I have been in the same league for 8 years and about 7 of the 10 are originals. I guess you could add tendencies into this theory too. I mean some guys you know are gonna take a QB first and the KC fan is gonna take Gonzo in the 3rd. I think we gave Rob Z alot to answer maybe we should just sit back and wait to hear from him on all of these questions.
Fantasy Football - Footballguys Forums
Welcome to Our Forums. Once you've registered and logged in, you're primed to talk football, among other topics, with the sharpest and most experienced fantasy players on the internet.
ZVBD (1 Viewer)
- Thread starter Rob Zarzycki
- Start date
OK Zolak, I know what value my time has & like most here understand & utilize VBD. But my point wasn't a knock on VBD it was simply that the dynamic application is starting with unquestionably flawed projections/rankings. Therefore, what little benefit that is derived from the application is, IMO, a waste of time. Time, that I would prefer to enjoy while I draft. I love watching the pinheads scramble through magazine cheatsheets while I suck back my 2nd, 3rd, 4th, 5th, ...... Sam Adams.This ain't rocket science. Lets not delude ourselves into thinking that there's a magic potion out there that will assure us of championships. We need to do our homework, plan accordingly and execute our draft plans.
The Z Machine
Footballguy
Here's a thought I just had. Instead of blindly predicting how many (QBs, RBs, WRs, etc.) go in each round, what about incorporating a statistical analysis instead. I.e. in the first round you look at the mean and standard deviation of the number of players selected at each position. This would be based on historical model of drafts with that particular lineup/scoring system. This way, you're not "guessing" how many players go, and instead you have a statistical basis for adjusting the ZVBD baseline. These numbers can be taken from Antsports in the case of the WCOFF format or the traditional 1QB/2RB/3WR/1TE/1D/1K, 10/6, 30/4 league. This seems much better than arbitrary guessing.If you are experienced in mock drafts you already have a good idea of how many positions go in each round, but my suggestino is a little more mathematically sound IMO.
Mungo Burrows
Footballguy
On point (a) I agree. There seems to be some built-in artificial inflation of positions like running back. I can't yet figure out if it is artificially over-inflating the value or whether it reflects the real demands at the position. On point (b) I think this is actually what the system is trying to address. The fact that you need more runningbacks means you will eventually have to dig deeper into that pool thus you project out more rounds. Although I don't remember this in the article, I assume that when you take a player at a position, you decrease the number of rounds out you look at that position. I.e, if your original RB number is 4, and you have one rb, you only look 3 rounds out the next time. That way it incorporates your own demand into the projection. Perhaps Rob can shed light on this, as I am only speculating.
The Z Machine
Footballguy
Yeah but not just a casual look, an actual in depth analysis, round by round, to see what type of distribution you get, so that your positions/round estimates can be that much more exact.
Maybe I'm being picky, but don't you think since FBG decided to publish the article, that it should be a part of the app. I mean, can they honestly think that we're going to do these calculations by hand every round. It seems a bit too much.I mean pay Skippy for his sheet!
Greg R
Footballguy
One thing jumps out at me. In step 4, you predict the entire draft. Or at least, the rounds where you pick starters and maybe primary backups.If you're going to do this, a fairly easy-to-write computer program could check all combinations of players available to you, and find the starting lineup that will score the most points. The best that VBD, DVBD, ZVBD, gut, stud RB, or any other strategy could do is to lead you to pick this combo that scores the most. But you have no assurance any other draft strategies will do that. The only way you can be sure to get that combo is to check all the combos. So if you trust your predictions of that much of the draft to do step 4, you've done the background work necessary to find the absolute best answer possible. So I would find that best answer by letting a computer program check every combination for me. There is no need to do the rest of the ZVBD steps.This isn't really a knock on ZVBD. It's just that ZVBD's requirements give you all the inputs you need for the perfect answer, so you might as well find it.
Last edited by a moderator:
Ditto.I'm all for crunching the numbers. (Am I the only one who wishes they had a Ph.D. in statistics sometimes? I just know I could put it to use for something important -- like my team!)I know there are those that claim luck is responsible for fantasy success. I'll just say that that has not been my experience.
Maybe they've had a different experience.Blutarsky90
Footballguy
I drafted two teams using the average draft info at AntSports and the FBG projections. Lineup reqs were 1/2/3/1. I left off kickers and defense. I only did nine rounds because AntSports only list the top 100 picks (and then there is the fact that I was at work at the time). The projected points are based on the scoring in my league. The ZVBD team looked like this (picks listed in order).1-Marshall Faulk RB 3262-Charlie Garner RB 2513-William Green RB 2314-Laverneus Coles WR 2045-Amani Toomer WR 1966-Rod Gardner WR 1947-Curtis Conway WR 1938-Todd Pinkston WR 1809-Chad Pennington QB 255For the second team, I used straight VBD and a little common sense with regard to tiering. In an actual draft I can't imagine waitng until the sixth round to take my second RB, but that's the way it worked out in this instance. That team came out like this:1-Tomlinson RB 3422-Vick QB 3293-Horn WR 2214-Gonzo TE 1435-Coles WR 2046-Staley RB 1987-Gardner WR 1938-A. Smith RB 1759-Troy Brown WR 188You can draw your own conclusions. I like the way the ZVBD team looks, but if you average the starters over 16 weeks (minus the tight end since the ZVBD team doesn't have one), they would lose 93-89. The difference between Gonzo and the next available TE would widen that gap. Interesting results. Think I'll try it in a mock where I'm not making all the picks. The manual calculations weren't really that big of a deal after the first couple of rounds. Anyway, it filled a couple of hours that would have been otherwise wasted actually earning my salary.
The Z Machine
Footballguy
So you are suggesting that a computer program constantly cycle through and "select" the players it thinks it will draft as the draft occurs in order to find the best route to take, i.e. RB Rd. 1, RB Rd. 2, QB Rd. 3. It almost looks like an AI program to me. A good program should be able to do this pretty easily. This could be the super add-on to the Draft dominator.Personally, I'm waiting for a neural network program that "learns" how to draft and plugs away at the calculations that way. It could be really cool.
Kasparov
Footballguy
While it is true that projecting out more rounds for a position will lower the baseline and raise the value of that position (simulating your demand for that position), but at what cost? If I project out 4 rounds for RBs at the beginning of the draft and only 1.5 rounds for QBs, guess who will be the most valuable? Looks like I'll be drafting a RB in round 1! Which is fine of course, but this just forces the Stud RB theory on your values. What it will probably do during the rest of the draft is pre-determine which position to take in which round. Again, that's fine if that's how you want to approach your draft. In my mind, your teams demand for a position should be based on, among other things, your supply at that position, i.e., I have drafted 2 RBs already and I intend on drafting a total of 5, therefore I have filled 40% of the order for RBs, etc. Something along those lines. Varying the rounds-out you are projecting seems a bit severe to me. But I could be convinced!
I will like to try this system out because i don't want to bring a laptop to my draft(don't have one). For years, as I am sure everyone else does, during I keep track of everyones pick during the draft and if the guys after me all have there starting QB and the ZVBD says to take a QB I'm, as would the rest of us, going to wait until my next pick to take a QB. Plus what about using it for the whole draft and no just for the starters? I just want to hear from Rob Z again to see whta he has to say about all of these great questions in this thread. I'll probably study mocks at antsports and come up with something. Let us all not forget it's Fantasy Football not Rocket Science 
Greg R
Footballguy
Yes, I'm suggesting the computer program would be sure to give you the best answer based on your predictions. VBD, DVBD, ZVBD could at best tie this result. If your prediction was wrong, then yes, you'd recalculate with an updated prediction, and changing your inputs so draft picks you've already used only have the selected player as the choice.But, I have doubts that you can predict that far ahead accurately enough to base decisions on, so I'm not necessarily saying I'd use this method myself. I believe the error in projecting that many rounds ahead is going to be larger than the number of points that you stand to gain by going from a good answer to the perfect answer. Using dynamic VBD last year, I twice predicted I'd hit the end of a tier and was wrong and got dropped to the next tier at my picks. And that was for predicting only 2 rounds ahead (I was picking back to back so had 22 picks before I went again each time). Those errors in my draft predictions were pretty big differences in points, much larger than what I stood to gain from going from a good combo of players to the best combo of players.So I'm saying if you expect to be off by 8 FP (or .5 FPG) in your draft predictions, I don't know that I'd use that prediction as a basis for a decision that gains me 4 FP. But if you are going to make those draft predictions and not improve them as the draft goes on anyway, it would seem to me that you might as well go all the way and go for the best combo those predictions suggest.
Rob Zarzycki
Footballguy
Please forgive me ladies and gentlemen if I do not reply in a few weeks, I will be in San Diego as part of my personal World Poker Tour. My other hobby is texas hold 'em poker. --Z
Where are you going to play in San Diego? I played at Viejas last night and the poker room was packed.
Hobbes
Footballguy
I'm not sure what the exact inner-workings are of DVBD...but the thing that I like about the proposed ZVBD is that it not only takes into account what the people around you will be drafting, but it also changes based on what you have already drafted...I have not seen another system that takes this into account, just ones that tell you what the best value remaining is.The thing that I don't like is the arbitrary 1.5 * starters multiplier. I think it would make more sense if this number was the "target" number of players that you want to draft at each position. So in a 15 round draft where you start 1,2,3,1,1,1 you may want 2,4,5,2,1,1 or something like that...then if you use these numbers as your multipliers, you are projecting out the number of rounds equal to the number of that position that you want less the number you have already drafted. It still takes into account the VBD and the projected draft positions...but it would allow you to be more flexible about the number of each position that you want to draft. 
Any thoughts?
Any thoughts?
Hobbes
Footballguy
Tad,I'm with you on this one. I don't get it
. It seems to me that your baseline should be based on the number of players picked between your picks. Adding the rounds seems to be artificially inflating the baseline numbers (which is going to be in favor of the RB) when it's not really necessary (IMO). Has anyone done any analysis to see if this step is really necessary?[The thing that I don't like is the arbitrary 1.5 * starters multiplier. I think it would make more sense if this number was the "target" number of players that you want to draft at each position. So in a 15 round draft where you start 1,2,3,1,1,1 you may want 2,4,5,2,1,1 or something like that...then if you use these numbers as your multipliers, you are projecting out the number of rounds equal to the number of that position that you want less the number you have already drafted.]I agree, there should be a multiplier for the total of positions draft some how. I am going to try and play with the multiplier. Being that if there are 10 teams and drafting 2 QB's per thats 20 QB's total. How can we incorporate that into our moving baseline, or does it not matter because the baseline will never reach over the #20 QB. any thoughts?
MrJimiT
Footballguy
I just began a mini-mock with all ten teams using ZVBD and the same projections list. With no intuitive interjection, I simply took the highest value number that the ZVBD suggested to draft. I am 40 picks into it, and not 1 QB has been drafted. Maybe it should be that way, but in my league, I am positive more QB's will be drafted long before pick #40. (I used my previous two years of league drafts for a draft list).There are two concerns that I have:#1 - The 1.5 multiplier used as the B-value in ZVBD. In my setup, a QB will never get drafted before 20 or so RB's have been drafted.#2 - The turnaround picks (1,2,3,8,9,10) with combining two-round results. The difference in baselines from pick #3 to #4 and pick #7 to #8 were tremendous. I'm not comfortable yet with this calculation.I believe it was truly useful to go through the steps for I have a real good grasp on my league's drafting tendancies. However, my 10 team mock with all teams using ZVBD, did not resemble my league's drafting tendancies whatsoever. Maybe this is a good thing, but I still have some questions about the aforementioned calculations.
I'm with you Mr. JimiI reviewed a couple of expert mocks to see how I would do at the 5th draft position and I didn't end up with a QB until the very late. I also ended up very thin at RB.Subtracting one from the B-value after drafting a position can also make you pass up on what I felt were good picks.I think there is something too this, but I believe a hybrid between the Z and the basic VBD is closer to my comfort level.
I'm not sold on it. You are depending on 100% predictability on what all the other owners will do.The VBD alone is all you need. You just do minor adj. during the draftbased on what/how others are drafting. You have the proj. ptscalculated and draft the player who should get you the most pts.from the position you have open. Drafting an extra rb or wr before a TE,K,Def seems best to me as depth means alot.There's not much more you need to put into extra charts and theories.I finish 1-2-3 alot.
Hobbes
Footballguy
I disagree...I think the fact that you include projections on draft order and that you consider positions that you have already taken potentially would make this strategy stronger than normal VBD.With VBD, you are depending 100% on what you think the players will do....this is also flawed. But just because you can not make an exact prediction does not mean that it is wrong...just that there is risk. I think that making an educated guess at something that you do not know for sure is far superior to leaving it out of the analysis all together.
I had a question about the original article:In the example #1, you ask for the baseline number for QB in the second round (3) but when you ask who the baseline QB player is, you use the first round draft sheet and use Manning as your baseline player? Shouldn't your baseline player for the first round be McNabb b/c your baseline # would be 2? Sorry if this sounds confusing.I've been making my own spreadsheet with some projections and I'm also confused about the 1.5 multiple. I dont' really have an answer, but I have yet to hear a convincing arguement. It seems like most people aren't picking QB's with this system until around the 7th round or so; which I don't mind, b/c it is what I do normally. I really like the theory behind this, but I'm not quite sure if/how to apply it yet.
Rob Zarzycki
Footballguy
I'm playing in a tournament at Suycan reservation I think on 15th and 16th. I'm leaving in a couple days.Getting back to fantasy football, THE ZVBD DOESN'T VALUE QUARTERBACKS THAT HIGH AT ALL. RUNNINGBACKS ALSO LOSE VALUE RELATIVE TO WIDE RECEIVERS (COMPARED TO NORMAL STYLE). THIS MAY SEEM DIFFERENT FROM YOUR NORM BUT THAT'S MY METHOD. LOVE IT OR HATE IT. I HAVE MORE ARTICLES COMING OUT SUPPORTING THIS STRATEGY.--Z Factor
Rob Z, I have been playing with your system for a few days now. I am using mock drafts to to average out the position picks per round. I do have the 5th pick this year and that is considering your example. My question is, when going through the mocks you skip over the position picked at my slot, if so I would be missing a position taken in that round for my set up of your D values. Is this correct or does it not really mater since that would be my pick supposedly in the mock anyway. I hope it is not a confusing question. Again I am using already finished mock draft to set up the D value table.
shffl,I don't want to answer for Z, but I had this problem at first as well. According to Ron, you would skip your pick, hence the 10/8 alternating rounds in his example. Since there are 10 teams, Ron's other 2 picks account for the "missing" 2 picks.Also, how do you feel about the ".5's" for the B numbers?
Tick
Footballguy
A fundamental question for DVBD or ZVBD for me is one I've been unable to answer so far:
Q) When projecting what other teams will pick, should I use the average number of each position taken between my draft picks, or should I use the average number of each position taken up to my next draft pick?
To elaborate, if the draft isn't going the way I expected it to go, should I adapt my projections of how many players will be picked between my picks?
Let's say that before my third round pick, I have projected that four QBs will have been taken. Between my third and fourth round pick, I project that two QBs will be taken. However, in the actual draft, when my third round pick comes up, zero QBs have been taken. Do I keep my projections for the number of QBs to be taken before my fourth round pick at two, or do I make it six?
Q) When projecting what other teams will pick, should I use the average number of each position taken between my draft picks, or should I use the average number of each position taken up to my next draft pick?
To elaborate, if the draft isn't going the way I expected it to go, should I adapt my projections of how many players will be picked between my picks?
Let's say that before my third round pick, I have projected that four QBs will have been taken. Between my third and fourth round pick, I project that two QBs will be taken. However, in the actual draft, when my third round pick comes up, zero QBs have been taken. Do I keep my projections for the number of QBs to be taken before my fourth round pick at two, or do I make it six?
Ok,I'm going to try to explain this the least confusing way possible. I created a spreadsheet (that is not complete yet) to help me understand this. I entered all of the values that you gave in your example. Originally, I had 2 QB's project in each of the first two rounds. Because none were taken, I replaced each 2 with a 0. Now, I needed to add 2 more positions to each of those rounds in order for the totals to match. So for the sake of making it easy, I added 1 WR and 1 RB each round.
Now, I had to take those 4 QB's and added that to the 2 projected between rounds 3 & 4. But I also had to remove 2 WR & 2 RB from that same round so the totals would match. It does change the baseline player for each position, but it seems like it should work if you felt like doing this for the entire draft. I don't have the time to test it on an ANT draft right now, but hopefully next week.
So to answer your question, I would bump the QB projections to 6 and decrease the positions that were drafted in place of the QB's in the earlier rounds.
Great explaination Tiger Fan. I believe you have got it. I ask Rob Z a similar question on the first page of this thread. He answered in the quote above. But I think your explaination is what he was getting at. I think there would be enough time to do this at the draft if it was practiced ahead of time. Nice job Tiger Fan.
Like I said, I'm working on a spreadsheet; but I'm trying to gear up for the MBSL II draft also; so after I prepare for that, I'll test out the ZVBD with an ANTSPORTS mock and let you guys know the result. I just don't think I'll be able to use it with my early serious drafts b/c I'm not sure about the time restraints. I still can't find a good answer for the .5's on the B numbers???? Any help?
I'm sorry to keep bring this thread back up, but the system intrigues me. The owners in my have been pretty much the same fro the last 3 years or so. I have my commish giving me the draft results of the last 3 seasons and I am going to use those results to create my postion drafted chart. This way I also have the other owners tendencies in the results as well. I think it will give me good baselines. I took last years and made up a chart and I'm trying it on Antsports. I am following Rob Z's instructions to the word. I will keep all of you posted. Thats if anyone cares
Brock Middlebrook
Footballguy
you're welcome (see the sig)
Sam Quentin
Footballguy
he didn't have to call it "zvbd." that was taking it too far -- that was rob z making it all about him, when the it should have been about Brock (primarily) and GregR and Mungo and Kasparov etc..
Something interesting from the MBSL 1 Survivor League draft Link to Round 1. I decided to act like I had the 6th pick in the draft and strictly use the zVBD with the readjustments that I posted above to draft the majority of my team.The VBD looks like this for the first round.I'm sorry to keep bring this thread back up, but the system intrigues me. The owners in my have been pretty much the same fro the last 3 years or so. I have my commish giving me the draft results of the last 3 seasons and I am going to use those results to create my postion drafted chart. This way I also have the other owners tendencies in the results as well. I think it will give me good baselines. I took last years and made up a chart and I'm trying it on Antsports. I am following Rob Z's instructions to the word. I will keep all of you posted. Thats if anyone cares
1. LT 252
2. Faulk 238
3. Holmes 213
4. Williams 212
5. Portis 209
6. Duece 197
7. Henry 196
8. Marvin 192
When the 6th pick came up, I ran the calculations and got this:
QB RB WR
Baseline Player Vick Zereoue Rice
Baseline Points 393 207 222
Best Player Culp. Holmes Harrison
Best Points 413 339 345
Difference 20 132 123
So using zVBD, I "selected" Holmes.
[Edited b/c I used the overall column originally to get the RB baseline player. Sorry]
Last edited by a moderator:
bueno
In a class by himself
This is an idea worth following through on. If I've already drafted 2 rbs, then the value of RB drops relative to the value of other positions where I don't have my starter.One thing I don't like about DVBD and ZVBD is the inaccuracy of guessing how many will be taken at each position. I agree that it is easier to predict a range where players may be taken, even develop a probability distribution for picks. I would be more tempted to use published cheatsheets and mock ADPs in concert though. Even the Antsports SMDs are compromised by computer selections. The cheatsheets do not accurately reflect real drafts either.Before someone jumps all over me about Antsports, let me explain: look at the computer selections in an SMD draft-by-draft. You'll see the same players are always taken - the ones Steve's programs are over-valuing. Until Steve develops a program that ties his computer selections to his ADP, that will likely always be the case. Therefore, if you have six computer selections, a player could conceivably be taken and half round later in the mocks than he would be taken in a "real" draft. Given the precision we are looking to acquire here, that isn't acceptable.Now, the comment about cheatsheets - I'm seeing RBs go way earlier than they ought to if you use SVBD and Joe's cheatsheets. If I calculate an X value for PKs and apply it blindly, they would go much early than they are too. Therefore, a straight mock built around the cheatsheets will not reflect reality either.You are therefore left with a large degree of subjectivity when you use either DVBD or ZVBD. You have to get the projections fairly accurate, then you have to get the drfat fairly accurate. Projections I can deal with - predicting the draft is a lot tougher.
I agree with almost everything you are saying (see my above posts.) If you have accurate historical data of your league, a possible solution would be to have the multiplyers reflect the average number of players drafted at each position per team.For example, over the past 2 years the teams in my league select on average, rounded to nearest .25 (zVBD multiplyers in parenthesis):2.5 QB (1.5)4.0 RB (3)5.0 WR (4.5)2.0 TE (1.5)2.0 K (1.5)1.75 DEF (1.5)Surprisingly, they are somewhat similar; but I would think that my numbers would be more accurate given this relatively untested theory.
groomsie
Footballguy
shffl
Actually, if you read the specifics of Steve's Antsports drafts, his average draft position does NOT factor in computer selections. Supposedly, if the mock drafter did not actually draft (or pre-draft) the player, then the player will not count in the ADP.
Here is the quote: "Computer Selections have never been included in this info. I exclude any computer selection. That's why the number of drafts a player in varies so much. If the computer selects Faulk often, he will have a smaller number of Drafts he is included in vs another player. "
ADP Front Page
I can't say if this is really the case, but that is what da man sez.
Edited to add quote from ADP site and link to site.
Last edited by a moderator:
groomsie
Footballguy
I am interested to see if anyone is using this method and what their comments are.
I am using the ZVBD in an antsports mock draft (SMD 680 - Team is "Clone Me"). I have not been correcting for actual picks made (I'll do this next time, as I do think this will make a difference). So far I am not sold.
The reason I am not sold is that for picks 1-3 it made no difference in my selections. This might have been expected. In round 4 it resulted in me going to TE when I'd likley have gone WR. So far I have:
1.09 RB - Henry
2.02 WR - Moss
3.09 QB - Culpepper
4.02 TE - Shockey
5.09 RB - Mack
6.02 WR - L.Coles
7.09 WR - Conway
8.02 Def - Eagles
I do believe that correcting as you go would make a difference in selections, but I need to reset my spreadsheet. My observations (Zarzycki, can you comment?):
1. I am not sure there is enough difference to justify the extra hassle with evaluating baselines for the average league. If you are in league full of sharks (who will be-round to round-reasonably predictable in selections) it may help. I may change my tune after trying again with updated actual draft tendencies...
2. Many times the baseline is 1, which essentially means you wont select a player (ex: RB baseline is = 1, then if Bennett is the top player available with 150 projected points, 150 - 150 = 0 = no selection). This sometimes forces overlooking value.
3. As with any system, this is only a guide in my opinion. I took Mack in round 5 despite the fact the ADP indicated I could have drafted him later. But, he was the best player on my board. But, if I'm going to only use this as a guide, why hassle with the baselines? I could do the same thing with my VBD sheet and my ADP sheet (ie: make the judgement call on when to pull the trigger).
I am using the ZVBD in an antsports mock draft (SMD 680 - Team is "Clone Me"). I have not been correcting for actual picks made (I'll do this next time, as I do think this will make a difference). So far I am not sold.
The reason I am not sold is that for picks 1-3 it made no difference in my selections. This might have been expected. In round 4 it resulted in me going to TE when I'd likley have gone WR. So far I have:
1.09 RB - Henry
2.02 WR - Moss
3.09 QB - Culpepper
4.02 TE - Shockey
5.09 RB - Mack
6.02 WR - L.Coles
7.09 WR - Conway
8.02 Def - Eagles
I do believe that correcting as you go would make a difference in selections, but I need to reset my spreadsheet. My observations (Zarzycki, can you comment?):
1. I am not sure there is enough difference to justify the extra hassle with evaluating baselines for the average league. If you are in league full of sharks (who will be-round to round-reasonably predictable in selections) it may help. I may change my tune after trying again with updated actual draft tendencies...
2. Many times the baseline is 1, which essentially means you wont select a player (ex: RB baseline is = 1, then if Bennett is the top player available with 150 projected points, 150 - 150 = 0 = no selection). This sometimes forces overlooking value.
3. As with any system, this is only a guide in my opinion. I took Mack in round 5 despite the fact the ADP indicated I could have drafted him later. But, he was the best player on my board. But, if I'm going to only use this as a guide, why hassle with the baselines? I could do the same thing with my VBD sheet and my ADP sheet (ie: make the judgement call on when to pull the trigger).
Saint
Footballguy
First of all. This is one of the most interesting articles I have read. Many Thanks to all the contributors and to the creators. I am working on taking this to the next level. Rob Z I will need your help on this if you have some time!
I have taken the FBG VBD spreadsheet and built a ZVBD spreadsheet on top of it. So the way it works, is that the FBG VBD spreadsheet is used to predict scoring and overall fantasy points. Once run, the ZVBD spreadsheet is opened. The starting positions and flex values are automatically pulled from the ZVBD spreadsheet and prepopulated along with the VBD fantasy point values.
As the user, you are required to enter the following items:
* Draft Position
* If Flex available, identify which positions can be flexed (For Survivor drafts, you can flex QB, PK, D)
* Round Number
* Estimated Projections for League by Round (I only have 16 rounds to date)
During the draft, you must do the following items:
* Mark on the draft sheet which player is taken (Easy, just put a character in the field)
* Update baseline projections (This is more difficult, I am in the process of working this out to be automatic based on the estimated projections given at the beginning.... Statistical analysis here is a good thing)
* Identify which position you have drafted
* Click the [Value] Button
A little time consuming, but takes about 30 seconds to do. When the button is clicked, the spreadsheet performs all the baseline calculations based on the round and gives you the best value position. It DOES account for all of the items Rob has mentioned in his article.
Currently, I am in the process of validating calculations and tweaking it. Hopefully, it'll pan out. I did one mock Survivor draft and here was the results:
1.11: Ahman Green
2.02: M. Harrison
3.11: S. Mack
4.02: J. Garcia
5.11: A. Thomas
6.02: C. Conway
7.11: M. Alstott
8.02: Eagles-D
9.11: J. Thrash
10.02: D. Jolley
11.11: J. Fiedler
12.02: Kevin Johnson
13.11: J. Kitna
14.02: R. Longwell
15.11: J. Chandler
16.02: Jets - D
Scoring:
Passing:
1pt / 25 yards
4pt / TD
2pt / 2 pnt conv
-2pt / Int
Rushing:
1pt / 10 yards
6pt / TD
2pt / 2 pnt conv
-2pt / Fumble Receiving:
1pt / 10 yards
6pt / TD
2pt / 2 pnt conv
-2pt / Fumble Kicking:
1pt / XP
3pt / FG Defenses:
1pt / Sack
2pt / INT
2pt / Fumble Recovery
4pt / Safety
6pt/ TD
I was very surprised by some of the value at the positions... Mainly at taking S. Mack in the 3rd round. In this first example I did not update the baseline "guesstimated" positions for the draft. I am making tweaks based on this draft and I will test again on my next draft. In addition, I will post the results after the draft for review. I should be drafting again on Thursday night to try this out.
Sorry for the long post, but I think it's been some interesting discovers into the world of ZVBD.
I have taken the FBG VBD spreadsheet and built a ZVBD spreadsheet on top of it. So the way it works, is that the FBG VBD spreadsheet is used to predict scoring and overall fantasy points. Once run, the ZVBD spreadsheet is opened. The starting positions and flex values are automatically pulled from the ZVBD spreadsheet and prepopulated along with the VBD fantasy point values.
As the user, you are required to enter the following items:
* Draft Position
* If Flex available, identify which positions can be flexed (For Survivor drafts, you can flex QB, PK, D)
* Round Number
* Estimated Projections for League by Round (I only have 16 rounds to date)
During the draft, you must do the following items:
* Mark on the draft sheet which player is taken (Easy, just put a character in the field)
* Update baseline projections (This is more difficult, I am in the process of working this out to be automatic based on the estimated projections given at the beginning.... Statistical analysis here is a good thing)
* Identify which position you have drafted
* Click the [Value] Button
A little time consuming, but takes about 30 seconds to do. When the button is clicked, the spreadsheet performs all the baseline calculations based on the round and gives you the best value position. It DOES account for all of the items Rob has mentioned in his article.
Currently, I am in the process of validating calculations and tweaking it. Hopefully, it'll pan out. I did one mock Survivor draft and here was the results:
1.11: Ahman Green
2.02: M. Harrison
3.11: S. Mack
4.02: J. Garcia
5.11: A. Thomas
6.02: C. Conway
7.11: M. Alstott
8.02: Eagles-D
9.11: J. Thrash
10.02: D. Jolley
11.11: J. Fiedler
12.02: Kevin Johnson
13.11: J. Kitna
14.02: R. Longwell
15.11: J. Chandler
16.02: Jets - D
Scoring:
Passing:
1pt / 25 yards
4pt / TD
2pt / 2 pnt conv
-2pt / Int
Rushing:
1pt / 10 yards
6pt / TD
2pt / 2 pnt conv
-2pt / Fumble Receiving:
1pt / 10 yards
6pt / TD
2pt / 2 pnt conv
-2pt / Fumble Kicking:
1pt / XP
3pt / FG Defenses:
1pt / Sack
2pt / INT
2pt / Fumble Recovery
4pt / Safety
6pt/ TD
I was very surprised by some of the value at the positions... Mainly at taking S. Mack in the 3rd round. In this first example I did not update the baseline "guesstimated" positions for the draft. I am making tweaks based on this draft and I will test again on my next draft. In addition, I will post the results after the draft for review. I should be drafting again on Thursday night to try this out.
Sorry for the long post, but I think it's been some interesting discovers into the world of ZVBD.
Last edited by a moderator:
Hobbes
Footballguy
I have used the zVBD in a couple of antsports drafts and there is definitely some tweaking that I have had to do for it to make sense for my drafting style. First, I would suggest that instead of using an arbitrary 1.5 * starters multiplier that you use different multipliers for each part of your draft.For instance, if you know that you want to take 3RB, 3WR, and 1QB before you take anything else. Then use these numbers as your multiplier. Then you can let the zVBD determine the order that those 7 players are chosen. After those 7 rounds, you could then change the multipliers to fill out the rest of your team.The problem that I had with only using one set of multipliers was that it kept telling me to take a running back...like 5 rounds in a row. Well, I took the RB with my first 3 picks...but it is hard to justify taking a RB like A. Smith in the 4th round when you already have 3 running backs and there are WRs the caliber of P. Burress still on the board. For informational purposes, my multipliers to start were: 2QB, 5RB, 4WR, 2TE, 2DT, 1K. These numbers represented the number of players that I wanted on my roster at the end of the draft. What this doesn't take into account is that I won't touch my second QB, tight end, or either my kicker or DT until I have atleast 3-4RB and 2-3WR or that I want to fill out my starting 2 wide outs before picking my 2nd backup RB.Any ideas on a better way to tweak this system for your own drafting preferences?
