The easiest way to explain VBD that I have heard is the price required to replace a player on your team. Say you draft Manning first and he is out for the year. How easy is it to replace his stats vs a runningback like LT2? It is much easier to replace a WR/QB that goes down for the year then a runningback. That is why you see them go so much higher in the draft even though the VBD numbers start to get jumbled up. Now Manning is impossible to replace if you assume he will repeat last years stats. That is why you see some people take him first. If you think last year was a fluke you would never think of it.
It's the same thing as VBD, and this explaination did a nice job of explaining for me as well as saying exactly how to calculate them yourself in excel.
Keep in mind that doing it for football is much simpler than for baseball- just turn your projections into total points for the season, and use that as your only category (unlike roto baseball, where you have 5-10 categories to deal with).
Here's the thing: after you've calculated all player X values, i assumed you simply rank players from top to bottom by their X value in order to generate your draft list. That seems like the logical way to proceed, then tweak your list based on ADP to maximize value at each pick.
But at least one fantasy mag i read through goes to the trouble of explaining VBD and calculating player X values for each position, but then deviates from ordering based on X values in the overall draft list. The players are in order by position, but across positions they are not in order of their X values, which is the whole point of VBD, i thought, to compare players at different positions. So it seems like there's something else going on, some step that i'm missing...
Not sure if anyone has anything else to say on this subject, but i think it's an interesting puzzle, given the fact that VBD seems to present a coherent draft strategy. I mean, why the hell does this mag (Fantasy Football Pro Forecast, i think) endorse VBD, calculate and publish X values for each player based on their projections, and then not order the players according to those X values? it bothers me...
probably because they themselves don't endorse VBD but want to give readers that option.
VBD is not the end all be all of drafting. If you just take the player with the highest value all of the time you will end up with a team composed of high value, but a poorer starting lineup. You may have tons of great rbs and wrs, but your qb may be dilfer and yet your backs that sit on the bench make up for this in value. The point is, you can't ignore starting requirements.
To maximize your draft use VBD for at least 4 or 5 rounds regardless of position. However, also use Average draft position information (adp) to figure out what other owners might do/think about your pick. If in your league Tony G is the top ranked player by Value, and yet he goes much later every year, don't draft him in the first. That was an obvious example, but typically when its my pick I look at players similar in value (according to their X values) and then asses who would go the earliest. This way I have a chance of getting two of the guys out of the group that are similar in value (one this round and one the next). If you don't do this, you may take the highest value player too early when he would have slipped another round. Its a fine line to walk when deciding whether to wait on a higher value player, but it can pay off so long as you don't sacrifice too much.
ADMIN edit: please note that the overall sig limit is 12k - thanks! (Thanks Leber) AIM is like multiplayer notepad
hardboiled wrote:The results were surprising--among some of the early discrepancies, my VBD list showed Daunte as a late first round pick. I also had Randy Moss and the other top receivers going earlier in the first and second rounds than usual. Basically, my VBD list placed a higher premium on the top players of other positions compared to the second tier of RBs.
Good post Hardboiled.
The discrepancies you talk about are largly due to projections. Sincwe VBD is based on projections, they are only as accurate as those projections. I personally dont rank Cpepp in the 1st rd because I have him falling back to his career averages at best in my projections, while on the other hand, Moss is in my top 5 since I think he'll put up close to 20 TDs this season. Its all about your projections. Understanding VBD is the first step in fundamental drafting, and it looks like you are well on your way.
As for the Mag - I wouldnt worry about it too much - most of the time those things are written and produced by people looking to gain notoriety more than accuracy.
The first thing in the article is to make stat predictions for all the players who will be drafted. How do you guys who use VBD come up with the predictions? Is the Footballguys service worth signing up for?
ipljeff wrote:The first thing in the article is to make stat predictions for all the players who will be drafted. How do you guys who use VBD come up with the predictions? Is the Footballguys service worth signing up for?
I personally prefer to make the predictions myself. I just take a look at how they performed the past year and years before that, and also look at how their situation has changed (i.e. Coaching changes, Offensive line improvements, personel changes....etc). And just predict how well I think they will do in their situation for the next season. I never draft strictly using VBD so I don't think it's real important to go out and sign up for things, I would just make the predictions yourself.
I use Avt, Since I am nice I will quote a nice definition:
Fantasy Football is a game of statistics. Unlike real football, fantasy owners are not rewarded if their players give a lot of autographs, have great chemistry on and off the field, do community work, or start a charity. Conversely, fantasy owners are not penalized if their player doesn’t have any of these attributes. In fantasy land, the only thing that counts is statistics. Therefore, your first priority in preparing for the draft is to figure out what kind of statistics will result from this year’s crop of players. These statistics are called projections.
Obviously, owners with more accurate projections than their opposition can reap huge rewards. Unfortunately, most people will tell you that creating projections is a difficult and time-consuming task. Consequently, they either take a half-hearted approach or let others do it for them. Fortunately, I have a method that takes less than a day to do, and the results are better than the hard-working experts who spend weeks and months doing them.
I use a little-known concept called the Average Value Theory (AVT). AVT is a powerful, proven, and objective method for doing projections. Best of all, it is quick and easy. Here’s how I do it.
Step 1) Find a source for NFL statistics. You’ll need to look up stats such as yardage, TDs, receptions, and so on. Basically, you need access to any stat that would be used in your fantasy league’s scoring system.
Step 2) Convert each player’s statistics into fantasy points. The conversions will be based on your fantasy league’s scoring system. You’ll need to calculate the fantasy points scored by every player at each position for the past three years . Unfortunately, this entails a lot of calculations; too many to do by hand or by calculator. Fortunately, there are many online programs available that do all the work for you.
Step 3) Rank the players in each year according to the fantasy points scored. You will do each position separately. If you are using an online program, this part is automatically done for you.
Step 4) Remove the names of the players. That’s right. The names of the players are unimportant in AVT. Instead the focus is on the numerical rankings and the fantasy points scored by those rankings. The ranking of actual NFL players and names are not part of the AVT method and will be touched upon in the next chapter.
Step 5) Determine how many AVs to calculate for each position. “AV” is the average value, or average fantasy points scored, calculated for a particular position and rank. In other words, the AV is the actual projection. You’ll need AVs for each player you have on your cheatsheet . For example, if you are going to have 30 QBs listed on your cheatsheet then you need 30 AVs. I recommend the following system in determining the number of AVs needed: Multiply the number of maximum starters you can have in a position by the number of teams in the league by two. Therefore, if you have one starting quarterback in a 12-team league, then you will need 1x12x2=24 QB projections to fill your cheatsheet. Here’s another example: If you are in a 10-team league that requires two RBs, but a third back can be started as a “flex” player, you would need 3x10x2=60 running backs.
Step 6) Calculate the AVs for each rank and position, doing no more than the required number of AVs determined from Step 5. To calculate an AV, add up the fantasy points scored by a particular rank and position over the last three years and divide by three. For example, the No. 1 QB may have scored a combined 333 fantasy points over the past three years. In this case, the 3-year AVT projection for the No. 1 QB is 111 fantasy points. This means that the AVT projects that the QB will score 111 fantasy points this coming year. When you are finished with this step you should have a ranking and projection list for each position. If I ask you what the projection for this year’s 12th ranked WR is, you should be able to tell me from your list.
Congratulations. That’s it! These AVs are your player projections for the upcoming year.
I use this to project player stats, and create my X-values.... Etc