Extreme Positional Scoring Disparities (1 Viewer)

[oso diablo]

I just got hooked into a new league with my neighbor. It has odd scoring rules, where QBs end up tallying 2-3x as many points as RBs or WRs. For example, Peyton Manning is projected (in DD) to score well over 600 points, but Chris Johnson is in the mid-200s.

I have been using VBD principles (and the Draft Dominator) for years, but always in leagues where the different positions were fairly close to each other, scoring-wise. So, the main question here is...

Can the principles of VBD be stretched too far in a league like this, or have you found the concept to be solid no matter how extreme the positional differences are?

A related question: is it proper to calculate VBD as a mere subtraction problem that way FBG does it, or would it make more sense to use some type of division formula? That is, player X is either 100 points above the baseline, or he is 50% above the baseline. Two different approaches that might yield different results when comparing across positions with extreme disparities.

What do you guys think?

 

[Jeff Pasquino]

I just got hooked into a new league with my neighbor. It has odd scoring rules, where QBs end up tallying 2-3x as many points as RBs or WRs. For example, Peyton Manning is projected (in DD) to score well over 600 points, but Chris Johnson is in the mid-200s.

I have been using VBD principles (and the Draft Dominator) for years, but always in leagues where the different positions were fairly close to each other, scoring-wise. So, the main question here is...

Can the principles of VBD be stretched too far in a league like this, or have you found the concept to be solid no matter how extreme the positional differences are?

A related question: is it proper to calculate VBD as a mere subtraction problem that way FBG does it, or would it make more sense to use some type of division formula? That is, player X is either 100 points above the baseline, or he is 50% above the baseline. Two different approaches that might yield different results when comparing across positions with extreme disparities.

What do you guys think?

Using a percentage would be skewed. For example, would you take a kicker projected for 120 points over a QB projected for 600 if the PK baseline was 80 and QB was 400? That PK is 50% better than the baseline (but only 40 pts). The QB is only 33% better but 200 points. Which would you rather have in your lineup?Leagues with odd rules (I've played in some - like 2 points per pass attempt and 2 more for a completion - crazy stuff). Regardless of the rules, the VBD math always works.

 

[theToes]

I just got hooked into a new league with my neighbor. It has odd scoring rules, where QBs end up tallying 2-3x as many points as RBs or WRs. For example, Peyton Manning is projected (in DD) to score well over 600 points, but Chris Johnson is in the mid-200s.

I have been using VBD principles (and the Draft Dominator) for years, but always in leagues where the different positions were fairly close to each other, scoring-wise. So, the main question here is...

Can the principles of VBD be stretched too far in a league like this, or have you found the concept to be solid no matter how extreme the positional differences are?

A related question: is it proper to calculate VBD as a mere subtraction problem that way FBG does it, or would it make more sense to use some type of division formula? That is, player X is either 100 points above the baseline, or he is 50% above the baseline. Two different approaches that might yield different results when comparing across positions with extreme disparities.

What do you guys think?

Using a percentage would be skewed. For example, would you take a kicker projected for 120 points over a QB projected for 600 if the PK baseline was 80 and QB was 400? That PK is 50% better than the baseline (but only 40 pts). The QB is only 33% better but 200 points. Which would you rather have in your lineup?Leagues with odd rules (I've played in some - like 2 points per pass attempt and 2 more for a completion - crazy stuff). Regardless of the rules, the VBD math always works.

For my league's scoring system VBD has 3 qb's in the top 7 rankings. Should VBD be followed here?

 

[Jeff Pasquino]

I just got hooked into a new league with my neighbor. It has odd scoring rules, where QBs end up tallying 2-3x as many points as RBs or WRs. For example, Peyton Manning is projected (in DD) to score well over 600 points, but Chris Johnson is in the mid-200s.

I have been using VBD principles (and the Draft Dominator) for years, but always in leagues where the different positions were fairly close to each other, scoring-wise. So, the main question here is...

Can the principles of VBD be stretched too far in a league like this, or have you found the concept to be solid no matter how extreme the positional differences are?

A related question: is it proper to calculate VBD as a mere subtraction problem that way FBG does it, or would it make more sense to use some type of division formula? That is, player X is either 100 points above the baseline, or he is 50% above the baseline. Two different approaches that might yield different results when comparing across positions with extreme disparities.

What do you guys think?

Using a percentage would be skewed. For example, would you take a kicker projected for 120 points over a QB projected for 600 if the PK baseline was 80 and QB was 400? That PK is 50% better than the baseline (but only 40 pts). The QB is only 33% better but 200 points. Which would you rather have in your lineup?Leagues with odd rules (I've played in some - like 2 points per pass attempt and 2 more for a completion - crazy stuff). Regardless of the rules, the VBD math always works.

For my league's scoring system VBD has 3 qb's in the top 7 rankings. Should VBD be followed here?

Take one first then decide. I wouldn't say Take 2 if you can only start one. That's why the Draft Dominator cuts the VBD number down once you have one.I'd grab a QB2 earlier than some other drafts, however. You don't want to not have a viable starter.

 

[SSOG]

VBD is a sound concept, but in practice it's only as good as the projections you use to calculate it (and the baselines you set). I believe that the "Joe's Secret Formula" VBD baseline just takes the last player at each position projected to go in the first 8 rounds (or first 10 rounds, I forget how many rounds exactly), and sets that player as the "baseline". If 6 QBs are projected to go in the first 8 rounds (or however many rounds), then QB6 becomes the baseline. Obviously that method of determining the baseline assumes that leagues adhere roughly to ADP. In leagues that get absolutely QB-crazy, for instance, that particular baseline isn't going to do you much good.

Also, a projection is just a static number and doesn't give any indication as to how likely a player is to perform somewhere around that projection. For instance, I'd say the odds of Antonio Gates performing in the neighborhood of his projection are very, very high because Gates is a known quantity and because TE production tends to be very stable and predictable. The odds of the NYJets defense performing in the neighborhood of their projection are very, very low because defenses are the most unpredictable fantasy "player" from year to year (in large part because they're the "player" most impacted by strength of schedule). As a result, VBD tends to overrate the New York Jets because it spits out a (very, very high) raw VBD number while not adding the caveat that the Jets defense is extremely unlikely to actually finish right around that number. At the same time, it tends to underrate players like elite TEs because it doesn't give a bonus based on the fact that they're a near-lock to live up to their projection (for instance, Gates has finished in the top 40 in season-ending VBD for 6 straight years, and has finished in the top 20 in 5 of those 6 years, despite routinely having an ADP in the 40s or higher).

In my mind, in order to justify a high draft pick, a player needs three elements working in his favor.

Element #1- He must be valuable. This one goes without saying. There's no sense on drafting a player in the 1st round if he's not going to outproduce his peers. VBD is the perfect tool for capturing player value.

Element #2- He must be predictable. This is why the Jets should never be drafted in the 1st, even if your league has the most defensive-heavy scoring you've ever seen. There's simply a much lower probability that they live up to their preseason expectations than there is with an Aaron Rodgers, Maurice Jones-Drew, or Andre Johnson. Don't spend a premium pick on a crapshoot position.

Element #3- He must be difficult to replace. This one requires knowledge of your league. If Peyton Manning has the top VBD value in your league... but you know your leaguemates are RB-RB proponents and that Tony Romo will likely still be available in the 6th, then you shouldn't draft Peyton Manning in the first because you can get nearly identical production at a fraction of the cost. In the Draft Dominator, the tool tries to estimate this for you by anticipating the dropoff at each position and making its recommendations in the "Best Value Pick" window... but at the end of the day, there's no substitute for actual experience, here.

Now, if a player can meet all three criteria, then you should absolutely spend a high draft pick on him. If Peyton Manning is projected to have a huge VBD, and you anticipate your league going extremely QB-heavy early (resulting in a steep dropoff in QB quality), then take Peyton Manning in the first without hesitating. If Peyton has a strong VBD advantage over CJ3 and you expect your leaguemates to draft 8 QBs in the first two rounds, then go ahead and draft him #1 overall. If you feel he doesn't meet criteria #1 or criteria #3, then stick with a stud RB at the top of the round.

 

[oso diablo]

Leagues with odd rules (I've played in some - like 2 points per pass attempt and 2 more for a completion - crazy stuff). Regardless of the rules, the VBD math always works.

probably more psychological than anything. i get the concept, and believe it should work, and that it does work. Didn't ever think twice about passing on, say, a 250-point player to take a 200 point player. But there's something just a bit unnerving about passing on a 600 point player to take a 200 point player, even when the math works out that way.p.s. makes it even tougher to join a league like that and not know its history, draft tendencies.

 

[cobalt_27]

VBD is a sound concept, but in practice it's only as good as the projections you use to calculate it (and the baselines you set). I believe that the "Joe's Secret Formula" VBD baseline just takes the last player at each position projected to go in the first 8 rounds (or first 10 rounds, I forget how many rounds exactly), and sets that player as the "baseline". If 6 QBs are projected to go in the first 8 rounds (or however many rounds), then QB6 becomes the baseline. Obviously that method of determining the baseline assumes that leagues adhere roughly to ADP. In leagues that get absolutely QB-crazy, for instance, that particular baseline isn't going to do you much good.Also, a projection is just a static number and doesn't give any indication as to how likely a player is to perform somewhere around that projection. For instance, I'd say the odds of Antonio Gates performing in the neighborhood of his projection are very, very high because Gates is a known quantity and because TE production tends to be very stable and predictable. The odds of the NYJets defense performing in the neighborhood of their projection are very, very low because defenses are the most unpredictable fantasy "player" from year to year (in large part because they're the "player" most impacted by strength of schedule). As a result, VBD tends to overrate the New York Jets because it spits out a (very, very high) raw VBD number while not adding the caveat that the Jets defense is extremely unlikely to actually finish right around that number. At the same time, it tends to underrate players like elite TEs because it doesn't give a bonus based on the fact that they're a near-lock to live up to their projection (for instance, Gates has finished in the top 40 in season-ending VBD for 6 straight years, and has finished in the top 20 in 5 of those 6 years, despite routinely having an ADP in the 40s or higher).In my mind, in order to justify a high draft pick, a player needs three elements working in his favor.Element #1- He must be valuable. This one goes without saying. There's no sense on drafting a player in the 1st round if he's not going to outproduce his peers. VBD is the perfect tool for capturing player value.Element #2- He must be predictable. This is why the Jets should never be drafted in the 1st, even if your league has the most defensive-heavy scoring you've ever seen. There's simply a much lower probability that they live up to their preseason expectations than there is with an Aaron Rodgers, Maurice Jones-Drew, or Andre Johnson. Don't spend a premium pick on a crapshoot position.Element #3- He must be difficult to replace. This one requires knowledge of your league. If Peyton Manning has the top VBD value in your league... but you know your leaguemates are RB-RB proponents and that Tony Romo will likely still be available in the 6th, then you shouldn't draft Peyton Manning in the first because you can get nearly identical production at a fraction of the cost. In the Draft Dominator, the tool tries to estimate this for you by anticipating the dropoff at each position and making its recommendations in the "Best Value Pick" window... but at the end of the day, there's no substitute for actual experience, here.Now, if a player can meet all three criteria, then you should absolutely spend a high draft pick on him. If Peyton Manning is projected to have a huge VBD, and you anticipate your league going extremely QB-heavy early (resulting in a steep dropoff in QB quality), then take Peyton Manning in the first without hesitating. If Peyton has a strong VBD advantage over CJ3 and you expect your leaguemates to draft 8 QBs in the first two rounds, then go ahead and draft him #1 overall. If you feel he doesn't meet criteria #1 or criteria #3, then stick with a stud RB at the top of the round.

This is, quite possibly, the best posting ever made on FBG.

 

[GregR_2]

...

For my league's scoring system VBD has 3 qb's in the top 7 rankings. Should VBD be followed here?

VBD is a tool that tells you how valuable a player at one position is compared to a player at another position. That is, it tells you QB3 is more valuable than RB4 (or whoever your 6th rated player is).

It doesn't tell you where the best place is to draft a player is unless everyone else in the league is going to draft off the same sheet as you are. It doesn't tell you if, for example, RB4 taken now and QB5 taken in round 2 will outscore as a group QB3 in the 1st and RB15 in the second. Even though out of those 4 players, QB3 is by himself the most valuable... the fact that everyone else is drafting RBs causes your net team to be worse if you wait to take the RB. Put another way, VBD shows you value in the supply of players. But it doesn't deal at all with the value created by the demand of other owners and the way they draft.

So you need to combine things that show you both of those to actually figure out the best draft strategy. When you combine VBD with Average Draft Position or your own beliefs about where players will be drafted, it's a great way to identify players who are likely to go later than they should and thus represent great values. If you see that QB3 is going in the early third round when you found he's the 7th most valuable player in the league, that is a good indication you should test out some sample teams you could put together taking him there, and see if he's part of the combination of players that will give you the best team. Identify a number of these high value players and figure out how to maneuver your draft so that you hit as many of these high value players as possible.

A novice would probably do better drafting straight off a VBD ranking than he would drafting on his own. But drafting straight off a VBD sheet isn't maximizing your draft completely.

And it's worth noting that Draft Dominator's suggestions do take the value from both supply and demand into account... though it is normally only looking between now and your next pick. You should really investigate your possibilities before the draft and know the implications of any decision. "If I take my QB in the 1st then I feel I need to take a 3rd RB by the 6th to have sufficient depth, and that is going to push back getting my WR3 to where he drops off this many points." Those are the kind of trade offs you want to be aware of before your draft ever starts, and that requires some work sitting down and going through possibilities so you learn what the later impact will be of a decision you make now. VBD, nor any tool, is going to be able to do that for you. But they can help point you to the areas you should investigate.

 

[HALtheCOW]

VBD is a sound concept, but in practice it's only as good as the projections you use to calculate it (and the baselines you set). I believe that the "Joe's Secret Formula" VBD baseline just takes the last player at each position projected to go in the first 8 rounds (or first 10 rounds, I forget how many rounds exactly), and sets that player as the "baseline". If 6 QBs are projected to go in the first 8 rounds (or however many rounds), then QB6 becomes the baseline. Obviously that method of determining the baseline assumes that leagues adhere roughly to ADP. In leagues that get absolutely QB-crazy, for instance, that particular baseline isn't going to do you much good.Also, a projection is just a static number and doesn't give any indication as to how likely a player is to perform somewhere around that projection. For instance, I'd say the odds of Antonio Gates performing in the neighborhood of his projection are very, very high because Gates is a known quantity and because TE production tends to be very stable and predictable. The odds of the NYJets defense performing in the neighborhood of their projection are very, very low because defenses are the most unpredictable fantasy "player" from year to year (in large part because they're the "player" most impacted by strength of schedule). As a result, VBD tends to overrate the New York Jets because it spits out a (very, very high) raw VBD number while not adding the caveat that the Jets defense is extremely unlikely to actually finish right around that number. At the same time, it tends to underrate players like elite TEs because it doesn't give a bonus based on the fact that they're a near-lock to live up to their projection (for instance, Gates has finished in the top 40 in season-ending VBD for 6 straight years, and has finished in the top 20 in 5 of those 6 years, despite routinely having an ADP in the 40s or higher).In my mind, in order to justify a high draft pick, a player needs three elements working in his favor.Element #1- He must be valuable. This one goes without saying. There's no sense on drafting a player in the 1st round if he's not going to outproduce his peers. VBD is the perfect tool for capturing player value.Element #2- He must be predictable. This is why the Jets should never be drafted in the 1st, even if your league has the most defensive-heavy scoring you've ever seen. There's simply a much lower probability that they live up to their preseason expectations than there is with an Aaron Rodgers, Maurice Jones-Drew, or Andre Johnson. Don't spend a premium pick on a crapshoot position.Element #3- He must be difficult to replace. This one requires knowledge of your league. If Peyton Manning has the top VBD value in your league... but you know your leaguemates are RB-RB proponents and that Tony Romo will likely still be available in the 6th, then you shouldn't draft Peyton Manning in the first because you can get nearly identical production at a fraction of the cost. In the Draft Dominator, the tool tries to estimate this for you by anticipating the dropoff at each position and making its recommendations in the "Best Value Pick" window... but at the end of the day, there's no substitute for actual experience, here.Now, if a player can meet all three criteria, then you should absolutely spend a high draft pick on him. If Peyton Manning is projected to have a huge VBD, and you anticipate your league going extremely QB-heavy early (resulting in a steep dropoff in QB quality), then take Peyton Manning in the first without hesitating. If Peyton has a strong VBD advantage over CJ3 and you expect your leaguemates to draft 8 QBs in the first two rounds, then go ahead and draft him #1 overall. If you feel he doesn't meet criteria #1 or criteria #3, then stick with a stud RB at the top of the round.

This is, quite possibly, the best posting ever made on FBG.