My top 250 rankings aren't telling the whole story. As I've thought about it, there are at least three ways to construct pre-draft rankings:
(1) Predicting Fantasy Production
On this draft board, the predictor is saying the #1 player will score the most fantasy points, the #2 player the second most, the 250th player the 250th most, etc. In standard scoring leagues--which is the way my draft board's set up--this approach makes no sense. Last season QBs comprised the top 6 overall scorers. No one wants to see a draft board where the first six guys are QBs.
That said, if we lived in a world where one QB was likely to score 400 fantasy points and the next five QBs were likely to score 290-325--and if the #1 flex player is not projected to have that much distance between himself and the next best flex players--then hell yeah, I'm telling everyone to draft the QB first overall.
So fantasy production-based rankings are useful only in those cases where straight-up points take precedent over positional eliteness.
(2) Predicting Positional Tiers
This is how my rankings are set up, in line with the approach described in the book FF4W. My goal is to grab each positional tier leader. So if a middle-of-the-pack third tier RB is available on my pick, I'll pass on him for a high-end third tier WR. Pretty intuitive stuff. It keeps us from overreaching for certain guys when comparable players at more elite positional levels are still around.
This approach predicts fantasy production, but not in such a linear way. Higher upside often trumps higher downside. In other words, WR Leonard Hankerson is ranked better than WR Roddy White using this system, because even though White's the #2 WR in Atlanta and arguably has a higher floor (Hankerson is largely unproven as a highly targeted, 16-game receiver), Hankerson's ceiling propels him above his post-prime teammate.
(3) Drafting Order
I aspire to create a top 250 list using this approach. The idea is that you can plug this list directly into a draft queue. When it's your turn, the top 2-3 players available are the ones you choose from. So if you've drafted RB-RB and in the third round your top three guys are an RB, QB, and WR, there's some flexibility if--in observing draft behavior so far--you conclude that this is the time to grab the best QB.
Also, this queue approach allows us to anticipate when drafters are likely to take each guy; we incorporate that insight into our draft ordering. So although Marvin Jones is my 80th ranked player, I won't take him with the 80th pick because his overall ADP is 140. So I'd probably rank him around 110th in this system--low enough where I can grab an even higher-value guy with the 80th pick, and high enough where there's little risk someone else will draft him.
---
So let's use Roddy White as an example of how these three ranking systems would play out. In the first style (Predicting Fantasy Production), where low floors matter just as much as high ceilings, he'd probably be ranked around 42nd among WRs and 100th overall.
In the second system, he's currently ranked 51st among WRs and 123rd overall. His higher floor vs. Hankerson is meaningless here. This system isn't about floors; it's about ceilings. So he moves down my draft board in favor of guys who might be crap--or who might bust out as every week contributors.
In the third system--Drafting Order--I'd place White around 200th overall, or might not include him at all. The fact is, I have no use for White. Maybe I'm wrong about him, but I believe in essentially every round I can find someone with higher upside. So he wouldn't make it onto my draft board. His overall ADP is 71. Someone's drafting him in the sixth or seventh round. Why throw him into a draft queue when (a) I don't want him, and (b) I'm not gonna get him.
(1) Predicting Fantasy Production
On this draft board, the predictor is saying the #1 player will score the most fantasy points, the #2 player the second most, the 250th player the 250th most, etc. In standard scoring leagues--which is the way my draft board's set up--this approach makes no sense. Last season QBs comprised the top 6 overall scorers. No one wants to see a draft board where the first six guys are QBs.
That said, if we lived in a world where one QB was likely to score 400 fantasy points and the next five QBs were likely to score 290-325--and if the #1 flex player is not projected to have that much distance between himself and the next best flex players--then hell yeah, I'm telling everyone to draft the QB first overall.
So fantasy production-based rankings are useful only in those cases where straight-up points take precedent over positional eliteness.
(2) Predicting Positional Tiers
This is how my rankings are set up, in line with the approach described in the book FF4W. My goal is to grab each positional tier leader. So if a middle-of-the-pack third tier RB is available on my pick, I'll pass on him for a high-end third tier WR. Pretty intuitive stuff. It keeps us from overreaching for certain guys when comparable players at more elite positional levels are still around.
This approach predicts fantasy production, but not in such a linear way. Higher upside often trumps higher downside. In other words, WR Leonard Hankerson is ranked better than WR Roddy White using this system, because even though White's the #2 WR in Atlanta and arguably has a higher floor (Hankerson is largely unproven as a highly targeted, 16-game receiver), Hankerson's ceiling propels him above his post-prime teammate.
(3) Drafting Order
I aspire to create a top 250 list using this approach. The idea is that you can plug this list directly into a draft queue. When it's your turn, the top 2-3 players available are the ones you choose from. So if you've drafted RB-RB and in the third round your top three guys are an RB, QB, and WR, there's some flexibility if--in observing draft behavior so far--you conclude that this is the time to grab the best QB.
Also, this queue approach allows us to anticipate when drafters are likely to take each guy; we incorporate that insight into our draft ordering. So although Marvin Jones is my 80th ranked player, I won't take him with the 80th pick because his overall ADP is 140. So I'd probably rank him around 110th in this system--low enough where I can grab an even higher-value guy with the 80th pick, and high enough where there's little risk someone else will draft him.
---
So let's use Roddy White as an example of how these three ranking systems would play out. In the first style (Predicting Fantasy Production), where low floors matter just as much as high ceilings, he'd probably be ranked around 42nd among WRs and 100th overall.
In the second system, he's currently ranked 51st among WRs and 123rd overall. His higher floor vs. Hankerson is meaningless here. This system isn't about floors; it's about ceilings. So he moves down my draft board in favor of guys who might be crap--or who might bust out as every week contributors.
In the third system--Drafting Order--I'd place White around 200th overall, or might not include him at all. The fact is, I have no use for White. Maybe I'm wrong about him, but I believe in essentially every round I can find someone with higher upside. So he wouldn't make it onto my draft board. His overall ADP is 71. Someone's drafting him in the sixth or seventh round. Why throw him into a draft queue when (a) I don't want him, and (b) I'm not gonna get him.