Rob Zarzycki
Footballguy
Okay guys and gals, what do you like or dislike about my $22,000 system. Is it good, bad, .... give it to me baby! give it to me!!! -- Z man
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
Rob Zarzycki
Footballguy
I felt more comfortable rounding down, that's all. If you round up it may overvalue those positions, but you can do it if you feel those positions deserve the extra strength.
Blutarsky90
Footballguy
Shouldn't the position number in Table B and Table C be the same? They are except for rounds 4 and 5.
yes and no...if you have an auction league, these rules can still apply...you just need to spin it a bit...baseline would be factored towards a dollar value, accordingly each players 'value' changes and players come off the board... in my opinion, it is quite possible, but I am far too lazy to try build a macro where this could be done...in theory it could be a simple as projected total points per player at a position, then some 'Q' coefficient, that increases players dollar value, as other players are removed from the board... every player technically gets more valuable, to teams lacking palyers at that position, as players from his postion are 'bought'...the key is determing that 'Q' value...dynamic based on your league..but possibly predictable based on keepers (maybe, many sallary keepers are anomolies) or prior season...I guess this only matters if youre in a league that does a draft.
Last edited by a moderator:
Loss By Design
Footballguy
I do have a few questions, not about the technique, but about the results. How does this method stack up against straight VBD, or DVBD? Have you ever logged each pick and what the suggested VBD and DVBD picks would have been? If it doesn't deviate very much from VBD, and especially DVBD, then might it be overkill?I'm all in favor of new (more complex) ways to make the best selection, but one thing I noticed that was lacking from your article was a definition of why it works. It would be nice to see an editorial about why it works, instead of strictly how it works. I suspect that players serious enough to attempt using ZVBD will not want to be led blindly down the draft path. It would be nice to know what considerations went into the decision to use 1.5 rounds ahead, as opposed to 1 or 2.
This system could work but the only thing I don't like is you have to guess how people are going to draft. Predicting where a player might go in the draft is a little more realistic. There is always one or two boneheads tht could through this hole system off during the draft. then you are left scrambling. In the survivor draft here at footballguys I think a QB wasn't taken until late round 2 or even round 3. I just think its hard to prdict what 9 or even 12 other guys are going to do. The idea is good though I just might take it and modify it some way. I will post what I come up with for argument 
Iwannabeacowboybaby!
Footballguy
I think it's pretty complex but it looks like it could work IF your projections are pretty good. If you rank Trent Greene over Drew Bledsoe and Greene flops, no matter what system you use you are in trouble.I know you won your competition using this system and congrats. You must have done a great job in doing your projections last season. I just wonder if you would have used other quality systems with your same projections and gut instincts would you have won also. My guess is yes but who knows for sure. I too have won various leagues and made good picks in drafts but have used different systems than yours. I guess my question is do you really think this system is what won you that competition or was it your quality of projections? I ask that with no sarcasm intended, just want to know what you think.
Last edited by a moderator:
The Z Machine
Footballguy
Rob, if you like, I will be the spokesman for your VBD theory. I feel that my screen name is more marketable than "Rob Zarzycki".On a serious note, I feel there should be a VBD showdown as alluded to before where the different theories are stacked up against each other. I'm not sure exactly how this should be done, but it could be kinda interesting. Also, is there a way to incorporate this into the existing FBG VBD apps?
NE_REVIVAL
Footballguy
I mean no disrespect but I believe properly ranking players and luck have a lot more to do with winning than any complex systems.Would love to see the results of 1 (or 2) 12 team league(s) where 6 teams are built using ZVBD (or VBD) and 6 are built the old fashioned way (by gut). Count me in as one of the 6 non believers if someone decides to set up such a challenge league(s). I would like nothing more than to be convinced that there really is a system that works.
I just feel that predicting how a player will do in a season (which I do do my own predictions) is easier that predict how a draft will go. If there was a way to modify the baseline during the draft as players are drafted that would help in picking the right valued player. Thats the idea his system gave me.
Camenthead
Footballguy
Z.Since you are the only one who has actually used this model -- did you adjust your expectations of future draft picks as a result of actual picks in earlier rounds? or did you keep predraft assumptions the same throughout the draft?
Skippy the one eyed Beagle
Footballguy
OK. I just read the column and apparently I already do this (or something very close to this) but I simply called it Dynamic VBD.
If anyone is interested in an application that basically does this check out my app at:
http://www.thecoordinator.com/
There is a fully functioning DEMO that allows you to draft 1 round.
If anyone is interested in an application that basically does this check out my app at:
http://www.thecoordinator.com/
There is a fully functioning DEMO that allows you to draft 1 round.
Rob Zarzycki
Footballguy
I did compare it to the traditional VBD, especially since that's the system I used before I developed the ZVBD. I did like 50 comparisons, each in a different draft using different variations, and that's how I developed it to where it stands today. The ZVBD should always give you more value because it's a dynamic version of VBD. (loss by design) I didn't have to change my future draft pick expectations b/c I was fortunate enough to have a league that drafted relatively normal. But even with "boneheads" you'll find the average number of players drafted from each position to remain quite consistent (camenthead)I am in discussion with David Dodds on integrating ZVBD into the VBD App but I think he has his own dynamic system somewhere up his sleeve. By the way, love your name (Z machine)If 5 QBs went intead of 2, you would want to adjust the other rounds, so yes is the answer. In a simpler example, let's say you had no TEs going in the first round, 1 in the second round (Gonzo), and none in the third round, but then notice someone takes Gonzo in the first round, then you should adjust for that by removing the 1 in the second round. It gets very complicated, I'm working on something accounting for unusual streaks of drafting players. It would be like a dynamic dynamic ZVBD. (shffl)The Draft Dominator isn't dynamic, maybe it will be this year. I haven't looked at it yet. DYNAMIC is the key fellas -- for those of you who aren't familiar, that means the baselines change as the draft proceeds.Good work everyone, appreciate the comments so far.
jplvr
Footballguy
That one beats mine: 2 years ago, Jamal Lewis gets injured and prior to the draft someone says "I wonder who'll draft him since he's injured." There was a big discussion about it. We're going around the room announcing players for bidding and a guy who was sitting right by the discussion calls out "Jamal Lewis 15 points." Dead silence. The commish says, "I think you got him buddy." Celebration ensues/laughter soon follows. The running gag for the rest of the draft was: "Jamal Lewis... already taken?!?!"
17seconds
root of all aliai
After reading through ZVBD, it appears as though it allows you to have greater confidence in picking a position based on your roster rules, and some level of insight into how the draft will go. To me, this is a great idea since these are the main things that can bite you if you just blindly use VBD.Also, I trust FBG. They spend a lot more time on projections than I would ever want to spend, so I just use their projections. Call me lazy, but other than tweaking a few players projections myself, I don't touch any of it. ZVBD could be very helpful to me, but probably only if there is an app I can run during the draft. As commissioner of my league it's hard enough recording all the picks, and crossing them out on my own sheet to do ZVBD calculations too!!
Mungo Burrows
Footballguy
We want the theory behind this!
Ok, let's see what we can make of this...
You project out players taken per position per round from mocks/etc.
Did you consider aiming for target players taken by a certain point instead of just players per round? E.g., My spreadsheet says to expect 8 QBs to be taken by Round X therfore my baseline is the #8 QB. This might keep you more on track if there are unexpected runs or a position is unexpectedly neglected, because don't these exceptions usually correct themselves? It doesn't matter that 3 more QBs were taken in round 2 than you expected because between rounds 2 and 4 there will probably be 3 less Qbs taken to compensate. Or is that not the case?....interesting question: When a draft deviates from the norm at a position, does it tend to correct itself quickly, or does it remain out of whack for a while.
You take the number of players you start at the position and multiply by 1.5
why 1.5? Is this to simulate the number of "likely to start at some point" players at the position (this is my assumption)? or is it just some fudge factor? Where did this come from?
Then you set your baseline at the player who would be left at (starters*1.5) rounds from now.
My take (read: guess) is you're calculating "what's the best I can hope for among players I might have to start at this position if I don't take one now." E.g., My RB number is 4 - To get the 4 RBs I would likely have to start at some point this season I obviously need to draft them in 4 rounds. I could take all 4 RBs for 4 straight rounds starting right now for the highest possible combined projections at that position(obviously neglecting other positions). If I don't take someone at that position now, I will have to get someone in the N+4th round or later. Looking four rounds out tells me who the player I will have in my lineup would be if I don't take the top player at that position now. Is this the theory behind what you're trying to calculate? That seems to be the best rationalization I could come up with.
Am I close?
Ok, let's see what we can make of this...
You project out players taken per position per round from mocks/etc.
Did you consider aiming for target players taken by a certain point instead of just players per round? E.g., My spreadsheet says to expect 8 QBs to be taken by Round X therfore my baseline is the #8 QB. This might keep you more on track if there are unexpected runs or a position is unexpectedly neglected, because don't these exceptions usually correct themselves? It doesn't matter that 3 more QBs were taken in round 2 than you expected because between rounds 2 and 4 there will probably be 3 less Qbs taken to compensate. Or is that not the case?....interesting question: When a draft deviates from the norm at a position, does it tend to correct itself quickly, or does it remain out of whack for a while.
You take the number of players you start at the position and multiply by 1.5
why 1.5? Is this to simulate the number of "likely to start at some point" players at the position (this is my assumption)? or is it just some fudge factor? Where did this come from?
Then you set your baseline at the player who would be left at (starters*1.5) rounds from now.
My take (read: guess) is you're calculating "what's the best I can hope for among players I might have to start at this position if I don't take one now." E.g., My RB number is 4 - To get the 4 RBs I would likely have to start at some point this season I obviously need to draft them in 4 rounds. I could take all 4 RBs for 4 straight rounds starting right now for the highest possible combined projections at that position(obviously neglecting other positions). If I don't take someone at that position now, I will have to get someone in the N+4th round or later. Looking four rounds out tells me who the player I will have in my lineup would be if I don't take the top player at that position now. Is this the theory behind what you're trying to calculate? That seems to be the best rationalization I could come up with.
Am I close?
MrJimiT
Footballguy
Probably for a different topic/thread...you should try to get a league secretary for draft night (wife, girlfriend, significant other) so you can put all of your marbles towards your own team. I'd think you would be at a severe disadvantage otherwise.As commissioner of my league it's hard enough recording all the picks, and crossing them out on my own sheet to do ZVBD calculations too!!
I'm uncomfortable agreeing with another Pats fan (don't want to look like a smoke-blowing homer) but I think we're adding complexity where it really doesn't need to be added. As Pat-Patriot noted, projections & rankings are far more important a draft-day formula. If you haven't projected/ranked well, any and all draft systems will fail.
Rob,I read your ZVBD article and thought it was great. Last year, I hand calculated total projected points (Marvin Harrison + Projected RB I would get in next round vs. Top RB + WR I would get) for each pick, and dominated. The ZVBD basically creates a system to do this. The one thing I don't understand is adding rounds for the first 3 or last 3 picks. Since you are estimating the total players of each position taken between your picks, doesn't this already take draft position into accout? If you are pick #2, you have 20 picks between round 1 and 2 and only 2 picks between rounds 2 and 3. I don't understand why you are adding rounds to adjust.Thanks,Tad
Nemesis
Footballguy
It's hard enough to keep track of the draft by:
writing down the picks of the draft and keeping it progressing smoothly (as a commish),
crossing out the names on my cheatsheet and depth charts,
circling my picks,
writing down the current rosters,
ensuring my bye weeks align with the stars,
discussing strategy with my co-owner,
tallying how many QBs, etc....have been taken so I decide when to pick one,
talking smack and making fun of other owner's picks,......
that the last thing I want is to have to constantly update some formulas while my fellow owners are yelling at me to Hurry up and Pick!
The draft is stressful enough that I'd like to at least leave some room to enjoy my "Christmas Morning".
Call me crazy (YOU'RE CRAZY! :wacko: ),.....but thanks, no thanks.
writing down the picks of the draft and keeping it progressing smoothly (as a commish),
crossing out the names on my cheatsheet and depth charts,
circling my picks,
writing down the current rosters,
ensuring my bye weeks align with the stars,
discussing strategy with my co-owner,
tallying how many QBs, etc....have been taken so I decide when to pick one,
talking smack and making fun of other owner's picks,......
that the last thing I want is to have to constantly update some formulas while my fellow owners are yelling at me to Hurry up and Pick!
The draft is stressful enough that I'd like to at least leave some room to enjoy my "Christmas Morning".
Call me crazy (YOU'RE CRAZY! :wacko: ),.....but thanks, no thanks.
Skippy the one eyed Beagle
Footballguy
Check out my app. It tracks the entire draft for you and then has reporting. I still am unclear how ZVBD differs from DVBD. My app supports the process he described. Free Demo available at: www.thecoordinator.comAfter reading through ZVBD, it appears as though it allows you to have greater confidence in picking a position based on your roster rules, and some level of insight into how the draft will go. To me, this is a great idea since these are the main things that can bite you if you just blindly use VBD.Also, I trust FBG. They spend a lot more time on projections than I would ever want to spend, so I just use their projections. Call me lazy, but other than tweaking a few players projections myself, I don't touch any of it. ZVBD could be very helpful to me, but probably only if there is an app I can run during the draft. As commissioner of my league it's hard enough recording all the picks, and crossing them out on my own sheet to do ZVBD calculations too!!
Mungo Burrows
Footballguy
DVBD says "How much do I drop going into the next round if I don't pick that position here. If I read the article correctly, ZVBD looks out deeper at positions where you have more starters/draft more players. At QB you only look one and a half or two rounds out, but at RB with backups you've got 4 roster spots to fill, so delaying a RB now affects all four RB spots so the baseline looks more rounds out. ...I think
FightinAmish
Footballguy
As a fellow Pats Fan, I hate to disagree with you, but I think you're slightly missing the point of VBD and ZVBD as I understand them. Value Based Drafting is not a substitute for projections, or luck, for that matter. VBD is simply a system to help owners to better assess "the value of players at different positions in order to display their relative worth." If you haven't read Bryant's article, it is definitely worth a few minutes of your time. In fact, his article clearly acknowledges that all projections are flawed and subject to the ole horseshoe in the sphincter. Lets face it, everything we do is flawed, but systems like those presented by the Footballguys are there to help us eliminate some of the inevitable risk and maximize our draft day performance.
Rob Zarzycki
Footballguy
two things my compadres...1) the ZVBD takes time to get used to. i had to practice implementing it 5+ times and then it quickly becomes second nature. It's just simple addition/subtraction really. If i had a laptop i'd of already created a program to do the work, but since i don't have a laptop i'm stuck with standard conventional paperwork.2) a lot of you are asking about the theory behind it, lookout for my upcoming articles "static baselines" and "dynamic baselines", those will give you a lot of underlying theory.
Kasparov
Footballguy
I've ALWAYS felt that DVBD simply meant using dynamic, or changing baselines. This idea of looking at two rounds at a time (which is what you are describing here MB?) comes from GregR and, while it's a worthy endeavor, it's not really what I consider to be DVBD. If VBD is drafting for value by using baselines, then DVBD is drafting for value by using baselines that change from round-to-round. Dynamic Valuing requires the values, or X-Factor to speak back to Joe's original article, to change as the draft progresses. The X-Factors change because the situation in which you are drafting is different each and every round so the thought process is that the players "values" will also change. Dynamic baselining (DVBD) tries to take this into consideration. ZVBD is dynamic baselining (DVBD) on paper.
GregStarling
Footballguy
that's what I've always thought too.
Mungo Burrows
Footballguy
I realize that DVBD can encompass a multi-round approach - I tried to stick my head into some of those conversations between you and GregR but most of the math (linear programming?) was beyond me. However, when most people talk about DVBD (I'll call this "standard DVDB), they refer to a process of setting your baseline at the players you think will be available at each position when you pick in the next round.I think what Rob is proposing here is having dynamic baselines, but the baselines at each position look more rounds ahead at positions where you need more players. Your baseline at one position may only be one round out while your baseline at another position could be 4 or 5 rounds out. It combines the "supply" aspect of standard DVDB (base choices on who is likely to be taken/available at a future pick) with some form of "demand" factor based on the number of players needed at the position.
I guess I'm somewhat interested in it because it addresses one of the weaknesses I saw in "standard" DVDB which is that it did not consider the fact that multiple players are needed at certain positions. (E.g, Not taking a RB in round 1 means your RB2 is probably also going to be worse). I'll need to look at this some more to see if it's something worth pursuing or if it gives me more ideas of how to organize my own draft, but it caught my attention.
Kasparov
Footballguy
This may very well be what many people's impressions are of DVBD, however nothing is written in stone here. Setting baselines is a lot like Nescafe: Tasters' Choice! If you want to set your dynamic baselines at two rounds out, go nuts! 3 rounds out? Fine! This idea of setting the baselines at different depths (rounds out) for different positions however, needs some explaining. I'm not sure about it's validity since it may artificially inflate/deflate the values of an entire position. Demand can be assumed to be tied to the baselines to begin with since they are reflecting, well, the demand for each position! So by setting the baselines deeper for RB's for example might make them seem more valuable than they need be. Just some thoughts off the top of the noggin'.....I just wanted to add that the Demand = Baseline idea that I speak of really only addresses half of the Demand equation. You must also work in what YOUR Demand for a particular position is and re-value accordingly.
Last edited by a moderator:
