Dynasty VORP Calculation (1 Viewer)

[Concept Coop]

Is an objective dynasty value calculator possible? Of course, the projections would be subjective. But is there an objective formula available?

I know this calculation is flawed, or incomplete at least, but I am having a hard time understanding where. This is what I have come up with:

RBA scores 300 pts/year for 4 years. Our baseline RB scores 200 pts/year for 2 years. RBA scores 1,200 (baseline: 400) which is 300% over baseline.

WRA scores 250 pts/year for 8 years. Our baseline WR scores 150 pts/year for 6 years. WRA scores 2,000 (baseline: 900) which is 222% over baseline.

Using this calculator, RBA offers more VORP than WRA. Help me understand how and why this calculation falls short. Or maybe it doesn't and I am over thinking things.

This is only using two baselines (points, longevity) and more would need to go into it (durability, et cetera). Also, the points/year would vary, especially for young players waiting for opportunity and older players losing a step. This is one area where it gets tricky. This would be my solution: chart the top 30 players at each position for the last 10 years or so, and break it down by rank:

QB5= 280 pts/yr

RB5= 260 pts/yr

WR5 = 240 pts/yr

This way, if I am not comfortable projecting 5 years out for Mark Ingram, I can decide that I think he is a future top 5 talent at that position and use the average. Or, if I think Jonathan Stewart scores as RB15 this season because he is splitting time, but next year and beyond, he is top 5, I can use those averages.

I would greatly appreciate any thoughts, comments, and opinions. Thanks in advance.

 

[Sabertooth]

I'm not smart enough to come up with this stuff. But I am interested in the Cliffs notes.

 

[Concept Coop]

I had hoped to do more with this this, but haven't had the time. I will keep you updated. I need to start tracking injuries (X% games missed at age X= Y% games misse for remainder of career), draft position for rookies, and a lot more. Maybe someday I'll have a working formula.

For now, I am using it and giving the player with the higher per year VORP projection a 5-10% increase.

 

[Joseph]

Is an objective dynasty value calculator possible? Of course, the projections would be subjective. But is there an objective formula available? I know this calculation is flawed, or incomplete at least, but I am having a hard time understanding where. This is what I have come up with:RBA scores 300 pts/year for 4 years. Our baseline RB scores 200 pts/year for 2 years. RBA scores 1,200 (baseline: 400) which is 300% over baseline.WRA scores 250 pts/year for 8 years. Our baseline WR scores 150 pts/year for 6 years. WRA scores 2,000 (baseline: 900) which is 222% over baseline.Using this calculator, RBA offers more VORP than WRA. Help me understand how and why this calculation falls short. Or maybe it doesn't and I am over thinking things. This is only using two baselines (points, longevity) and more would need to go into it (durability, et cetera). Also, the points/year would vary, especially for young players waiting for opportunity and older players losing a step. This is one area where it gets tricky. This would be my solution: chart the top 30 players at each position for the last 10 years or so, and break it down by rank:QB5= 280 pts/yrRB5= 260 pts/yrWR5 = 240 pts/yrThis way, if I am not comfortable projecting 5 years out for Mark Ingram, I can decide that I think he is a future top 5 talent at that position and use the average. Or, if I think Jonathan Stewart scores as RB15 this season because he is splitting time, but next year and beyond, he is top 5, I can use those averages. I would greatly appreciate any thoughts, comments, and opinions. Thanks in advance.

I think you definitely need to factor in some type of time value of VBD into the equation. IMO, 300 points in Y1 > 300 points in Y2 > 300 points in Y3.What that factor is though, is probably going to be up to an individual owner.

 

[Concept Coop]

Is an objective dynasty value calculator possible? Of course, the projections would be subjective. But is there an objective formula available? I know this calculation is flawed, or incomplete at least, but I am having a hard time understanding where. This is what I have come up with:RBA scores 300 pts/year for 4 years. Our baseline RB scores 200 pts/year for 2 years. RBA scores 1,200 (baseline: 400) which is 300% over baseline.WRA scores 250 pts/year for 8 years. Our baseline WR scores 150 pts/year for 6 years. WRA scores 2,000 (baseline: 900) which is 222% over baseline.Using this calculator, RBA offers more VORP than WRA. Help me understand how and why this calculation falls short. Or maybe it doesn't and I am over thinking things. This is only using two baselines (points, longevity) and more would need to go into it (durability, et cetera). Also, the points/year would vary, especially for young players waiting for opportunity and older players losing a step. This is one area where it gets tricky. This would be my solution: chart the top 30 players at each position for the last 10 years or so, and break it down by rank:QB5= 280 pts/yrRB5= 260 pts/yrWR5 = 240 pts/yrThis way, if I am not comfortable projecting 5 years out for Mark Ingram, I can decide that I think he is a future top 5 talent at that position and use the average. Or, if I think Jonathan Stewart scores as RB15 this season because he is splitting time, but next year and beyond, he is top 5, I can use those averages. I would greatly appreciate any thoughts, comments, and opinions. Thanks in advance.

I think you definitely need to factor in some type of time value of VBD into the equation. IMO, 300 points in Y1 > 300 points in Y2 > 300 points in Y3.What that factor is though, is probably going to be up to an individual owner.

I have definitely considered this, but ultimately decided against it. It would weight value greatly towards RBs and away from WR/QB/TE. I think the nature of the position adds enough value towards the elite RBs and want more of a plain picture. But I definitely get what you're saying. I just wanted to take team position in the league and personal preference out of it.

 

[Joseph]

Is an objective dynasty value calculator possible? Of course, the projections would be subjective. But is there an objective formula available? I know this calculation is flawed, or incomplete at least, but I am having a hard time understanding where. This is what I have come up with:RBA scores 300 pts/year for 4 years. Our baseline RB scores 200 pts/year for 2 years. RBA scores 1,200 (baseline: 400) which is 300% over baseline.WRA scores 250 pts/year for 8 years. Our baseline WR scores 150 pts/year for 6 years. WRA scores 2,000 (baseline: 900) which is 222% over baseline.Using this calculator, RBA offers more VORP than WRA. Help me understand how and why this calculation falls short. Or maybe it doesn't and I am over thinking things. This is only using two baselines (points, longevity) and more would need to go into it (durability, et cetera). Also, the points/year would vary, especially for young players waiting for opportunity and older players losing a step. This is one area where it gets tricky. This would be my solution: chart the top 30 players at each position for the last 10 years or so, and break it down by rank:QB5= 280 pts/yrRB5= 260 pts/yrWR5 = 240 pts/yrThis way, if I am not comfortable projecting 5 years out for Mark Ingram, I can decide that I think he is a future top 5 talent at that position and use the average. Or, if I think Jonathan Stewart scores as RB15 this season because he is splitting time, but next year and beyond, he is top 5, I can use those averages. I would greatly appreciate any thoughts, comments, and opinions. Thanks in advance.

I think you definitely need to factor in some type of time value of VBD into the equation. IMO, 300 points in Y1 > 300 points in Y2 > 300 points in Y3.What that factor is though, is probably going to be up to an individual owner.

I have definitely considered this, but ultimately decided against it. It would weight value greatly towards RBs and away from WR/QB/TE. I think the nature of the position adds enough value towards the elite RBs and want more of a plain picture. But I definitely get what you're saying. I just wanted to take team position in the league and personal preference out of it.

But does that create a situation where a guy who will perform slighly over the baseline for many years is more valuable than the high end stud who will give you one year of stud performance, one year of middling production and then retire?If so, should it?

 

[Concept Coop]

Is an objective dynasty value calculator possible? Of course, the projections would be subjective. But is there an objective formula available? I know this calculation is flawed, or incomplete at least, but I am having a hard time understanding where. This is what I have come up with:RBA scores 300 pts/year for 4 years. Our baseline RB scores 200 pts/year for 2 years. RBA scores 1,200 (baseline: 400) which is 300% over baseline.WRA scores 250 pts/year for 8 years. Our baseline WR scores 150 pts/year for 6 years. WRA scores 2,000 (baseline: 900) which is 222% over baseline.Using this calculator, RBA offers more VORP than WRA. Help me understand how and why this calculation falls short. Or maybe it doesn't and I am over thinking things. This is only using two baselines (points, longevity) and more would need to go into it (durability, et cetera). Also, the points/year would vary, especially for young players waiting for opportunity and older players losing a step. This is one area where it gets tricky. This would be my solution: chart the top 30 players at each position for the last 10 years or so, and break it down by rank:QB5= 280 pts/yrRB5= 260 pts/yrWR5 = 240 pts/yrThis way, if I am not comfortable projecting 5 years out for Mark Ingram, I can decide that I think he is a future top 5 talent at that position and use the average. Or, if I think Jonathan Stewart scores as RB15 this season because he is splitting time, but next year and beyond, he is top 5, I can use those averages. I would greatly appreciate any thoughts, comments, and opinions. Thanks in advance.

I think you definitely need to factor in some type of time value of VBD into the equation. IMO, 300 points in Y1 > 300 points in Y2 > 300 points in Y3.What that factor is though, is probably going to be up to an individual owner.

I have definitely considered this, but ultimately decided against it. It would weight value greatly towards RBs and away from WR/QB/TE. I think the nature of the position adds enough value towards the elite RBs and want more of a plain picture. But I definitely get what you're saying. I just wanted to take team position in the league and personal preference out of it.

But does that create a situation where a guy who will perform slighly over the baseline for many years is more valuable than the high end stud who will give you one year of stud performance, one year of middling production and then retire?If so, should it?

The formula, or my intensions for it, is not to give either of the two hypothetical player models an advantage over the other by design. It is in place to measure them objectively, so that they can be measured against each other. For example, in the formula, even if Arian Foster never played another down of football (and I had a time machine and knew that) he would still have more VORP this time a year ago than a guy like Knowshon Moreno would for his entire career if I projected Moreno to score 10 points over baseline for 8 years.

 

[C & C]

I'm interested in this, but can only slightly contribute right now...

My first concern is these two players get rated evenly...

Player X

year 1 - 75

year 2 - 75

year 3 - 75

year 4 - 75

Player Y

year 1 - 10

year 2 - 20

year 3 - 120

year 4 - 150

I can find a bunch of players who will be similar to Player X, but it's likely worth holding onto Player Y for two years for him to develop.

 

[Concept Coop]

I'm interested in this, but can only slightly contribute right now...My first concern is these two players get rated evenly...Player Xyear 1 - 75year 2 - 75year 3 - 75year 4 - 75Player Yyear 1 - 10year 2 - 20year 3 - 120year 4 - 150I can find a bunch of players who will be similar to Player X, but it's likely worth holding onto Player Y for two years for him to develop.

They are rated evenly, only because they are, during those 4 years, in raw points over replacement. The calculation isn't to project who will do what - you have to input the projections. It is only to measure those projections in value over replacement.But, your point is very valid. Player Y could very well be the one most likely to win a championship. I am open to finding an objective way to equate for that. But it doesn't seem very likely that you would project someone to score above the baseline by 10, then by 140 3 years later. But, again, you point is valid and worth consideration.

 

[DoubleG]

I'm interested in this, but can only slightly contribute right now...

My first concern is these two players get rated evenly...

Player X

year 1 - 75

year 2 - 75

year 3 - 75

year 4 - 75

Player Y

year 1 - 10

year 2 - 20

year 3 - 120

year 4 - 150

I can find a bunch of players who will be similar to Player X, but it's likely worth holding onto Player Y for two years for him to develop.

They are rated evenly, only because they are, during those 4 years, in raw points over replacement. The calculation isn't to project who will do what - you have to input the projections. It is only to measure those projections in value over replacement.But, your point is very valid. Player Y could very well be the one most likely to win a championship. I am open to finding an objective way to equate for that. But it doesn't seem very likely that you would project someone to score above the baseline by 10, then by 140 3 years later. But, again, you point is valid and worth consideration.

And herein lies the tricky part in dynasty leagues:The bolded part only holds true if you surround Player Y with players whose production follows a similar trajectory. Put another way, if you have a "start 3WR" team half full of Steve Smith (CAR) and Santana Moss - and the other half is Torrey Smith (BAL) and Greg Little (CLE) - you could have a team that is slightly above average for a long time, but never quite "peaks". In other words, looking at the values in terms of overall team VORP over time may also be helpful.

Sorry if I just created more work.

 

[CalBear]

I would probably approach this with a relatively significant risk discount; if I predict Arian Foster to score 300 points this year, I would discount him by, say, 30% per year going forward. So if the replacement player is 100 points, and Foster is projected to score 300/210/147/103 points over the next four years (including 30% risk discount), he'd be projected for 200+110+47+3 future VORP points (360) for dynasty purposes. I'd play with the risk discount number until the tail looked reasonable; it would probably be different for different positions.

 

[crnerblitz]

If you can conquer this equation, then you should be working on algorithm to predict buy/sell points through technical indicators for equity trading.

Seriously.

 

[griff321]

Evaluating risk factors is what gets me. You can take general consistency ratings and apply them to tiers to get an idea, but obviously some factors have to affect consistency more then others. Then how would you weigh them or even predict the likeliness of player X being afflicted with limiting factor Y?

 

[SSOG]

Jason and I toyed with a similar problem (how to create plausible cross-position values) over at DR.net when we were trying to create a trade calculator. We actually established a pretty decent working framework, I think, but we got bogged down in how to adjust those values in many-for-few player trades. We had some decent ideas, but ultimately I think we just ran out of energy maybe 20% away from a plausible solution.I forget the exact methodology we settled on, but here's a rough overview:Step 1: Come up with 5-year VBD totals by position and sort them (i.e. the best WR scored XXX VBD over this span, the 8th best WR scored YYY VBD, and so on down the line). This was the easiest step. We settled on 5-year totals because 1-year totals create a lot of noise at RB (who most frequently have huge seasons and then disappear), while 10-year totals would skew QBs and WRs (who are the only players who have 10 productive years). No special reason why 5 is the magic number, it just seemed like a good total- 5 years sounds like a reasonable lifespan for a stud player at any position.Step 2: Adjust VBD totals by how likely you are to start that player in any given week (because points on the bench mean nothing). This was more complicated- basically, starting in 2008, Jason has been tracking how frequently a player in the preseason top-X wound up appearing as a starter in Dodds' weekly projections. He then created multipliers based on that data, and used it to create marginal rates. The marginal production difference between QB1 and QB9 was valued at 100%, the marginal production between QB10 and QB18 was somewhere around 50%, the marginal production from QB19 to QB27 was somewhere around 25%, and so on down the line. The nice thing about this methodology is that it can be adjusted based on league size and starter counts (i.e. the marginal rates in a 12-team start 2 QB league would obviously be radically different than the marginal rates in an 8-team start 1 QB league), which was important to us because our goal was to create a universal one-size-fits-all calculator. I think we even created input boxes so you could put in how many people are rostered at each position in your own particular league, and the algorithm would adjust to match (if only 30 QBs are rostered, QB35 obviously isn't very valuable).Step 3: We took our custom value score and smoothed it out (rolling averages) and applied it over top of my rankings (so my #1 rated WR had a value score equal to the value score of the top WR over the 5 year average). We then created a mechanism to subtly adjust these values to more properly reflect my projections going forward (mostly, to change what is a smoothed-out historical production curve into a set of rankings with discernible tier breaks). We experimented with various methods, but iirc we settled on an algorithm that over-weighted players in the same tier and under-weighted players on the other side of the tier break.Final result: I just rank the players, set the tier breaks, and let the website automatically generate "Dynasty Value Scores" (although we'll give it a catchier acronym- we've been calling it jVal behind the scenes).As I said, the project still needs a bit more work and polish (we need to test it more to make sure it isn't spitting out nonsensical values at any point), but hopefully we'll have a chance to get back to it this season, polish it up, and get some actual values rolled out for public consumption at some point.Edit: in case I wasn't clear, the system is designed to take subjective inputs (my rankings) and generate objective outputs (dynasty value scores). Obviously the output will only be as good as the inputs, which is why I'm talking in generics (WR1 = XX, RB5 = YY).

I think you definitely need to factor in some type of time value of VBD into the equation. IMO, 300 points in Y1 > 300 points in Y2 > 300 points in Y3.What that factor is though, is probably going to be up to an individual owner.

NFL teams think like this, too... which is why New England always has a zillion draft picks. They turned a 1st rounder years ago into a steady stream of extra 2nd rounders year after year after year after year after year. I think New England's widely-acclaimed strategy puts the lie to the whole "immediate assets are worth more than future assets!" mindset.The only reason to overrate current assets, in my mind, is if you are worried that the league will fold. If there's only a 50% chance that the league is still around 3 years from now, then production 3 years from now should be weighted 50%. If you feel certain the league will last, though, and your goal is to maximize long-term success, then future production is worth no less than current production, and you can take advantage of anyone who thinks otherwise to become the New England of your dynasty league, doing things like grabbing Vincent Jackson when everyone has him ranked barely in the top 20 last year and then turning around and watching everyone rocket him up into the top 10 this year because all of a sudden his production, which was always inevitable, is now imminent.