What's new
Fantasy Football - Footballguys Forums

This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

Draft Dominator (1 Viewer)

Liquid Tension

Liquid Tension

Footballguy
This has been a perplexing issue with me and I am looking for some guidance. I wouldn't care much, but the difference between Joe's secret formula and the worst starter method is enormous and because of this it makes me gun shy. For example, with the worst starter method, TE Witten moved up from 36 (Joe's secret formula) to 10!

Can someone shed some light on this because the difference is HUGE.

Thanks

 
Last edited by a moderator:
Liquid Tension said:
This has been a perplexing issue with me and I am looking for some guidance. I wouldn't care much, but the difference between Joe's secret formula and the worst starter method is enormous and because of this it makes me gun shy. For example, with the worst starter method, TE Witten moved up from 36 (Joe's secret formula) to 10!Can someone shed some light on this because the difference is HUGE.Thanks
Click to expand...
I can't imagine that I am the only one who has an issue with this as it almost makes you question the validity?One separate question on the VBD is that kickers tend to appear to be overvalued in the VBD application. Any thoughts folks?
 
I'll be truthful, I don't really understand the draft dominator stuff. So, I use the VBD with my leagues scoring, etc. inputed in it. I understand it better and is easy for me to follow, if I wish, during the draft.

 
drm1125 said:
I'll be truthful, I don't really understand the draft dominator stuff. So, I use the VBD with my leagues scoring, etc. inputed in it. I understand it better and is easy for me to follow, if I wish, during the draft.
Click to expand...
Joe's formula is also a factor in the VBD app. I have never really looked into the difference, but maybe it is something I want to do now.
 
dino259 said:
drm1125 said:
I'll be truthful, I don't really understand the draft dominator stuff. So, I use the VBD with my leagues scoring, etc. inputed in it. I understand it better and is easy for me to follow, if I wish, during the draft.
Click to expand...
Joe's formula is also a factor in the VBD app.
Click to expand...
I always fiddle with it in the VBD app for my auctions to weigh things differently. I have a mac so I have little experience with DD which bothers me, wish it worked on my platform too.
 
I've been playing around with Joe's Secret Forumla vs. my league's projected draft baselines for a few days on the DD. I'm in a 10-team, 10-player keeper league with a 10-round draft (we keep 10 from last year and draft 10 more).

Using my league's projected draft baselines -- 32 QBs, 58 RBs, 54 WRs, 15 Ks and 20 defenses -- the DD says that 4 QBs should go 1-4 at the draft.

Using Joe's formula, the DD says that the top picks should be RB-WR-WR-RB.

Given my league's draft history, there's usually 5 RBs, 4 WRs and 1 QB drafted in the opening round, so using Joe's formula seems to give me a more realistic feel to the draft projection.

The DD using baseline players seems to push the 4 QBs because they are projected to score more points than the remaing RBs and WRs. While that's probably true, since everyone in my league already keeps at least one QB, owners tend to wait on drafting their backup QBs and those guys can be drafted in the second or third rounds.

I think using baselines in DD gives you a cold hard look at the numbers, but Joe's formula gives you those numbers with positional scarcity/need factored in as well.

 
Personally, I don't like Joe's formula for most situations. His formula is based on an approximation of the number of players who will be taken in the first five rounds. There are two major problems with that: one, it doesn't account for league differences, and two, it makes the assumption that the way people tend to draft is optimal. What if RBs are overvalued in typical drafts? By using Joe's Secret Formula, which sets the baseline very low for RBs, they will continue to be overvalued.

Worst starter is probably not the best in most leagues, either; I do use worst starter in Anarchy, because it's all-play, but in a league with a bench you probably need to be smarter about what you define as the baseline.

I try to think about baselines as "the level below which I consider players more or less equivalent." In my auction league, I try to peg baselines to players I wouldn't pay more than $1 for, which results in baselines lower than worst starter. Or you could look at it as "the level below which the starters pretty much suck", which might be something like QB10, RB20, WR30, TE6. Or you could look at your league's historical point totals and see if there's a point total where scoring tends to level off, and set the baseline to that.

Philosophically, you could look at setting the baseline at replacement value; if you could pick up an equivalent player for little or no cost, you've hit the baseline.

 
Last edited by a moderator:
Lets compare the results, just opening up a new league and using the default settings, (12 team, 1 QB, 2 RB, 3 WR, 1 TE) and see what the drop offs are.

With Joe's method, the baseline player is:

QB11

RB29

WR31

TE6

With worst starter the baseline of course is:

QB12

RB24

WR36

TE12

So relating the two with a modifier that turns last starter into Joe's you get:

.92 * last starter

1.2* last starter

.86 * last starter

.5 * last starter

At this point this is where you need to ask yourself if that is how you think it should be. I think I can probably state some reasoning behind why it might be that way. Note I'm not critiquing it or saying I agree, just giving some reasons I think would have gone into it.

For example, at QB while the elite few are often predictable, the bottom half of the starters often do not end up being who we think. That uncertainty could lead to valuing them less than last starter as you can get as good of QBs later and outside of your projected starters.

At RB, scarcity and injuries is the driving factor which increases their value over what it would have been.

WR is another position where it's hard to predict outside of the top few, who will end up there. Most of the top 60 WRs have a very legitimate shot at ending up in the 20-30 range.

TE is a position where the top 3 are normally pretty solid, the top 6 you have a fair shot at having gotten right or close to right, but after that there isn't a lot of separation, so not as high as value because you can get ones late in the draft likely to perform similar to the ones taken earlier.

Ok, so where does all of that get us? Really it just gives an insight into what Joe thinks about the value of the positions. If you don't agree with it, make your own baselines, or use last starter. But do realize that last starter may be a baseline we commonly use in discussion, but that it definitely does not reflect all of the significant components of value. Joe's formula tries to take into account the other general trends in the player pool and in scarcity and demand.

 
Liquid Tension said:
This has been a perplexing issue with me and I am looking for some guidance. I wouldn't care much, but the difference between Joe's secret formula and the worst starter method is enormous and because of this it makes me gun shy. For example, with the worst starter method, TE Witten moved up from 36 (Joe's secret formula) to 10!Can someone shed some light on this because the difference is HUGE.Thanks
Click to expand...
In regards to Witten, see my post before this one, where TE has the biggest change in Joe's formula. You're basically getting half as many TE starters with Joe's formula. If you have a league with PPR scoring that favors TE and/or start a lot of TEs, this could be a lot bigger change.Joe's formula is probably more appropriate for a standard scoring and standard setup for TE. If I were doing it I imagine I'd probably have used something similar if I had to set a single number. However in my own leagues we don't fit that profile (2 TEs plus a WR/TE flex, and 1 PPR for TE while WR only get .5 and RB .25), and the differences between the TEs going down the position chart start to look more like the other positions.
 
I use median starter for the first 6 rounds or so then switch to worst starter.

I always look at ADP to see if I can wait a round or more and never pick someone who I think has a better than 50% chance of being there.

Part of the reason is I play in several "deep" leagues (2QB, 2RB, 3WR + 2 flex) and some are PPR. So Joe's formula does not well reflect the realities.

 
I play in a 10 team (1 player keeper ) league starts 1 qb, 2 rb, 2 wr, 1 te, 1 k and 1 Bonus (bonus can be any type player)

Scoring: 5 pts passing td 0.5 pts per 10 yards

6 pts rushing td 1.0 pts per 10 yards

I tend to use Joe's Secret formula, but then I go back and tweek the number of starters usually QB and RB to get a top 30 or so that is similar to how things looked in our final points average from the prior year or two.

Example if I think I need an average of 12-15 qb's in my top 30(our league setup puts a higher value on qbs than most leagues) I may have to change the numbers of starters at qb to some number a little more or less than 2.0 say 2.3 or 1.8 what ever I think it needs to be. Same thing with running backs.

Idon't know if it is the right way to go about it, and I second guess myself alot but it seems to work for me.

 
CalBear, GregR and others...thanks for your opinions; it really is appreciated and I am sure not just by me.

I need to play around with it to understand this better as I need to get my mind around what you are saying.

I have different leagues and while I don't use it as the bible, once I alter the projections, adjust the scoring and setup the application per the league requirements (teams rounds etc...) my expectation is that this application would give me the values of the players taking it all into context?

It appears this is not correct as the starters change the context more than I expected (I assumed it was already in there).

Can I assume the scoring and roster size aren't a factor (for adjusting the formula), but the amount of starters are a factor?

if so, perhaps some of us can list what our starting Requirements are and then we could get a discussion on what you guys think we should do with the app to get a realistic value sheet?

Here are 2:

Standard league: 1 QB, 2 RB, 3 WR, 1 TE, 1 K, 1 Def

What about this league:

2 QB's, 3 RB's, 3 WR's, 2 TE's, 2 kickers and the IDP of 2 CB's, 2 safeties, 3 DL and 3 LB's

These were 2 extremes that I thought may get our minds around what you are saying.

Thanks again!

 
I do a custom baseline.

We start 8 in my primary league. What I do is take a 3-2-1 weighted average of the total number of players from each position drafted in the first 8 rounds over the past three years.

This way, it's customized to my league's drafting tendencies and by not using the "worst starter" method, DD isn't telling me every round from the 8th on that I should be drafting a kicker (advice, which, of course, I ignore).

 
Liquid Tension said:
...I have different leagues and while I don't use it as the bible, once I alter the projections, adjust the scoring and setup the application per the league requirements (teams rounds etc...) my expectation is that this application would give me the values of the players taking it all into context?It appears this is not correct as the starters change the context more than I expected (I assumed it was already in there).Can I assume the scoring and roster size aren't a factor (for adjusting the formula), but the amount of starters are a factor?
Click to expand...
I think that is probably a correct statement. Changing bench space and scoring probably won't change where Joe's formula puts the baseline, and doing a test right now with 2 pt WR receptions and doubling bench space seems to confirm that.Now I'm sitting here trying to think of whether there are scenarios where you would WANT them change due to changes in bench or scoring, even though Joe's doesn't. So far I don't think I don't think I've come up with any realistic situations that would be so. Well... maybe changes to TE that stagger PPR in his favor, as that could create more separation in the TE curve to where I'd want to extend the baseline further down the curve, but it would have to be significant. And probably coupled with starting more TEs.
if so, perhaps some of us can list what our starting Requirements are and then we could get a discussion on what you guys think we should do with the app to get a realistic value sheet?Here are 2:Standard league: 1 QB, 2 RB, 3 WR, 1 TE, 1 K, 1 DefWhat about this league:2 QB's, 3 RB's, 3 WR's, 2 TE's, 2 kickers and the IDP of 2 CB's, 2 safeties, 3 DL and 3 LB'sThese were 2 extremes that I thought may get our minds around what you are saying.Thanks again!
Click to expand...
In the first league, I think Joe's formula is probably a pretty good baseline.In the second league, I would not use Joe's. Some of these I think are pretty solid reasons, others are my personal preference. I'll try to be clear on which.In the second league, QB scarcity is definitely a concern. Assuming 12 teams, there are not enough QBs to go around. Joe's formula would have the baseline not be as far down the chart as last starter, while I would want the baseline to be somewhere below the last starter. 12 team league I'd probably set at QB 26. In fact, no probably... my leagues both have a 2nd QB via a flex that I treat as if it should always be QB, and the one that uses a draft instead of an auction I put my baseline at QB26.RBs are also even more important in your setup. It looks like Joe's puts it at RB46. My first thought was that might be a little too far, but then I thought I'd probably want most all of the RBBC guys in it, and it looks like that is about the cut off point, so I would probably be ok with that.WR, Joe's is probably good.TE I might move the baseline. If you have PPR, and especially PPR that favors TE over other positions (like mine are 1 PPR TE, .5 WR, .25 RB), then I'd definitely change it. Joe's formula has it at TE 12, which only gives Witten's 121 points a 51 VBD. But the 24th TE, Royal, is only scoring 47. I think when you only start 12 TE then Joe's formula isn't bad since TEs 6-12 are probably fairly interchangeable. But I think there is a fair difference between TEs in the 12-15 range like Shockey, Scheffler, Heath Miller... vs guys in the 20-24 range like Chris Baker, Bo Scaife, Robert Royal, etc. Possibly 2 fantasy points per week. So I think I'd want to move that baseline lower than 12. I might set it at about 16.Ok those were the ones I think my reasoning is pretty solid. Kickers and IDPs you may disagree with me on, especially IDPs. I consider the vast majority of IDPs to be a crapshoot. I think only the top few players at each position are guys that you can put much faith in a projection for. Guys who are beasts and should always get numbers like Jared Allen, Mario Williams, Patrick Willis, Adrian Wilson, etc. Even they are not sure things... but once I get past the top 3-5 IDPs at each position, I can normally find guys rated in the 40s or even 60s in FBGs who I think will do as well as guys they have rated 10th. I've done some studies on my leagues, and on my teams in my leagues, and I seldom end the season with the IDP I drafted unless I drafted those top guys. The other teams in my leagues, normally about 60% of their end of the season IDPs came from waivers at some point, for leagues that start 2 DL, 2 LB, 2 DB.So I use a small baseline for IDP. Kicker same thing, not very predictable. I just checked my league we have a draft in... for 12 teams and starting 2 DL, 2 LB, and 2 DB, I have set the DL, LB and DB baseline to 8 for all three positions. Last starter would be 24, it looks like Joe is using 12 for DB and DL and 14 for LB. So I'm setting mine to about 1/3 of last starter, while Joe is setting his to about 1/2 of last starter. Since you start 3 DL and 3 LB I'd maybe push those back to 12, and with CB and safety separated.... um... probably I'd stick with around 8.Ok, now all that said, that's what I'd start with. Then I'd generate the overall list and take a look at it. It may not end up looking how you expect. And then you have a tough question... does that mean the baseline is set wrong... or does it mean that your prior conception of how it should look is off, and that some position has more or less value than you thought.There's no right answer. About all I could suggest to double check yourself is to monkey with it, see how things change, maybe mock draft a few teams vs the ADP to see which set of baselines results in a better team overall. But you've been in FF for awhile I know, so I'm sure you probably have a decent grasp on the value the league should have from seeing what strategies have been successful in it. That's a good guide. Harder to figure out for a newbie.
 
Last edited by a moderator:
GregR said:
Liquid Tension said:
...I have different leagues and while I don't use it as the bible, once I alter the projections, adjust the scoring and setup the application per the league requirements (teams rounds etc...) my expectation is that this application would give me the values of the players taking it all into context?It appears this is not correct as the starters change the context more than I expected (I assumed it was already in there).Can I assume the scoring and roster size aren't a factor (for adjusting the formula), but the amount of starters are a factor?
Click to expand...
I think that is probably a correct statement. Changing bench space and scoring probably won't change where Joe's formula puts the baseline, and doing a test right now with 2 pt WR receptions and doubling bench space seems to confirm that.Now I'm sitting here trying to think of whether there are scenarios where you would WANT them change due to changes in bench or scoring, even though Joe's doesn't. So far I don't think I don't think I've come up with any realistic situations that would be so. Well... maybe changes to TE that stagger PPR in his favor, as that could create more separation in the TE curve to where I'd want to extend the baseline further down the curve, but it would have to be significant. And probably coupled with starting more TEs.
if so, perhaps some of us can list what our starting Requirements are and then we could get a discussion on what you guys think we should do with the app to get a realistic value sheet?Here are 2:Standard league: 1 QB, 2 RB, 3 WR, 1 TE, 1 K, 1 DefWhat about this league:2 QB's, 3 RB's, 3 WR's, 2 TE's, 2 kickers and the IDP of 2 CB's, 2 safeties, 3 DL and 3 LB'sThese were 2 extremes that I thought may get our minds around what you are saying.Thanks again!
Click to expand...
In the first league, I think Joe's formula is probably a pretty good baseline.In the second league, I would not use Joe's. Some of these I think are pretty solid reasons, others are my personal preference. I'll try to be clear on which.In the second league, QB scarcity is definitely a concern. Assuming 12 teams, there are not enough QBs to go around. Joe's formula would have the baseline not be as far down the chart as last starter, while I would want the baseline to be somewhere below the last starter. 12 team league I'd probably set at QB 26. In fact, no probably... my leagues both have a 2nd QB via a flex that I treat as if it should always be QB, and the one that uses a draft instead of an auction I put my baseline at QB26.RBs are also even more important in your setup. It looks like Joe's puts it at RB46. My first thought was that might be a little too far, but then I thought I'd probably want most all of the RBBC guys in it, and it looks like that is about the cut off point, so I would probably be ok with that.WR, Joe's is probably good.TE I might move the baseline. If you have PPR, and especially PPR that favors TE over other positions (like mine are 1 PPR TE, .5 WR, .25 RB), then I'd definitely change it. Joe's formula has it at TE 12, which only gives Witten's 121 points a 51 VBD. But the 24th TE, Royal, is only scoring 47. I think when you only start 12 TE then Joe's formula isn't bad since TEs 6-12 are probably fairly interchangeable. But I think there is a fair difference between TEs in the 12-15 range like Shockey, Scheffler, Heath Miller... vs guys in the 20-24 range like Chris Baker, Bo Scaife, Robert Royal, etc. Possibly 2 fantasy points per week. So I think I'd want to move that baseline lower than 12. I might set it at about 16.Ok those were the ones I think my reasoning is pretty solid. Kickers and IDPs you may disagree with me on, especially IDPs. I consider the vast majority of IDPs to be a crapshoot. I think only the top few players at each position are guys that you can put much faith in a projection for. Guys who are beasts and should always get numbers like Jared Allen, Mario Williams, Patrick Willis, Adrian Wilson, etc. Even they are not sure things... but once I get past the top 3-5 IDPs at each position, I can normally find guys rated in the 40s or even 60s in FBGs who I think will do as well as guys they have rated 10th. I've done some studies on my leagues, and on my teams in my leagues, and I seldom end the season with the IDP I drafted unless I drafted those top guys. The other teams in my leagues, normally about 60% of their end of the season IDPs came from waivers at some point, for leagues that start 2 DL, 2 LB, 2 DB.So I use a small baseline for IDP. Kicker same thing, not very predictable. I just checked my league we have a draft in... for 12 teams and starting 2 DL, 2 LB, and 2 DB, I have set the DL, LB and DB baseline to 8 for all three positions. Last starter would be 24, it looks like Joe is using 12 for DB and DL and 14 for LB. So I'm setting mine to about 1/3 of last starter, while Joe is setting his to about 1/2 of last starter. Since you start 3 DL and 3 LB I'd maybe push those back to 12, and with CB and safety separated.... um... probably I'd stick with around 8.Ok, now all that said, that's what I'd start with. Then I'd generate the overall list and take a look at it. It may not end up looking how you expect. And then you have a tough question... does that mean the baseline is set wrong... or does it mean that your prior conception of how it should look is off, and that some position has more or less value than you thought.There's no right answer. About all I could suggest to double check yourself is to monkey with it, see how things change, maybe mock draft a few teams vs the ADP to see which set of baselines results in a better team overall. But you've been in FF for awhile I know, so I'm sure you probably have a decent grasp on the value the league should have from seeing what strategies have been successful in it. That's a good guide. Harder to figure out for a newbie.
Click to expand...
:shock: GregR...thanks. This was an excellent post giving me a much better understanding of the thought process behind it. BTW, the "2nd" league has 10 teams in it so what impact would that have on your baselines for the positions? You mentioned the scarcity of the QB's, but there is a scarcity at RB as well because of starting 3. Oh and it .5 PPR but for all positions
 
Last edited by a moderator:
Chumpson said:
dino259 said:
drm1125 said:
I'll be truthful, I don't really understand the draft dominator stuff. So, I use the VBD with my leagues scoring, etc. inputed in it. I understand it better and is easy for me to follow, if I wish, during the draft.
Click to expand...
Joe's formula is also a factor in the VBD app.
Click to expand...
I always fiddle with it in the VBD app for my auctions to weigh things differently. I have a mac so I have little experience with DD which bothers me, wish it worked on my platform too.
Click to expand...
Use Boot Camp to create a partition and then install windows on said partition. I did this yesterday, works great, downloaded DD, installed MS Office, and then installed VBD. You need to have an Intel Mac with a large enough harddrive. If youre interested, Google "Boot Camp" and I am sure youll figure it out.ETA: I don't really use the applications, but I figured I would take a look - still looks like a foreign language to me.
 
Last edited by a moderator:
whitem0nkey said:
please make a mac version. one that does not require us mac users to install windows and a few other softwares.
Click to expand...
I've been asking for 5 years, not going to happen, I just installed parallels on my MacBookPro, works flawlessly.
 
Last edited by a moderator:
Liquid Tension said:
GregR said:
Liquid Tension said:
...I have different leagues and while I don't use it as the bible, once I alter the projections, adjust the scoring and setup the application per the league requirements (teams rounds etc...) my expectation is that this application would give me the values of the players taking it all into context?It appears this is not correct as the starters change the context more than I expected (I assumed it was already in there).Can I assume the scoring and roster size aren't a factor (for adjusting the formula), but the amount of starters are a factor?
Click to expand...
I think that is probably a correct statement. Changing bench space and scoring probably won't change where Joe's formula puts the baseline, and doing a test right now with 2 pt WR receptions and doubling bench space seems to confirm that.Now I'm sitting here trying to think of whether there are scenarios where you would WANT them change due to changes in bench or scoring, even though Joe's doesn't. So far I don't think I don't think I've come up with any realistic situations that would be so. Well... maybe changes to TE that stagger PPR in his favor, as that could create more separation in the TE curve to where I'd want to extend the baseline further down the curve, but it would have to be significant. And probably coupled with starting more TEs.
if so, perhaps some of us can list what our starting Requirements are and then we could get a discussion on what you guys think we should do with the app to get a realistic value sheet?Here are 2:Standard league: 1 QB, 2 RB, 3 WR, 1 TE, 1 K, 1 DefWhat about this league:2 QB's, 3 RB's, 3 WR's, 2 TE's, 2 kickers and the IDP of 2 CB's, 2 safeties, 3 DL and 3 LB'sThese were 2 extremes that I thought may get our minds around what you are saying.Thanks again!
Click to expand...
In the first league, I think Joe's formula is probably a pretty good baseline.In the second league, I would not use Joe's. Some of these I think are pretty solid reasons, others are my personal preference. I'll try to be clear on which.In the second league, QB scarcity is definitely a concern. Assuming 12 teams, there are not enough QBs to go around. Joe's formula would have the baseline not be as far down the chart as last starter, while I would want the baseline to be somewhere below the last starter. 12 team league I'd probably set at QB 26. In fact, no probably... my leagues both have a 2nd QB via a flex that I treat as if it should always be QB, and the one that uses a draft instead of an auction I put my baseline at QB26.RBs are also even more important in your setup. It looks like Joe's puts it at RB46. My first thought was that might be a little too far, but then I thought I'd probably want most all of the RBBC guys in it, and it looks like that is about the cut off point, so I would probably be ok with that.WR, Joe's is probably good.TE I might move the baseline. If you have PPR, and especially PPR that favors TE over other positions (like mine are 1 PPR TE, .5 WR, .25 RB), then I'd definitely change it. Joe's formula has it at TE 12, which only gives Witten's 121 points a 51 VBD. But the 24th TE, Royal, is only scoring 47. I think when you only start 12 TE then Joe's formula isn't bad since TEs 6-12 are probably fairly interchangeable. But I think there is a fair difference between TEs in the 12-15 range like Shockey, Scheffler, Heath Miller... vs guys in the 20-24 range like Chris Baker, Bo Scaife, Robert Royal, etc. Possibly 2 fantasy points per week. So I think I'd want to move that baseline lower than 12. I might set it at about 16.Ok those were the ones I think my reasoning is pretty solid. Kickers and IDPs you may disagree with me on, especially IDPs. I consider the vast majority of IDPs to be a crapshoot. I think only the top few players at each position are guys that you can put much faith in a projection for. Guys who are beasts and should always get numbers like Jared Allen, Mario Williams, Patrick Willis, Adrian Wilson, etc. Even they are not sure things... but once I get past the top 3-5 IDPs at each position, I can normally find guys rated in the 40s or even 60s in FBGs who I think will do as well as guys they have rated 10th. I've done some studies on my leagues, and on my teams in my leagues, and I seldom end the season with the IDP I drafted unless I drafted those top guys. The other teams in my leagues, normally about 60% of their end of the season IDPs came from waivers at some point, for leagues that start 2 DL, 2 LB, 2 DB.So I use a small baseline for IDP. Kicker same thing, not very predictable. I just checked my league we have a draft in... for 12 teams and starting 2 DL, 2 LB, and 2 DB, I have set the DL, LB and DB baseline to 8 for all three positions. Last starter would be 24, it looks like Joe is using 12 for DB and DL and 14 for LB. So I'm setting mine to about 1/3 of last starter, while Joe is setting his to about 1/2 of last starter. Since you start 3 DL and 3 LB I'd maybe push those back to 12, and with CB and safety separated.... um... probably I'd stick with around 8.Ok, now all that said, that's what I'd start with. Then I'd generate the overall list and take a look at it. It may not end up looking how you expect. And then you have a tough question... does that mean the baseline is set wrong... or does it mean that your prior conception of how it should look is off, and that some position has more or less value than you thought.There's no right answer. About all I could suggest to double check yourself is to monkey with it, see how things change, maybe mock draft a few teams vs the ADP to see which set of baselines results in a better team overall. But you've been in FF for awhile I know, so I'm sure you probably have a decent grasp on the value the league should have from seeing what strategies have been successful in it. That's a good guide. Harder to figure out for a newbie.
Click to expand...
:goodposting: GregR...thanks. This was an excellent post giving me a much better understanding of the thought process behind it. BTW, the "2nd" league has 10 teams in it so what impact would that have on your baselines for the positions? You mentioned the scarcity of the QB's, but there is a scarcity at RB as well because of starting 3. Oh and it .5 PPR but for all positions
Click to expand...
10 team league some of what I said won't apply, or at least not as much. QB scarcity is not as big of a problem. Joe's is setting at around .9 * # of QB starters and for 12 teams I was setting it at about 1.1 of QB starters. I might slide it between what I'd said for 12 teams and what Joe has... QB 20, maybe QB21 which would be just about last starter.RB, Joe's formula is probably still good for 10 team. Same with WR.TE I'd also back off how far I'd have moved it... since I think TEs 20+ are a crapshoot but now with 10 teams you're only starting 20, I might back it off to TE 13 or 14, while Joe probably has it at TE 10. IDP's I'd just stay proportional... if I was going with 8 for starting 24, if you're only starting 20 I'd make it 7 or 6. Although I pretty much ignore the cheat sheet when it comes to IDPs. If I see a chance to get one of those top guys after I'm happy with my starters, or at least when I think I can fill my starters in later rounds with equiv players, then I'd go after one. For me personally, setting baselines for IDPs is more valuable for my auction league, where I am using DD to express auction values and I want those top guys to be accurate price and everyone after them at IDP that I consider to be a crapshoot to be a minimum bid.One other note, because different positions decline at different rates, it's hard to have a single baseline that really does a great job for all the starters at all points in the draft. Generally the top players you can get to be most accurate and hopefully the middle guys fairly accurate. But by the time you get down to the bottom 1/3 of a position like RB or WR then I'm looking at upside and replaceability as much or even more than I'm looking at projected points or x-value. And actually I probably almost ignore the X-value for those bottom players unless I go in and move the baseline to start dealing with backups. This is why the VBD excel tool mentions to stop using the cheat sheet once you get down near the backup level.Just good to keep in mind that it's a tool to help with decision making, but there are other factors (bye weeks, upside, risk of losing starting job, etc) that you can't factor into a single number in your projection. So I think the tool of value based analysis is best suited to the top guys, where the projection probably does a really good job of summing up your thoughts on the player, much better than it does when you're looking at a Robert Meachem whose value probably is driven more by the possibility he takes over a starting job on a great offense and in the handcuffing of Lance Moore, rather than his projection which is probably what it is thought he'll score as the WR3.
 
Last edited by a moderator:
sjacksonfan said:
What is Joe's Formula/worst starter?
Click to expand...
They are baselines used to express the value of a player within a position. That is... if the top TE and top QB each score 250 in some scoring system, that doesn't mean they are necessarily equal. If TE's 2 through 12 tend to score less than QBs 2 through 12, that would mean that TE1 is a bigger benefit to your team. He outscores the TEs other teams will start by more than QB1 will outscore the QBs other teams will start.One way to express this is to normalize each players points based on what his position scores by subtracting some baseline number of points from each player in the position. Last starter you subtract the last fantasy starter (12 teams starting 1 QB each, then last starter is QB12). Some people average the starter points at the position and use that as the baseline. Joe's formula is an option coded into the VBD tool and Draft Dominator that is a way of calculating what baseline player to use based on what Joe thinks does the best job of properly ranking the players for normal league setups. Taking into account things like RB scarcity, easier replaceability of WRs and TEs, etc.
 
mjm3773 said:
I do a custom baseline.We start 8 in my primary league. What I do is take a 3-2-1 weighted average of the total number of players from each position drafted in the first 8 rounds over the past three years.This way, it's customized to my league's drafting tendencies and by not using the "worst starter" method, DD isn't telling me every round from the 8th on that I should be drafting a kicker (advice, which, of course, I ignore).
Click to expand...
This sounds interesting to me. Can you explain in more detail how your are doing the 3-2-1 weighted average? What value are you assigning for the weighted average...the draft spot?
 
chihawk said:
mjm3773 said:
I do a custom baseline.We start 8 in my primary league. What I do is take a 3-2-1 weighted average of the total number of players from each position drafted in the first 8 rounds over the past three years.This way, it's customized to my league's drafting tendencies and by not using the "worst starter" method, DD isn't telling me every round from the 8th on that I should be drafting a kicker (advice, which, of course, I ignore).
Click to expand...
This sounds interesting to me. Can you explain in more detail how your are doing the 3-2-1 weighted average? What value are you assigning for the weighted average...the draft spot?
Click to expand...
Last year, in the first 8 rounds in my league, 11 qbs, 34 rbs, 31 wrs, 10 tes and 10 dsts were drafted. Since that is the most recent year, and thus most indicative of recent trends, I multiply each of those values by 3 (33, 102, 93, 30 and 30, respectively). I do the same for the positions taken in the 2007 draft, only just using 2 as the multiplier, and then just using one for the multiplier for the positions taken in the first 8 rounds of the 2006 draft.I then add up the totals for each position and divide by 6 (because of the 3-2-1 weighting) to yield an average number of how many of each position is taken in the first 8 rounds of our draft and then I use that number in the custom baseline setup for that position in the DD.
 
How to set Baselines: The baselines are what rates players across positions. For Example, a baseline of 8 for QBs & 40 for RBs and WRs means that the 8th best QB is equivalent in value (VBD) to the 40th ranked RB and WR. The key is finding a baseline that makes sense to you and the league you’re in. You must be comfortable that it is accurately telling you the value (VBD) of QB vs. RB vs. WR vs. TE vs. PK vs. DEF.

Here is what I would do to set baselines in my league (12 team PPR league with 18 draft slots):

Baseline #1 (Starter Baseline – 96 Total): What is the average number taken at each position (including keepers) after Round 8 is complete. This is your "starter" baseline. Why round 8? Excluding kicker’s and defense who aren’t usually starting to come off the board until after round 8 and don’t account for much fantasy value (in most leagues, including mine), 8 is the total number of starters (1 QB, 2 RB, 3 WR, 1 TE & 1 Flex, in my league). You need to review past draft data to get this information. In the Draft Dominator, go to the VBD Baseline tab in setup and enter this information in the user configurable by position rank area. This baseline shows what the league believes is the value of their starting team. I would use this baseline until your pick in round 8. (Example: QB – 8, RB – 40, WR – 40, TE – 8, PK – 0, DEF – 0).

Baseline #2 (Bench Baseline – 216 Total): I would use the average number (including keepers) of each position taken at the end of the draft (12 teams x 18 rounds = 216 total players selected). You need to review past draft data to get this information. You can either change the “user configurable by position rank” at the end of round 8, or what is easier is to use the "user configurable by point value" area for this. Just look in the player pool and find the projected points for the appropriate player at each position representing the last player of that position that should be taken in the draft based on the number of each position that should be drafted. These numbers will stay there and all you have to do during the draft is after round 8 is click to switch over to the Bench Baseline by changing from “user configurable by position rank” to “point value”.

Other Notes about baselines: Tweaking baselines for total players taken at each position will give you all positive VBDs.

Hope this helps,

5thQuarter

 
Last edited by a moderator:
GregR said:
Liquid Tension said:
GregR said:
Liquid Tension said:
...I have different leagues and while I don't use it as the bible, once I alter the projections, adjust the scoring and setup the application per the league requirements (teams rounds etc...) my expectation is that this application would give me the values of the players taking it all into context?It appears this is not correct as the starters change the context more than I expected (I assumed it was already in there).Can I assume the scoring and roster size aren't a factor (for adjusting the formula), but the amount of starters are a factor?
Click to expand...
I think that is probably a correct statement. Changing bench space and scoring probably won't change where Joe's formula puts the baseline, and doing a test right now with 2 pt WR receptions and doubling bench space seems to confirm that.Now I'm sitting here trying to think of whether there are scenarios where you would WANT them change due to changes in bench or scoring, even though Joe's doesn't. So far I don't think I don't think I've come up with any realistic situations that would be so. Well... maybe changes to TE that stagger PPR in his favor, as that could create more separation in the TE curve to where I'd want to extend the baseline further down the curve, but it would have to be significant. And probably coupled with starting more TEs.
if so, perhaps some of us can list what our starting Requirements are and then we could get a discussion on what you guys think we should do with the app to get a realistic value sheet?Here are 2:Standard league: 1 QB, 2 RB, 3 WR, 1 TE, 1 K, 1 DefWhat about this league:2 QB's, 3 RB's, 3 WR's, 2 TE's, 2 kickers and the IDP of 2 CB's, 2 safeties, 3 DL and 3 LB'sThese were 2 extremes that I thought may get our minds around what you are saying.Thanks again!
Click to expand...
In the first league, I think Joe's formula is probably a pretty good baseline.In the second league, I would not use Joe's. Some of these I think are pretty solid reasons, others are my personal preference. I'll try to be clear on which.In the second league, QB scarcity is definitely a concern. Assuming 12 teams, there are not enough QBs to go around. Joe's formula would have the baseline not be as far down the chart as last starter, while I would want the baseline to be somewhere below the last starter. 12 team league I'd probably set at QB 26. In fact, no probably... my leagues both have a 2nd QB via a flex that I treat as if it should always be QB, and the one that uses a draft instead of an auction I put my baseline at QB26.RBs are also even more important in your setup. It looks like Joe's puts it at RB46. My first thought was that might be a little too far, but then I thought I'd probably want most all of the RBBC guys in it, and it looks like that is about the cut off point, so I would probably be ok with that.WR, Joe's is probably good.TE I might move the baseline. If you have PPR, and especially PPR that favors TE over other positions (like mine are 1 PPR TE, .5 WR, .25 RB), then I'd definitely change it. Joe's formula has it at TE 12, which only gives Witten's 121 points a 51 VBD. But the 24th TE, Royal, is only scoring 47. I think when you only start 12 TE then Joe's formula isn't bad since TEs 6-12 are probably fairly interchangeable. But I think there is a fair difference between TEs in the 12-15 range like Shockey, Scheffler, Heath Miller... vs guys in the 20-24 range like Chris Baker, Bo Scaife, Robert Royal, etc. Possibly 2 fantasy points per week. So I think I'd want to move that baseline lower than 12. I might set it at about 16.Ok those were the ones I think my reasoning is pretty solid. Kickers and IDPs you may disagree with me on, especially IDPs. I consider the vast majority of IDPs to be a crapshoot. I think only the top few players at each position are guys that you can put much faith in a projection for. Guys who are beasts and should always get numbers like Jared Allen, Mario Williams, Patrick Willis, Adrian Wilson, etc. Even they are not sure things... but once I get past the top 3-5 IDPs at each position, I can normally find guys rated in the 40s or even 60s in FBGs who I think will do as well as guys they have rated 10th. I've done some studies on my leagues, and on my teams in my leagues, and I seldom end the season with the IDP I drafted unless I drafted those top guys. The other teams in my leagues, normally about 60% of their end of the season IDPs came from waivers at some point, for leagues that start 2 DL, 2 LB, 2 DB.So I use a small baseline for IDP. Kicker same thing, not very predictable. I just checked my league we have a draft in... for 12 teams and starting 2 DL, 2 LB, and 2 DB, I have set the DL, LB and DB baseline to 8 for all three positions. Last starter would be 24, it looks like Joe is using 12 for DB and DL and 14 for LB. So I'm setting mine to about 1/3 of last starter, while Joe is setting his to about 1/2 of last starter. Since you start 3 DL and 3 LB I'd maybe push those back to 12, and with CB and safety separated.... um... probably I'd stick with around 8.Ok, now all that said, that's what I'd start with. Then I'd generate the overall list and take a look at it. It may not end up looking how you expect. And then you have a tough question... does that mean the baseline is set wrong... or does it mean that your prior conception of how it should look is off, and that some position has more or less value than you thought.There's no right answer. About all I could suggest to double check yourself is to monkey with it, see how things change, maybe mock draft a few teams vs the ADP to see which set of baselines results in a better team overall. But you've been in FF for awhile I know, so I'm sure you probably have a decent grasp on the value the league should have from seeing what strategies have been successful in it. That's a good guide. Harder to figure out for a newbie.
Click to expand...
:goodposting: GregR...thanks. This was an excellent post giving me a much better understanding of the thought process behind it. BTW, the "2nd" league has 10 teams in it so what impact would that have on your baselines for the positions? You mentioned the scarcity of the QB's, but there is a scarcity at RB as well because of starting 3. Oh and it .5 PPR but for all positions
Click to expand...
10 team league some of what I said won't apply, or at least not as much. QB scarcity is not as big of a problem. Joe's is setting at around .9 * # of QB starters and for 12 teams I was setting it at about 1.1 of QB starters. I might slide it between what I'd said for 12 teams and what Joe has... QB 20, maybe QB21 which would be just about last starter.RB, Joe's formula is probably still good for 10 team. Same with WR.TE I'd also back off how far I'd have moved it... since I think TEs 20+ are a crapshoot but now with 10 teams you're only starting 20, I might back it off to TE 13 or 14, while Joe probably has it at TE 10. IDP's I'd just stay proportional... if I was going with 8 for starting 24, if you're only starting 20 I'd make it 7 or 6. Although I pretty much ignore the cheat sheet when it comes to IDPs. If I see a chance to get one of those top guys after I'm happy with my starters, or at least when I think I can fill my starters in later rounds with equiv players, then I'd go after one. For me personally, setting baselines for IDPs is more valuable for my auction league, where I am using DD to express auction values and I want those top guys to be accurate price and everyone after them at IDP that I consider to be a crapshoot to be a minimum bid.One other note, because different positions decline at different rates, it's hard to have a single baseline that really does a great job for all the starters at all points in the draft. Generally the top players you can get to be most accurate and hopefully the middle guys fairly accurate. But by the time you get down to the bottom 1/3 of a position like RB or WR then I'm looking at upside and replaceability as much or even more than I'm looking at projected points or x-value. And actually I probably almost ignore the X-value for those bottom players unless I go in and move the baseline to start dealing with backups. This is why the VBD excel tool mentions to stop using the cheat sheet once you get down near the backup level.Just good to keep in mind that it's a tool to help with decision making, but there are other factors (bye weeks, upside, risk of losing starting job, etc) that you can't factor into a single number in your projection. So I think the tool of value based analysis is best suited to the top guys, where the projection probably does a really good job of summing up your thoughts on the player, much better than it does when you're looking at a Robert Meachem whose value probably is driven more by the possibility he takes over a starting job on a great offense and in the handcuffing of Lance Moore, rather than his projection which is probably what it is thought he'll score as the WR3.
Click to expand...
Thanks again Greg! OK, based on what you said, here is what I was thinking. 10 teams.2 QB's - setting at 20-21 (the difference is surprisingly large)3 RB's - setting at 383 WR's - setting at 302 TE's - setting at 133 DL - setting at 113 LB's - setting at 14 4 DB's - setting at 13 (I need to separate the CB's and S when I draft)Intuitively I think the IDP baseline should be set higher but use the strategy of looking at the top guys and then adjusting downward after the first 3 or so go off the board. What I then did was look at the drafts and get an idea of where guys are going. With that in mind I adjusted the settings to match the actual draft and this is what I came up with.2 QB's - setting at 24 (the difference is surprisingly large)3 RB's - setting at 383 WR's - setting at 382 TE's - setting at 163 DL - setting at 113 LB's - setting at 30 4 DB's - setting at 13While adjustments were made it is interesting to note the main thing that stands out; the linebackers being so much higher than the DL. The changes didn't seem to have a large impact so moving it so much higher than the DL isn't as huge as it seems. The LB's consistency seems to be getting them drafted higher, but the value of the DL is pretty good because sacks are worth 8 points in this league (very high number). INT's are worth 8 also FWIW.You make a point that I agree with and follow; I only use the VBD for the first bunch of rounds or for starters.So, am I off by a lot?
 
mjm3773 said:
chihawk said:
mjm3773 said:
I do a custom baseline.We start 8 in my primary league. What I do is take a 3-2-1 weighted average of the total number of players from each position drafted in the first 8 rounds over the past three years.This way, it's customized to my league's drafting tendencies and by not using the "worst starter" method, DD isn't telling me every round from the 8th on that I should be drafting a kicker (advice, which, of course, I ignore).
Click to expand...
This sounds interesting to me. Can you explain in more detail how your are doing the 3-2-1 weighted average? What value are you assigning for the weighted average...the draft spot?
Click to expand...
Last year, in the first 8 rounds in my league, 11 qbs, 34 rbs, 31 wrs, 10 tes and 10 dsts were drafted. Since that is the most recent year, and thus most indicative of recent trends, I multiply each of those values by 3 (33, 102, 93, 30 and 30, respectively). I do the same for the positions taken in the 2007 draft, only just using 2 as the multiplier, and then just using one for the multiplier for the positions taken in the first 8 rounds of the 2006 draft.I then add up the totals for each position and divide by 6 (because of the 3-2-1 weighting) to yield an average number of how many of each position is taken in the first 8 rounds of our draft and then I use that number in the custom baseline setup for that position in the DD.
Click to expand...
So if you took only the last years draft and you drafted 28 RB's in the 1st 8 rounds, would you use 28 or 14 as your number?I also assume that you are only using the DD for the first 8 or so rounds?
 
5thQuarter said:
How to set Baselines: The baselines are what rates players across positions. For Example, a baseline of 8 for QBs & 40 for RBs and WRs means that the 8th best QB is equivalent in value (VBD) to the 40th ranked RB and WR. The key is finding a baseline that makes sense to you and the league you’re in. You must be comfortable that it is accurately telling you the value (VBD) of QB vs. RB vs. WR vs. TE vs. PK vs. DEF.

Here is what I would do to set baselines in my league:

Baseline #1 (Starter Baseline – 96 Total): What is the average number taken at each position (including keepers) after Round 8 is complete. This is your "starter" baseline. Why round 8? Excluding kicker’s and defense who aren’t usually starting to come off the board until after round 8 and don’t account for much fantasy value (in most leagues, including mine), 8 is the total number of starters (1 QB, 2 RB, 3 WR, 1 TE & 1 Flex, in my league). You need to review past draft data to get this information. In the Draft Dominator, go to the VBD Baseline tab in setup and enter this information in the user configurable by position rank area. This baseline shows what the league believes is the value of their starting team. I would use this baseline until your pick in round 8. (Example: QB – 8, RB – 40, WR – 40, TE – 8, PK – 0, DEF – 0).

Baseline #2 (Bench Baseline – 216 Total): I would use the average number (including keepers) of each position taken at the end of the draft (12 teams x 18 rounds = 216 total players selected). You need to review past draft data to get this information. You can either change the “user configurable by position rank” at the end of round 8, or what is easier is to use the "user configurable by point value" area for this. Just look in the player pool and find the projected points for the appropriate player at each position representing the last player of that position that should be taken in the draft based on the number of each position that should be drafted. These numbers will stay there and all you have to do during the draft is after round 8 is click to switch over to the Bench Baseline by changing from “user configurable by position rank” to “point value”.

Other Notes about baselines: Tweaking baselines for total players taken at each position will give you all positive VBDs.

Hope this helps,

5thQuarter
Click to expand...
This is interesting because this is the first someone has talked about the configurable point value...I need to review these and play around with it.As for the first paragraph, I took the DD as the application to help me understand the values across multiple positions based on my scoring and starting requirements. So if I am figuring that out then where is the value other than sorting it for me? I thought the idea was that the program "figured out" that the differential between the last starter and the first starter was greatest in the "RB" category because you are starting 3 RB's and then when scarcity is taken into account here is who you take?

So taking the above isn't there a way to take the real starting positions in the game and compare it to the starting requirements of the league to come up with the "scarcity level" and then the point differential based on the projections?

for example, if the league starts 3 RB's that is 30 starters (10 teams) out of 32 teams. I guess for RB's you have to include some of the RBBc guys, but let's just take this example at base value and say 94% of the positions will be drafted. TE's (starting 2) would be 20 out of 32 or 63%, same for qb's. WR's would be 30 out of 64 or 47%...DL and LB's would be harder to determine as some tackles would get a little value but mostly the DE's would be drafted so let's say (start 3) 30 out of 64 also, LB's I would argue that it is 30 out of 96 -31%-(because all 3 positions could be used) and CB's/safeties would be 40 out of 128 (31%)

Now maybe my % for IDP's are off, but you get my point.

Thoughts?

 
Last edited by a moderator:
Liquid Tension said:
As for the first paragraph, I took the DD as the application to help me understand the values across multiple positions based on my scoring and starting requirements. So if I am figuring that out then where is the value other than sorting it for me? I thought the idea was that the program "figured out" that the differential between the last starter and the first starter was greatest in the "RB" category because you are starting 3 RB's and then when scarcity is taken into account here is who you take?
Click to expand...
That's correct. The only thing the baselines do for you is they affect the order in which you players are listed in the player pool overall tad. Baseline's do this by assigning static VBD #'s to each player. That's how the players are ranked in the overall tab of the player pool, by static VBD. (Important Note: Baselines only determine each player's static VBD number and once determined, those players are ranked in the overall tab by VBD.) The static DVD is determined by two factors:1. the players projected total points, and

2. the positional baseline #

Liquid Tension said:
I thought the idea was that the program "figured out" that the differential between the last starter and the first starter was greatest in the "RB" category because you are starting 3 RB's and then when scarcity is taken into account here is who you take?
Click to expand...
There are many ways to determine position scarcity, baselines are only part of the equation and only help for rankings in the overall tab. The real value of the dominator is learning how the best value pick window works and what it is telling you. I encourage you to read the following explanatory topics on the draft dominator and the best value pick window:

Understanding Draft Dominator Features

Best Value Dropoff

Best Value Questions

Hope this helps,

5thQuarter

 
Last edited by a moderator:
Liquid Tension said:
...So, am I off by a lot?
Click to expand...
Sounds pretty good to me. As long as you agree with the ultimate result, that is the important thing. If you don't agree, then you can learn a fair amount by trying to figure out why... whether it's something in choosing the baseline that you didn't take into account, or whether the baseline is good but your preconceived notion of player value wasn't right.Also, one other thing worth mentioning. The overall cheatsheat even for the top players of course needs to be combined with where you think the player will go in the draft. ADP works, but of course if you can better guess at where he'll go in your particular league, all the better.The cheatsheet if done well should tell you where players SHOULD go. But as the league takes players in a different order, the optimum choices for your team may differ from your cheatsheet. The more players at any position who are being taken between your picks, the more correct it becomes to follow the herd. Eventually you reach a point where the guy at another position you would take now as the best value, is still there at your next pick, or a similar player is, so you're better off taking the position being decimated by the other drafters. You may also have a player you think you can get later who you're happy with filling the position you'd be drafting for from the cheatsheet. Maybe you love Benson as a RB2 and are happy waiting a few rounds and filling in other positions, so do so.Draft Dominator has a Best Value feature to help you try to look ahead and see what the drop off in value at each position will be.
 
Liquid Tension said:
mjm3773 said:
chihawk said:
mjm3773 said:
I do a custom baseline.We start 8 in my primary league. What I do is take a 3-2-1 weighted average of the total number of players from each position drafted in the first 8 rounds over the past three years.This way, it's customized to my league's drafting tendencies and by not using the "worst starter" method, DD isn't telling me every round from the 8th on that I should be drafting a kicker (advice, which, of course, I ignore).
Click to expand...
This sounds interesting to me. Can you explain in more detail how your are doing the 3-2-1 weighted average? What value are you assigning for the weighted average...the draft spot?
Click to expand...
Last year, in the first 8 rounds in my league, 11 qbs, 34 rbs, 31 wrs, 10 tes and 10 dsts were drafted. Since that is the most recent year, and thus most indicative of recent trends, I multiply each of those values by 3 (33, 102, 93, 30 and 30, respectively). I do the same for the positions taken in the 2007 draft, only just using 2 as the multiplier, and then just using one for the multiplier for the positions taken in the first 8 rounds of the 2006 draft.I then add up the totals for each position and divide by 6 (because of the 3-2-1 weighting) to yield an average number of how many of each position is taken in the first 8 rounds of our draft and then I use that number in the custom baseline setup for that position in the DD.
Click to expand...
So if you took only the last years draft and you drafted 28 RB's in the 1st 8 rounds, would you use 28 or 14 as your number?I also assume that you are only using the DD for the first 8 or so rounds?
Click to expand...
If I only used last year, I would use 28 as the baseline number. I just use the three year weighted average in case one year was really skewed in terms of positions drafted.This year, my baselines end up being qb-12, rb-36, wr-29, te-10, def-11.I still use the DD after the first 8 rounds because the VBD values still flow pretty well after that (with exceptions, of course, e.g obvious rb handcuffs).
 
Chumpson said:
dino259 said:
drm1125 said:
I'll be truthful, I don't really understand the draft dominator stuff. So, I use the VBD with my leagues scoring, etc. inputed in it. I understand it better and is easy for me to follow, if I wish, during the draft.
Click to expand...
Joe's formula is also a factor in the VBD app.
Click to expand...
I always fiddle with it in the VBD app for my auctions to weigh things differently. I have a mac so I have little experience with DD which bothers me, wish it worked on my platform too.
Click to expand...
I loved the DD when I was a PC guy, but now I'm a Mac guy. I miss it, but I don't want windows polluting my Mac.
 
mjm3773 said:
chihawk said:
mjm3773 said:
I do a custom baseline.We start 8 in my primary league. What I do is take a 3-2-1 weighted average of the total number of players from each position drafted in the first 8 rounds over the past three years.This way, it's customized to my league's drafting tendencies and by not using the "worst starter" method, DD isn't telling me every round from the 8th on that I should be drafting a kicker (advice, which, of course, I ignore).
Click to expand...
This sounds interesting to me. Can you explain in more detail how your are doing the 3-2-1 weighted average? What value are you assigning for the weighted average...the draft spot?
Click to expand...
Last year, in the first 8 rounds in my league, 11 qbs, 34 rbs, 31 wrs, 10 tes and 10 dsts were drafted. Since that is the most recent year, and thus most indicative of recent trends, I multiply each of those values by 3 (33, 102, 93, 30 and 30, respectively). I do the same for the positions taken in the 2007 draft, only just using 2 as the multiplier, and then just using one for the multiplier for the positions taken in the first 8 rounds of the 2006 draft.I then add up the totals for each position and divide by 6 (because of the 3-2-1 weighting) to yield an average number of how many of each position is taken in the first 8 rounds of our draft and then I use that number in the custom baseline setup for that position in the DD.
Click to expand...
Thanks for the clarification...I just didn't get the 3-2-1 of valuing the more reset draft more from the original post....makes sense.I think incorporating the history of your drafts may be a very useful way of tweaking the DD to have it be even more focused on your league. Unrelated to tweaking the DD, I am also currently looking at some stats from previous years drafts to see any positional drafting trends in my league. When was the 1st, 6th, and 12th Defense drafted for the past 3 years. When was the 1st, 6th, and 12th TE drafted for the past 3 years. 1st, 6th, and 12th PK drafted for the past 3 years....and so on. That may help me prepare me for when to expect runs on certain positions.
 
Agree with all of the above.

But setting it at median starter under setup "user configural rank by position" for the first 6 to 8 rounds and then switching to worst starter is just really easy as long as one keeps ADP in mind. And lots and lots of leagues have no past history to go by.

Just always remember that the VBD is just a piece of information relevant to value, not "who do I pick". In other words, it does not tell you when to take a kicker.

Standard league 12 owners: 1 QB, 2 RB, 3 WR, 1 TE, 1 K, 1 Def

Median starter set at: 6QB, 12RB, 18WR, 6TE, 6K, 1DEF.

Worst starter set DD at:12QB, 24RB, 36WR, 12TE, 12K, 12DEF.

How about 10 team league: 2 QB's, 3 RB's, 3 WR's, 2 TE's, 2 kickers and the IDP of 2 CB's, 2 safeties, 3 DL and 3 LB's

Median: 10QB, 15RB, 15WR, 10TE, 10K, 10CB, 10S, 15DL, and 15LB.

Worst: 20QB, 30RB, 30WR, 20TE, 20K, 20CB, 20S, 30DL, 30LB.

Easy-peasy.

 
ookook said:
Just always remember that the VBD is just a piece of information relevant to value, not "who do I pick". In other words, it does not tell you when to take a kicker.
Click to expand...
Actually, isn't that really the point of VBD? It is telling you when to draft a kicker, or a QB, or whatever. If the highest VBD player on your board is a kicker, it's basically telling you, "Hey bud, now's the time to take a kicker, because if you wait until your next pick the dropoff is more severe at this position than at any other."Now of course there are lots of other factors to keep in mind as well, such as your roster needs and whatnot, but those aren't factored into VBD. I'm not saying you just blindly follow theVBD, but the VBD number itself really is telling you which position to take at any point - you just have to take that advice in conjunction with all the other available information to make the best decision for your team.

 
tomarken said:
ookook said:
Just always remember that the VBD is just a piece of information relevant to value, not "who do I pick". In other words, it does not tell you when to take a kicker.
Click to expand...
Actually, isn't that really the point of VBD? It is telling you when to draft a kicker, or a QB, or whatever. If the highest VBD player on your board is a kicker, it's basically telling you, "Hey bud, now's the time to take a kicker, because if you wait until your next pick the dropoff is more severe at this position than at any other."
Click to expand...
What you're referring to there is not VBD. VBD is what is used to determine the order of the cheatsheet, which is an expression of player value based fairly exclusively on the player pool. The cheatsheet that you see in draft dominator includes nothing to do with dropoffs between now and the next time you pick.What you're talking about is a component of value based on both the player pool and the order in which teams will draft them. The method for expressing that numerically normally goes by the name Dynamic VBD or DVBD. But they are fairly different beasts despite both having VBD in the name.

Draft Dominator can do both. VBD is shown in the cheatsheet. DVBD is done in the Best Value window. So no, the cheatsheet does not tell you who to take, or you shouldn't use it that way anyway. It tells you who is the most valuable player (ie. your example kicker) outside of any context of a draft. But to include the information about dropoffs you want to use the Best Value part of Draft Dominator.

For drafts, VBD and DVBD are both great inputs that should both be used to a decision. In an auction on the other hand, VBD is probably about all you need.

 
My point was (an an obvious one at that) that there are times when you can wait several rounds to get the player that may have the most relative value. That is why following ADP is also important.

Beyond ADP and such, VBD and DVBD may not do a very good judge taking into account your team needs. The Best Value and DVBD tries to do this, but I find the need multiplier pretty far off how I like to draft.

Just one piece of information out of several...and perhaps not even the most important.

BTW, I almost always check both median and worst baselines. Before DD I did them in Excel and had them side by side. The declinations are not really very linear and this obviates it.

 
GregR said:
tomarken said:
ookook said:
Just always remember that the VBD is just a piece of information relevant to value, not "who do I pick". In other words, it does not tell you when to take a kicker.
Click to expand...
Actually, isn't that really the point of VBD? It is telling you when to draft a kicker, or a QB, or whatever. If the highest VBD player on your board is a kicker, it's basically telling you, "Hey bud, now's the time to take a kicker, because if you wait until your next pick the dropoff is more severe at this position than at any other."
Click to expand...
What you're referring to there is not VBD. VBD is what is used to determine the order of the cheatsheet, which is an expression of player value based fairly exclusively on the player pool. The cheatsheet that you see in draft dominator includes nothing to do with dropoffs between now and the next time you pick.What you're talking about is a component of value based on both the player pool and the order in which teams will draft them. The method for expressing that numerically normally goes by the name Dynamic VBD or DVBD. But they are fairly different beasts despite both having VBD in the name.

Draft Dominator can do both. VBD is shown in the cheatsheet. DVBD is done in the Best Value window. So no, the cheatsheet does not tell you who to take, or you shouldn't use it that way anyway. It tells you who is the most valuable player (ie. your example kicker) outside of any context of a draft. But to include the information about dropoffs you want to use the Best Value part of Draft Dominator.

For drafts, VBD and DVBD are both great inputs that should both be used to a decision. In an auction on the other hand, VBD is probably about all you need.
Click to expand...
Good stuff everyone. 5th Quarter, I will be reviewing the articles that you listed as well. :thumbup:

 
the one item that still is odd to me is that even with having all the settings the way I had them and moving kicker to a baseline of only 1, they are still overvalued (from a standpoint of where everyone drafts them). Even intuitively it appears the drop off isn't as large as other positions. Any thoughts on this? Is this where the DVBD makes sense as the kickers may not ever be the choice?

 
I use custom.

I have draft data for my league going back 5 years. At some point a long time ago, I don't remember how, I figured for my 12-team league with 14-player rosters that pick #111 was the baseline pick. So every year I use my draft data to calculate how many players at each position are taken by then, weight it for recency, and plug the numbers in. Made sense at the time, still works OK.

 
Philo said:
I use custom.I have draft data for my league going back 5 years. At some point a long time ago, I don't remember how, I figured for my 12-team league with 14-player rosters that pick #111 was the baseline pick. So every year I use my draft data to calculate how many players at each position are taken by then, weight it for recency, and plug the numbers in. Made sense at the time, still works OK.
Click to expand...
what determined that baselien pick? when all the good starters are gone?
 
Philo said:
I use custom.I have draft data for my league going back 5 years. At some point a long time ago, I don't remember how, I figured for my 12-team league with 14-player rosters that pick #111 was the baseline pick. So every year I use my draft data to calculate how many players at each position are taken by then, weight it for recency, and plug the numbers in. Made sense at the time, still works OK.
Click to expand...
Based on mjm3773's post above, I have done the same thing for my league. I used the first 96 picks as my baseline (first 8 rounds) over the past three years. I am going to drop that back to 9, 10, 11, rounds to see what the impact is.
 
Liquid Tension said:
the one item that still is odd to me is that even with having all the settings the way I had them and moving kicker to a baseline of only 1, they are still overvalued (from a standpoint of where everyone drafts them). Even intuitively it appears the drop off isn't as large as other positions. Any thoughts on this? Is this where the DVBD makes sense as the kickers may not ever be the choice?
Click to expand...
I think a big part of the reason that people draft kickers so late is that they are extremely hard to predict. A method like VBD or DVBD doesn't know this. It just uses the numbers provided as equally likely. Though yes, DVBD should be able to help as it will check the ADP to see how many kickers should go before your next pick, and if it is 0 it'll tell you.But even with DVBD, it still doesn't know how untrustworthy kicker projections are. In the end you need to make a decision based on something, and if the projections hold your best beliefs about players, that's as good a basis to start with as anything. But there are other factors you should incorporate along with the results a method like this gives. Difficulty in projecting something like kicker. The fact that projections are a single number while our beliefs about most players tend to inhabit a range of possibilities with different likelihoods. Two players may both have X points, but you may feel that one is very likely to actually score that, while the other guy will likely either score well above it or well below it but this is what he averages out to. And which you go with then may depend on the rest of your team's situation (are you looking for a solid backup, or a high upside guy to help make up for some weak starters ahead of him?).
 
GregR said:
Liquid Tension said:
the one item that still is odd to me is that even with having all the settings the way I had them and moving kicker to a baseline of only 1, they are still overvalued (from a standpoint of where everyone drafts them). Even intuitively it appears the drop off isn't as large as other positions. Any thoughts on this? Is this where the DVBD makes sense as the kickers may not ever be the choice?
Click to expand...
I think a big part of the reason that people draft kickers so late is that they are extremely hard to predict. A method like VBD or DVBD doesn't know this. It just uses the numbers provided as equally likely. Though yes, DVBD should be able to help as it will check the ADP to see how many kickers should go before your next pick, and if it is 0 it'll tell you.But even with DVBD, it still doesn't know how untrustworthy kicker projections are. In the end you need to make a decision based on something, and if the projections hold your best beliefs about players, that's as good a basis to start with as anything. But there are other factors you should incorporate along with the results a method like this gives. Difficulty in projecting something like kicker. The fact that projections are a single number while our beliefs about most players tend to inhabit a range of possibilities with different likelihoods. Two players may both have X points, but you may feel that one is very likely to actually score that, while the other guy will likely either score well above it or well below it but this is what he averages out to. And which you go with then may depend on the rest of your team's situation (are you looking for a solid backup, or a high upside guy to help make up for some weak starters ahead of him?).
Click to expand...
You make a few good points here. Kickers are the most difficult to predict and the floor/ceiling of players has to be built in somewhere. Jacobs when healthy is a beast, but you know you will miss him for 2 games a year
 
Last edited by a moderator:
mjm3773 said:
I do a custom baseline.We start 8 in my primary league. What I do is take a 3-2-1 weighted average of the total number of players from each position drafted in the first 8 rounds over the past three years.This way, it's customized to my league's drafting tendencies and by not using the "worst starter" method, DD isn't telling me every round from the 8th on that I should be drafting a kicker (advice, which, of course, I ignore).
Click to expand...
I've never drafted exclusively off the draft dominator, in large part because the league I have been in the longest has an unusual scoring system which leads to different positional selections than Joe's formula and I don't think worst starter is helpful as a baseline.I do however maintain draft histories of my leaguemates and use that to understand the likely players available in each round at each position. Of course, this assumes that the players will be drafted in nearly the same order as I rank them at each position...Anyway, back to the baseline. I decided I would look into using the similar 3-2-1 weighted average -- as I already use that exact proportion to forecast this year's position breakdown by round. I also took the idea one step further. Instead of having one baseline for the first eight rounds and then shifting over, as I was doing this in a VBA Excel macro, I simply added two more baseline references. I now have a spreadsheet with four VBDs next to the projected points for each player on my board, one each for what I call the VBD4 round baseline, VBD8 round baseline, VBD12 round baseline, and full draft VBD baseline.In sorting by the different baselines, players naturally shift around. For the most part, late in the draft this is meaningless, but I'm curious what it means -- if anything -- when players trade places based on a VBD4 round baseline and a VBD8 round baseline.Specifically, what seems to happen is that receivers are slightly more valued than running backs at the VBD4 round baseline (and actually at the VBD12 round baseline) than at the VBD8 round baseline. I realize that I am inputting my league's draft history and therefore it is influencing the baseline, but I'm not sure what it means.Should the VBD4 round baseline be used for the first few rounds of the draft before shifting to the VBD8 baseline?Since the league draft tendency is influencing the baseline, does this make the VBD less accurate for my pick purposes but more accurate in terms of understanding when players will be drafted by my opponents?I am thinking around in circles and would appreciate anyone who can help me with an outside perspective.Thanks.
 
Last edited by a moderator:
A simpler version of my post above:

Instead of using one intermediate baseline (through 8 rounds for some, pick #111 for Philo), is there any benefit to having two or even three intermediate baselines (say through 4 rounds and through 8 rounds, or pick #55 and #111)?

And, what does it mean when players rank differently under the two baselines? There will not be a huge shift, but the interpositional value will likely be affected by changing the baselines.

 
An indirect answer to the question I posed, from GregR in another thread, Joe's Secret Formula.

Another poster used league draft history to generate a baseline, and GregR responded in part with:

GregR said:
That's an interesting thing to do, and quite possibly worth while. But I just wanted to point out one thing which is kind of important and might get overlooked by others.

Doing that creates what amounts to a customized ADP for your league. Which is something that could be very useful. However, that result may not be the same as what a given owner should draft from.

As an example, my leagues have formats where QB, WR, and TE are equivalent in value to RB (starting more of each other position and having PPR that favors TE and WR). However, many owners who learned the stud RB rule without learning why continue to overdraft RBs though it isn't the right thing to do. Using the method you mentioned would be great to predict who other owners would take. But since they aren't figuring out value properly and aren't drafting properly as a result, it isn't something I should ever use to guide my own draft decision other than as an ADP list of what the other owners will do.

So yes, very useful. But people should realize too that isn't necessarily what they want to draft from themselves. They should set baselines that they think are correct for their league, not what the rest of the league things. Unless they think the rest of the league is right in which case the two are the same.
Click to expand...
So it seems like having the league draft history set up the baseline gives you a great idea of how the draft will unfold and how value among positions should break down, but as GregR said, unless you think exactly the same way as the rest of the league, you should not necessarily draft too tightly to the draft dominator's player pool suggestions.I'm now thinking of generating a second baseline on what I think should be the appropriate number of players drafted at two points in the draft just to give me a contrast and literally a draft list as opposed to a customized league ADP list.

 
The Jerk said:
An indirect answer to the question I posed, from GregR in another thread, Joe's Secret Formula.

Another poster used league draft history to generate a baseline, and GregR responded in part with:

GregR said:
That's an interesting thing to do, and quite possibly worth while. But I just wanted to point out one thing which is kind of important and might get overlooked by others.

Doing that creates what amounts to a customized ADP for your league. Which is something that could be very useful. However, that result may not be the same as what a given owner should draft from.

As an example, my leagues have formats where QB, WR, and TE are equivalent in value to RB (starting more of each other position and having PPR that favors TE and WR). However, many owners who learned the stud RB rule without learning why continue to overdraft RBs though it isn't the right thing to do. Using the method you mentioned would be great to predict who other owners would take. But since they aren't figuring out value properly and aren't drafting properly as a result, it isn't something I should ever use to guide my own draft decision other than as an ADP list of what the other owners will do.

So yes, very useful. But people should realize too that isn't necessarily what they want to draft from themselves. They should set baselines that they think are correct for their league, not what the rest of the league things. Unless they think the rest of the league is right in which case the two are the same.
Click to expand...
So it seems like having the league draft history set up the baseline gives you a great idea of how the draft will unfold and how value among positions should break down, but as GregR said, unless you think exactly the same way as the rest of the league, you should not necessarily draft too tightly to the draft dominator's player pool suggestions.I'm now thinking of generating a second baseline on what I think should be the appropriate number of players drafted at two points in the draft just to give me a contrast and literally a draft list as opposed to a customized league ADP list.
Click to expand...
:lmao: And to think the first couple of years I came to this site I didn't think a subscription was worth the $$ Boy was I stupid. I learn more about DD every time I come to this site. Draft Dominator alone is worth the price of admission, everything else is just gravy. :thumbup:
 
You must log in or register to reply here.

Similar threads

Ministry of Pain
Tell the Sharks what your team's Biggest Needs are going into the off season and NFL Draft '25
Replies
17
Views
818
dickey moe
dickey moe
T
Travdogg's Predictions for the 2024 NFL season
Replies
25
Views
3K
Moonlight
Moonlight
T
Travdogg's top-150 (8-21 update)
Replies
46
Views
3K
BLOCKED_PUNT
BLOCKED_PUNT

Users who are viewing this thread

Total: 1 (members: 0, guests: 1)
Back
Top