I would describe Dynamic Value Based Drafting (DVBD) differently though you can factor in starters using it, yes. But I wouldn't generally think of that as the major gain from it.
Value Based Drafting
Value Based Drafting (VBD) as hopefully most people know by now, is a way of expressing numerically the value of a player regardless of position, making for easy and quick comparisons in the heat of a draft. If you expect a RB and QB will both score 250 total points on the year that doesn't make them equal value. If average QBs are scoring 240 and average RBs are scoring 200, the RB is an above average player while the QB is roughly an average player, making the RB more valuable. We can even use those numbers to say the RB adds 40 more points of value to your team than does the QB over an average team. So it is a way of taking fantasy points and getting at a number that can be directly compared regardless of positions. (In math this kind of result is referred to as having been "normalized").
VBD is an expression of your belief of each player's value based solely on the pool of players available this year.
Many people then take those VBD values (sometimes called X-values) and generate a ranking from them and draft from the list. It's better than just wild ### guessing at a ranking... but it doesn't incorporate several things worth including. Which leads us to...
So what is Dynamic Value Based Drafting?
Let's say it is your 2nd round pick and you plan to take a RB and WR with your next 2 picks. You are considering your RB10 or WR5 at your 2nd round pick. Our gypsy fortune teller tells you that no WRs will be taken before your 3rd round pick, but 8 more RBs will be taken. VBD might say that WR5 is more valuable than RB10. But clearly taking WR5 now and then RB18 in the third is a worse team than taking RB10 and still getting the exact same WR5 in the third round.
Does this mean VBD is "wrong"? No, it means what I said in the italicized above. VBD only shows you value based on the pool of players. This example shows that for your team there is also a component of value based on the order in which players are selected in any given draft.
This is what DVBD tries to deal with. The basic gist is you look at the drop off in scoring at each position between your pick now and your next pick, which is easily done by just subtracting their fantasy points or their VBD numbers. And you generally want to pick the position of need that will drop off the most. If the difference between the QB available now and at your next pick is 60 points, for RB it is 50, for WR it is 40 and for TE it is 10, then generally the QB tends to be the better pick. Because the combination of the QB and a different position next round will add more value than any other combination.
To get back to your comment about DVBD is about focusing on starters... in Draft Dominator you can have it include a factor that basically says, "If I already have all my starters at this position then discount the drop off so it isn't saying taking a backup QB would be more valuable than taking a starting player." So yes it can include a measure of focusing on your starters, but you can do that in your head pretty well with most any draft strategy, and I wouldn't say that is the heart of what DVBD is.
DVBD is very simple conceptually. But it is not the end-all be-all either, anymore than VBD was. Consider the example I just gave and let me flesh it out more. So 2nd round you can take QB3 or 3rd round you can take QB5 who drops off 60 points, more than any other position. If that's all you consider, you would say "I should take QB3 now".
But how the rest of the draft unfolds can change whether that is the right answer. Let's say your QB5 is Peyton Manning and you don't believe anyone in your draft will touch him until the 10th round, while you think he has 3rd round value. If you pass on QB3 and instead take QB5 in the 9th round, the net gain at all the other positions that now got selected earlier is probably more than that 60 point drop off. So because he won't be taken until later rounds is what drove the right decision. One way you could look at it is we should really consider the drop offs for every position at every round we're taking starters. Unfortunately this is a very difficult math problem to solve given the number of variables involved, short of brute forcing it. And even that requires predictions of the order the players will be taken.
So what are you saying, neither is "right"?
Neither is designed to give a perfect draft strategy. What they both do well is express the component of value they deal with. If you assume either is the completely perfect answer that takes everything into account, then you're misusing it. Fortunately for many people, especially people new to fantasy football, even misusing VBD or DVBD likely leads to a better result than using total points or blindly adhering to stud RB. (Something I actually confirmed years back with a series of mock drafts where the only change was the strategy for 1 team.)
VBD is great for giving you an ordered list of your view of player value that you can refer to throughout your draft. That is invaluable. Dynamic VBD can show you in seconds the drop offs that are likely to come in the next few rounds. Even if you do it in your head, but particularly if you are using a good tool like Draft Dominator where all you have to do is tell it what picks you want to see the drop off between and it'll handle the calculations.
So how should you draft if neither of those is perfect?
But if you want the best, most robust draft method, you need to spend some time preparing so you have on hand other information that will affect the outcome of your draft. The biggest piece missing from our picture so far is a set of beliefs about what players will be undervalued and overvalued in your draft and by how much.
Essentially I'm saying ideally we would do DVBD for every round and every position... but that math problem is difficult to solve and our uncertainty in predicting the order players are taken make trying to get such exact results questionable. So instead let's just highlight key under and over valued players and work them into our draft process more organically.
A simple way to get at this is to take Average Draft Position from a site like antsports or MFL, use their overall ranking and then for each player subtract the ranking that VBD's cheatsheet says he should have. If a player is going as the 20th player and your VBD sheet has him as 28th, then he's not likely to be a good value pick. But if he's going as the 30th player and you have him as the 20th player, that's a good value player. If you can adjust ADP rankings to better fit your league's draft history all the better.
Once you have identified good value players, you can look at them collectively and try to see how to mesh a number of them together into a draft strategy that will give you the best overall team. Not every pick will be a high value player, but several should be. The most important thing here is don't just look at the names and wonder. Actually write down a team based on players likely to be available at each of your picks. Sum up the projected fantasy points of the starters, and don't forget to look at the depth on the team. That sounds common sense yet I think many people only try to visualize in their head which team they'd like more when they could add up the actual fantasy points they expect them to produce in a few seconds to include in their decision.
Then try it again with another combination. Try it again to test likely scenarios for a given strategy. Maybe your sleeper RB won't slide to the 3rd round like you hope. So do a version where you just miss out on him so you can see how it impacts your draft. If you don't like the results at all, now you have a feeling for the risk of assuming he'll be there, and can weigh that entire draft strategy versus others accordingly. And if you do go with that strategy, since you'll have been through a mock of that situation, should your RB not slide to the 3rd, you'll know exactly the impact on every other position as now you spend more mid round picks on RB depth to make up for it. You'll have already made all these decisions, at a moment when rather than 90 seconds on the clock you could take as much time as needed to weigh your options. And throwing in another plug, Draft Dominator can handle mocking for the other teams in your league as you do these multiple scenarios, and will then show you your expected points.
David Dodds I imagine goes through a similar thought process in coming out with his Perfect Draft series of articles, giving you the end results of what he thought the best strategy was. Those can also be a good guide.
So, summary... VBD and DVBD are both great drafting tools. Neither is an absolute best strategy by itself, or even with the two taken together.
