (Sorry, I was busy.)

I find the historic difference between the top players and best expected waiver wire players to be a good way to illustrate the relative value of players in the draft. This is especially useful to help get a good first handle on a new or novel scoring system, as it captures both position scarcity and increased value of the elite players. A (ranked) regression line could serve a similar purpose for selecting positions in the very early rounds.

The author you quoted started to suggest a value over replacement draft method, but the method he settled on in the end is very similar to Value-Based Drafting, which is a well-respected method which many have used to good effect. The main drawback is that in it's purest representation, it requires specific predictions (yards/TDs) for EVERY player likely to be in the running, which is too much of an effort for many. You could, of course, use someone else's predictions, but that rather takes the fun out of doing your rankings.

Beyond the high start up cost, it really doesn't have much of a drawback. A single predicted score does a poor job of representing risk/reward, such as players who are unsigned or in position battles. And it doesn't (by itself) deal well with common bye weeks. But these are minor or advanced considerations. Some purists will also say that it is rational that you overload on RBs beyond their relative points value because they have greater trade value during the season when RBs start getting injured on other teams. I do not subscribe to this theory and I have never seen it work in the groups I usually play with, but I suppose it could depend on your opponents/league.

I have seen some rather advanced implementations of VBD/VOR (which were automated/computerized) posted here, which seem quite good.

I personally generally go with the Tiering system, which accomplishes much the same purpose which allowing me to make my own rankings with a reasonable amount of effort.
Firemoth
Special Teams Staff
Special Teams Staff


Posts: 116
(Past Year: 5)
Joined: 25 Jul 2005
Home Cafe: Football


Return to Football Talk

Who is online

Users browsing this forum: No registered users and 6 guests

cron
Forums Articles & Tips Start & Sit Sleepers Rankings Leagues


Get Ready...
The 2014 NFL season kicks off in 16:02 hours
(and 41 days)
2014 NFL Schedule


  • Fantasy Football
  • Article Submissions
  • Privacy Statement
  • Site Survey 
  • Contact