By giving bonuses for yardage only for QBs or giving points for completions or something like that? I know a guy who runs a league with completions being worth 1 point and incompletions being worth -1 point. All his other settings are normal stuff. What this does is make QBs the highest scoring players generally. It makes them as important as in real football or so he says.
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.
Moving away from a 2RB Fantasy System (1 Viewer)
- Thread starter TheMathNinja
- Start date
Mr. Irrelevant
IBL Representative
Well, you wouldn't necessarily start the RB25 over the WR33 in that situation. But you would, and should, draft the RB25 before the WR33 in that situation, probably by at least a round or more. And it's precisely for the reason you yourself mentioned - the value of RBs drops off so rapidly after that point that you're leaving a ton of VAR on the table by passing on the RB in favor of the WR. You're going to get significantly more points on average out of the combo of RB25 plus the 1st WR off the WW than you would with WR33 plus the 1st WW RB.
In that sense, multiple flex positions impact the positional scarcity of RBs much less than would otherwise be expected.
Dez
Footballguy
Why don't we make it really easy for everyone and you can start 0 QB, 0 RB, 0 WR, 0 TE and 8 Flex and let someone start 8 QBs if they want ? That be fun......by saying you only have to start 1 RB, 1 WR and allowing 4 Flex you are taking a lot of strategy out of the game. If you have to start 2 RB at least you have to make some decisions about building and maintaining you team...Do you want to pay the price to secure a stud RB even though they don't last as long or take a WR that lasts more years but doesn't have the impact in your league as a top RB would.
NCPanthersFan
Footballguy
Performance bonus is yardage + TD length for all position players. 6 pts all around. Yardage & TD length bonus for QB's is heavier so that is where we look to make up the gap.. Getting a QB that finishes in the top 5 is a huge advantage but comes with some serious riskreward come draft time. We average 3-5 QB's in the 1st round each year. However based upon the depth this year I think I am betting it is closer to 3.
Ghost Rider
Footballguy
I like leagues with more flex options, as it gives the option of teams being built in different ways. Instead of everyone trying to get the same positions to fill the starting lineups, someone might go heavy on RBs, while someone might go heavy on WRs, or even TEs! That flexibility makes it more fun and unpredictable. The only drawback is trades are less likely to happen, but that doesn't bother me a lot.
CalBear
Footballguy
In a three-flex situation, it does not matter whether your flex players are RBs or WRs. It only matters who will score more points. Once you have filled out your mandatory starting requirements, RB and WR are essentially the same position. (With a slight modification for bye week and backup requirements).
So, no, you will not get more points out of RB25 and a replacement player than out of WR25 and a replacement player. Not if WR25 scores more than RB25. Because the replacement player is the same player in either scenario; the waiver-wire RB or WR who (you think) will score the most points.
Last edited by a moderator:
Bell went a couple rounds earlier than Luck in a startup dynasty a few weeks ago. I am trying to figure out a way to have a non-PPR league where we do not see 25 of the first 35 picks be running backs.
KellysHeroes
Footballguy
when I created my dynasty leagues 5 - 8 yrs ago I saw that even back then the RB2 spot was becoming devalued and the flex spot was becoming WR more often than not. So I kept most 12 Teamers 2 RB require but 14 and 16 teams became 1 RB with an extra flex.
I can tell you a couple of yrs ago that when RB and WR were equal in terms of line up requirements RB used to cost more but now that has changed a lot.
Top RB tiers (top 10% of the RBs) stil cost more than the Top WR tier bt when u get lower in the tiers the WRs now cost more than the RB.
For example, if you offer RB15 for WR15 it'll most likely get rejected, you now have to offer RB15 for WR20 to get a player for player deal done.
I can tell you a couple of yrs ago that when RB and WR were equal in terms of line up requirements RB used to cost more but now that has changed a lot.
Top RB tiers (top 10% of the RBs) stil cost more than the Top WR tier bt when u get lower in the tiers the WRs now cost more than the RB.
For example, if you offer RB15 for WR15 it'll most likely get rejected, you now have to offer RB15 for WR20 to get a player for player deal done.
Last edited by a moderator:
Backside Pursuit
Footballguy
How is it I missed this thread? A subject near and dear to my heart...
Bottom line for me is: I want fantasy re-drafts to resemble (theoretical) annual NFL skill-position dispersal drafts. I want fantasy innovations that increase realism rather than simply increase scoring. I greatly appreciate TheMathNinja's work (here and elsewhere) but, IMO, the god-awful PPR distortion and the RB-eligible Flex are not realistic or necessary counterbalances to the original 2-RB distortion. The ideal lineup for me would feature 6 skill position players only, just like real football:
1 QB
1 RB
2 WR
1 TE
1 WR/TE
etc...
cobalt_27
Footballguy
Exactly. The 2 RB start setup is one of the biggest distortions in FF lore and has dictated disproportionate value to the position for nearly 3 decades. Thankfully, some people (here in this thread especially) are wise to this and adopting either the sensible 1RB start or going with a 2 QB league. I prefer the former, but either one helps the equation.
How is it I missed this thread? A subject near and dear to my heart...
Bottom line for me is: I want fantasy re-drafts to resemble (theoretical) annual NFL skill-position dispersal drafts. I want fantasy innovations that increase realism rather than simply increase scoring. I greatly appreciate TheMathNinja's work (here and elsewhere) but, IMO, the god-awful PPR distortion and the RB-eligible Flex are not realistic or necessary counterbalances to the original 2-RB distortion. The ideal lineup for me would feature 6 skill position players only, just like real football:
1 QB
1 RB
2 WR
1 TE
1 WR/TE
etc...
You can hardly get an NFL team to invest a 1st rounder on a RB these days, so what is it that makes FF commissioners so absurdly out of touch that they keep setting up their leagues to force feed RBs to 8 of the top 10 teams every year (and 2 QB in the top 30).
So dumb.
msudaisy26
Footballguy
What league is this? I had pick 1.01 and traded down to 1.03 this year and took Bell I would have gladly given up either pick + for Luck
msudaisy26
Footballguy
To be more like real football wouldn't you go 1 qb 0 - 3 running backs, Green Bay has done it a few times, 0 - 5 receivers, and 0 - 3 tight ends? but you must start 6
How is it I missed this thread? A subject near and dear to my heart...
Bottom line for me is: I want fantasy re-drafts to resemble (theoretical) annual NFL skill-position dispersal drafts. I want fantasy innovations that increase realism rather than simply increase scoring. I greatly appreciate TheMathNinja's work (here and elsewhere) but, IMO, the god-awful PPR distortion and the RB-eligible Flex are not realistic or necessary counterbalances to the original 2-RB distortion. The ideal lineup for me would feature 6 skill position players only, just like real football:
1 QB
1 RB
2 WR
1 TE
1 WR/TE
etc...
cobalt_27
Footballguy
cobalt_27
Footballguy
Not terribly reflective of the league as whole. At least Bp's suggestion reflects the more common 11 or 12 personnel configurations.To be more like real football wouldn't you go 1 qb 0 - 3 running backs, Green Bay has done it a few times, 0 - 5 receivers, and 0 - 3 tight ends? but you must start 6
It doesn't have to be perfect and position scarcity and tradition will always skew RB value, at least in the foreseeable future. But, dear god, let's not compound the absurdity by starting 2 RB or, worse, starting 2 and then including them in the flex spot.
msudaisy26
Footballguy
The one thing I don't like is if a team is lucky enough or smart enough to find 3 good backs in the draft, free agency or trading why would you punish them and make them pick 2 every week. If you aren't going to to punish a team that is unlucky to have a late draft position and thus doesn't get one of the top running backs, or is lazy and doesn't pick up players in a timely fashion. Then why would you punish a team that found more than 1 reliable back?Not terribly reflective of the league as whole. At least Bp's suggestion reflects the more common 11 or 12 personnel configurations.To be more like real football wouldn't you go 1 qb 0 - 3 running backs, Green Bay has done it a few times, 0 - 5 receivers, and 0 - 3 tight ends? but you must start 6
It doesn't have to be perfect and position scarcity and tradition will always skew RB value, at least in the foreseeable future. But, dear god, let's not compound the absurdity by starting 2 RB or, worse, starting 2 and then including them in the flex spot.
cobalt_27
Footballguy
Why does this question only apply to RB? Couldn't this be said for every position.The one thing I don't like is if a team is lucky enough or smart enough to find 3 good backs in the draft, free agency or trading why would you punish them and make them pick 2 every week. If you aren't going to to punish a team that is unlucky to have a late draft position and thus doesn't get one of the top running backs, or is lazy and doesn't pick up players in a timely fashion. Then why would you punish a team that found more than 1 reliable back?Not terribly reflective of the league as whole. At least Bp's suggestion reflects the more common 11 or 12 personnel configurations.It doesn't have to be perfect and position scarcity and tradition will always skew RB value, at least in the foreseeable future. But, dear god, let's not compound the absurdity by starting 2 RB or, worse, starting 2 and then including them in the flex spot.To be more like real football wouldn't you go 1 qb 0 - 3 running backs, Green Bay has done it a few times, 0 - 5 receivers, and 0 - 3 tight ends? but you must start 6
See, this is the point. Get outside the small world RB box for a minute, catch the fresh air supplied by the rest of the arguably more important positions in the league, and you'll see how multi-dimensional this game can be.
Otherwise, cue up your flash cards and follow your RB-RB-RB draft routine.
cobalt_27
Footballguy
Why does this question only apply to RB? Couldn't this be said for every position? In fact, I'm arguing for folks to start 1 RB and considering diversification of the other positions early in the draft.See, this is the point. Get outside the small world RB box for a minute, catch the fresh air supplied by the rest of the arguably more important positions in the league, and you'll see how multi-dimensional this game can be.The one thing I don't like is if a team is lucky enough or smart enough to find 3 good backs in the draft, free agency or trading why would you punish them and make them pick 2 every week. If you aren't going to to punish a team that is unlucky to have a late draft position and thus doesn't get one of the top running backs, or is lazy and doesn't pick up players in a timely fashion. Then why would you punish a team that found more than 1 reliable back?Not terribly reflective of the league as whole. At least Bp's suggestion reflects the more common 11 or 12 personnel configurations.It doesn't have to be perfect and position scarcity and tradition will always skew RB value, at least in the foreseeable future. But, dear god, let's not compound the absurdity by starting 2 RB or, worse, starting 2 and then including them in the flex spot.To be more like real football wouldn't you go 1 qb 0 - 3 running backs, Green Bay has done it a few times, 0 - 5 receivers, and 0 - 3 tight ends? but you must start 6
Otherwise, cue up your flash cards and follow your RB-RB-RB draft routine.
Last edited by a moderator:
msudaisy26
Footballguy
So nice he said it twice!Why does this question only apply to RB? Couldn't this be said for every position? In fact, I'm arguing for folks to start 1 RB and considering diversification of the other positions early in the draft.See, this is the point. Get outside the small world RB box for a minute, catch the fresh air supplied by the rest of the arguably more important positions in the league, and you'll see how multi-dimensional this game can be.The one thing I don't like is if a team is lucky enough or smart enough to find 3 good backs in the draft, free agency or trading why would you punish them and make them pick 2 every week. If you aren't going to to punish a team that is unlucky to have a late draft position and thus doesn't get one of the top running backs, or is lazy and doesn't pick up players in a timely fashion. Then why would you punish a team that found more than 1 reliable back?Not terribly reflective of the league as whole. At least Bp's suggestion reflects the more common 11 or 12 personnel configurations.It doesn't have to be perfect and position scarcity and tradition will always skew RB value, at least in the foreseeable future. But, dear god, let's not compound the absurdity by starting 2 RB or, worse, starting 2 and then including them in the flex spot.To be more like real football wouldn't you go 1 qb 0 - 3 running backs, Green Bay has done it a few times, 0 - 5 receivers, and 0 - 3 tight ends? but you must start 6
Otherwise, cue up your flash cards and follow your RB-RB-RB draft routine.
Being serious though I am saying a person should be allowed to start all 3 backs if they want or none, start 5 receivers or none, start 3 tight ends or none. They are all valid and legal formations in the NFL and it gives the most options while not punishing any person/team for wanting running backs or drafting them well, the same can be said for receivers or tight ends.
Mario Kart
Footballguy
The league I am using these stats from have
1 QB, 1/20 yards, 6pt/TD, -2/INT, -1/Fum (Interception would be for other positions as well)
2 RB, 1/10 yards, 6pt/TD, -1/Fum
2/3 WR, 1/10 yards, 6pt/TD
1/2 TE, 1/5 yards, 6pt/TD
each position is 1pt/per reception
For the most part, the rankings of the positions are pretty equal from year to year. The QB 06 scores similar year to year as do the other positions year in and year out. A (4) denotes that four seasons were above the average meaning one year that position rank scored well below the deviation of that ranking. A (2) skews the other way in that there were two really good years that made the weighing in its favor. Overall though, the scoring is pretty spot on. In theory, if I could add the previous year, the rankings, in theory, would all have (3)'s.
Now, what to do? Are RB's heavily outweighed in the data? I would say that RB's are not heavily favored given the sample size. If this league were to draft a new dynasty draft... would 10 RB's go in the first round? Can't say for sure, however, my guess would be no.
What is skewed and why? If "equal points" is the goal which would, in theory, equalize each position, scoring should either be increased or decreased of the QB's. Meaning, since it is a start 1, the QB 06 should score in the 250's as does RB 06 and WR 06 and TE 06... or scoring should be increased and allow QB 06 to score nearer to 500 points. That would increase the QB value because it would put them on par or equal to their positional rank. As it stands now, a top QB is great to have as they will outscore a TE 06 and especially a TE 12 much more (I use TE because that is the only other position that is required to start 1).
Starting requirements - should the start 2 RB go away in this league? I would say... no. It's always better to have the top tier of players but in the above, where is the discrepancy? WR's is where it is at, it appears. And, that is with the same scoring as RB's. The requirements are at least 2 WRs but if you happen to have 3 in the top 18, you're sitting pretty good.
Drafting if done today. Maybe, just maybe the draft would be full of RB's early. But, why? I think the answer has to do with perception more than actuality. Perception is RB's score more. That may be true as in they may score the most points in one game versus a WR... ie. boom or bust therefore they go earlier. Or, they may go early because their window is smaller and thus more "predicable" in that sense. If I know one position is going to be in the top 12 (or have a better chance at top 12) in the next three years, I'm going to pick that guy. So, now, we are working with variability. Which position offers the least variability? The RB, overall. Why? Their window in much smaller.
Now, the above is a 12 team league. There are currently 7 starting positions. If there were to be a change, I would keep it at 7 but make the requirements start 1 QB, 1-2 RB, 2-4 WR, 1-3 TE. More variability for bye weeks and team strength. I would suggest the start 2 RB be limited to 12 team leagues though. Start making that requirement for 14+ and the value will skyrocket. As it stands with the above, it looks fine the way it is with possibly minor tweaks.
1 QB, 1/20 yards, 6pt/TD, -2/INT, -1/Fum (Interception would be for other positions as well)
2 RB, 1/10 yards, 6pt/TD, -1/Fum
2/3 WR, 1/10 yards, 6pt/TD
1/2 TE, 1/5 yards, 6pt/TD
each position is 1pt/per reception
Now, what is the goal for a fantasy league? Will it be to have "equal scoring" so drafts are not lopsided? Will it be to "spread the wealth" in that a draft will not overvalue one position over another? Let's look now at the data from the last five years from the above league.
For the most part, the rankings of the positions are pretty equal from year to year. The QB 06 scores similar year to year as do the other positions year in and year out. A (4) denotes that four seasons were above the average meaning one year that position rank scored well below the deviation of that ranking. A (2) skews the other way in that there were two really good years that made the weighing in its favor. Overall though, the scoring is pretty spot on. In theory, if I could add the previous year, the rankings, in theory, would all have (3)'s.
Now, what to do? Are RB's heavily outweighed in the data? I would say that RB's are not heavily favored given the sample size. If this league were to draft a new dynasty draft... would 10 RB's go in the first round? Can't say for sure, however, my guess would be no.
What is skewed and why? If "equal points" is the goal which would, in theory, equalize each position, scoring should either be increased or decreased of the QB's. Meaning, since it is a start 1, the QB 06 should score in the 250's as does RB 06 and WR 06 and TE 06... or scoring should be increased and allow QB 06 to score nearer to 500 points. That would increase the QB value because it would put them on par or equal to their positional rank. As it stands now, a top QB is great to have as they will outscore a TE 06 and especially a TE 12 much more (I use TE because that is the only other position that is required to start 1).
Starting requirements - should the start 2 RB go away in this league? I would say... no. It's always better to have the top tier of players but in the above, where is the discrepancy? WR's is where it is at, it appears. And, that is with the same scoring as RB's. The requirements are at least 2 WRs but if you happen to have 3 in the top 18, you're sitting pretty good.
Drafting if done today. Maybe, just maybe the draft would be full of RB's early. But, why? I think the answer has to do with perception more than actuality. Perception is RB's score more. That may be true as in they may score the most points in one game versus a WR... ie. boom or bust therefore they go earlier. Or, they may go early because their window is smaller and thus more "predicable" in that sense. If I know one position is going to be in the top 12 (or have a better chance at top 12) in the next three years, I'm going to pick that guy. So, now, we are working with variability. Which position offers the least variability? The RB, overall. Why? Their window in much smaller.
Now, the above is a 12 team league. There are currently 7 starting positions. If there were to be a change, I would keep it at 7 but make the requirements start 1 QB, 1-2 RB, 2-4 WR, 1-3 TE. More variability for bye weeks and team strength. I would suggest the start 2 RB be limited to 12 team leagues though. Start making that requirement for 14+ and the value will skyrocket. As it stands with the above, it looks fine the way it is with possibly minor tweaks.
Mario Kart
Footballguy
I guess I am posing the question of... Show the flaw in the 2RB system given the data above.
Believe me, I am not a fan of having to start two RB's but where is the flaw in the starting requirements? Many of us remember the days of requirements consisting of 2RB AND 2WR... adding a 3rd WR helped alleviate the value of the RB but where do we go with the above?
Believe me, I am not a fan of having to start two RB's but where is the flaw in the starting requirements? Many of us remember the days of requirements consisting of 2RB AND 2WR... adding a 3rd WR helped alleviate the value of the RB but where do we go with the above?
Backside Pursuit
Footballguy
I think that is a good ideological question and the answer may simply boil down to an inherent predisposition or personality trait. Some of us are more process-oriented and some of us are more results-oriented. But I certainly have respect for either approach and I think that your league can work and work well.
Generally speaking, I am not a fan of starting with a set of desired results and reverse-engineering a system that produces those desired results. It definitely can be done but it's never the cleanest, simplest, least artificial or most scalable approach, in my opinion. (And this is true of all systems, from economic systems to fantasy football systems.)
Let me give you an absurd hypothetical. Let’s imagine that I have devised the perfect “equitable” formula to encourage balanced fantasy team-building. TEs get a .82 PPR and WRs get a .64 PPR and everybody starts 1.5 RBs—everybody starts 2 RBs with 1 designated as getting half the points of whatever he produces. The results are ideally equitable and the system works. Why would I be against such a system? 1. It introduces unnecessary complexity. 2. It attempts to counterbalance one artificial construct with yet more artificial constructs. (Of which, PPR might very well be the worst.) 3. It’s not scalable. If my league expands from 12 teams to 14 teams, I now have to recalculate all of my formulas. (As you note when you say: "I would suggest the start 2 RB be limited to 12 team leagues though. Start making that requirement for 14+ and the value will skyrocket. As it stands with the above, it looks fine the way it is with possibly minor tweaks.") 4. It’s not realistic.
I don't want to play in a 2-QB league not because it cannot be constructed fairly or provide a valid challenge but because it does not correlate with real football. (Similarly, I would not want to play in a 2-kicker league or 7-WR league or a scoring bonus or PPR league, regardless of how challenging and equitable it might be.)
The league you describe sounds very balanced. Where would Andrew Luck be drafted in a redraft? The answer to that question (the "cobalt_27 test") is critical to whether I would consider it realistic or not.
I like simplicity. Give me 6 skill position starters, just like real football. (And blocking backs are not fantasy relevant. If people want to separate tailbacks from fullbacks, that's fine with me and people are welcome to draft and start a fantasy-unproductive fullback in their lineup using the I formation. 11 and 12 personnel groupings are the ~60% norm and ought to be the standard in FF as well.)
There is no flaw. Stick with it. It sounds like you’ve got a good system. Scarcity is what potentially separates RBs from WRs more so than PPG. Scarcity is what drives demand and inflates prices. Through the use of PPR and lineup requirements, your league counterbalances the scarcity demand.
By the way, if anybody would like to join a free, public (yes, public), 16-team, 1-RB league, I have one drafting this Sunday evening…
BigSteelThrill
Footballguy
Where they are drafted initially, is way different then having them maintain similar value once the season starts.
Mario Kart
Footballguy
Nothing to add to your post, I like what you say. One thing I disagree with is the "6 positional starters."
If fantasy football was to mimic real football, each league would have 32 teams in it. Obviously, this is not possible for everyone. To counteract the "6 positional starters" that is not entirely accurate. If fantasy football were to mimic real football, each team uses 53-man rosters. In theory that is ~25 offensive positions each week. Broken down, that would be 2-3 QB's, 2-4 RB's, 3-5 WR's, 1-3 TE's plus the offensive line.
Now, do each of those players play? Most likely, yes, but they each serve different purposes. The starters will certainly play more and so on. But, now, we have to shrink this supposed league to only 12 teams. Should a 12 team league start the above requirements? (in a perfect world there would be 36 NFL teams for this example but as such we don't have fantasy leagues with 10.67 teams). So, the numbers don't work out nicely but we can still use them.
In a 12-team league, given the number of real players in the NFL, any team could start 2-3 RB's comfortably and still mimic NFL teams. In this example, each fantasy team will "play" 1-2 QB, 2-3 RB, 3-5 WR, 1-3 TE... making each league have starting requirements of 8-10 players for each fantasy team... and not 6... for "true accuracy" of the NFL as each NFL team plays that amount of positional players each game. Limiting teams to 6 starters also increases "the luck factor" as there is less room for error. As it happens, it seems most leagues have settled on the 7-8 starters per game.
Ultimately, what I think is the biggest problem that no one has been able to solve, and will not be able to solve is the following. All of these numbers work in a bottle but we are talking real life mimicry now. What is the issue we cannot overcome at least when it comes to H2H play? The NFL uses a bye-week and in fantasy land, that is impossible. Thus, the implementation of total points leagues.
Ultimately #2... it all depends what you, each individual likes. There is no one formula but most need to understand that the more constraints there are, the more valuable some positions will become and that is natural.
If fantasy football was to mimic real football, each league would have 32 teams in it. Obviously, this is not possible for everyone. To counteract the "6 positional starters" that is not entirely accurate. If fantasy football were to mimic real football, each team uses 53-man rosters. In theory that is ~25 offensive positions each week. Broken down, that would be 2-3 QB's, 2-4 RB's, 3-5 WR's, 1-3 TE's plus the offensive line.
Now, do each of those players play? Most likely, yes, but they each serve different purposes. The starters will certainly play more and so on. But, now, we have to shrink this supposed league to only 12 teams. Should a 12 team league start the above requirements? (in a perfect world there would be 36 NFL teams for this example but as such we don't have fantasy leagues with 10.67 teams). So, the numbers don't work out nicely but we can still use them.
In a 12-team league, given the number of real players in the NFL, any team could start 2-3 RB's comfortably and still mimic NFL teams. In this example, each fantasy team will "play" 1-2 QB, 2-3 RB, 3-5 WR, 1-3 TE... making each league have starting requirements of 8-10 players for each fantasy team... and not 6... for "true accuracy" of the NFL as each NFL team plays that amount of positional players each game. Limiting teams to 6 starters also increases "the luck factor" as there is less room for error. As it happens, it seems most leagues have settled on the 7-8 starters per game.
Ultimately, what I think is the biggest problem that no one has been able to solve, and will not be able to solve is the following. All of these numbers work in a bottle but we are talking real life mimicry now. What is the issue we cannot overcome at least when it comes to H2H play? The NFL uses a bye-week and in fantasy land, that is impossible. Thus, the implementation of total points leagues.
Ultimately #2... it all depends what you, each individual likes. There is no one formula but most need to understand that the more constraints there are, the more valuable some positions will become and that is natural.
Backside Pursuit
Footballguy
No one expects exact alignment. It seems to me always preferable to err on the side of realism whenever possible, especially when it comes to easy and basic settings like scoring and lineups. Even if someone figures out how to score offensive linemen and long snappers, we are certainly not obligated to replicate 32 teams and 53-man rosters. We can do things on a smaller scale yet maintain realistic scoring and realistic lineups.Mario Kart said:
The number of starters and/or on-field personnel ought to be as straightforward as “touchdowns equal 6 points.” Imagine that the default fantasy lineup has historically always been:
2 QB
1 RB
2 WR
1 TE
1 K
Imagine that we’ve been using this default for decades. Why would the results be so distorted? Because there are never (or almost never) 2 QBs on the field at the same time (even if there are 2 QBs on the same team). So instead of fixing this anachronistic distortion we decide to create new ones—like “points per touches” or “YAC multipliers”—in order to counterbalance the original distortion. It works but it seems to me less than ideal. Why not just remove the original distortion?
Another factor in this is allowing for in-game substitutions. The I formation does not utilize 2 fantasy-relevant tailbacks but RBBC and rotations are realistic parts of the game and need to be accounted for. I hate the idea of 2-RB starting lineups because no team starts Barry Sanders and Curtis Martin in the same backfield at the same time. No team hands the ball off at that sort of volume. Also, if I start Jordy Nelson and he is injured in the first quarter, my team is down a man for the rest of the game but in the real world the Packers do not play shorthanded. It seems to me we might be able to figure out a backup plan for something like this. Either we can use a pre-designated injury half-substitute (where half the points earned by the designated substitute count) or we can unlock rosters at halftime and make second half substitutions. For example, I can pull Jordy Nelson at halftime of his game and substitute another WR on my roster for second half value only. The substitute WR might be one playing a game at the same time or might be one playing later in the day or the next day. So if I pull Nelson at halftime and substitute Denarius Moore playing a late start, I am allowed to combine only Moore’s second half points with Nelson’s first half points. I like the idea of allowing for a second RB to replace the primary RB in-game but I loathe the Tecmo insanity of 2 starting tailbacks piling up ridiculous numbers and skewing player values unrealistically.
Ace Matherton
Footballguy
I have one league that has been like this for 6 years and it is easily the most fun league to participate in. I've started 4 TE's and 1 WR in a bind, much more strategy to it imo.
Maelstrom
Footballguy
That's very similar to a league I commished for going on 10 years (dissolved now cause the company closed and people went their separate ways). We basically had required 1 RB, 3 WR and 1 TE, and 2 RB/WR/TE flex spots. So you could run 3-3-1, 2-4-1, 1-5-1, 1-4-2, 1-3-3. Then we adjusted the scoring for the relative positions (primarily through PPR) to equalize the positions a bit more. First, no one could ever consistently run the 3-3-1 lineup, just not enough quality RBs to go around. Plus with the scoring fairly equalized, building a team around a stud WR and TE was viable. Your first rate stud RBs were still king of the hill, but in the top 10 for scoring at the end of the year, there was always a couple WRs, a TE or 2 and a QB or 2. It worked out pretty well, and made for an interesting draft as you couldn't just cross of the next next 16 RBs right after your pick.
Mario Kart
Footballguy
I have liked the discussion the past couple of days in this thread. The posts following my long one with data, to me, have validated the 2RB requirement as being "okay" and not some demon that needs to be killed off. Sure, starting 1RB with other options might be "more than okay" but the topic of 2RB being bad has not been there in this thread.
I will reassert that if there are leagues with 14+ teams, 1RB is ideal. And, few reasons to get away from 2RB if in a 10-team league.
I will reassert that if there are leagues with 14+ teams, 1RB is ideal. And, few reasons to get away from 2RB if in a 10-team league.
BigSteelThrill
Footballguy
to both of you.Also you sitting on those TEs causes them to raise in value. While this is "odd", we have all seen people horde the other positions, particularly RB.
ghostguy123
Footballguy
How is that "much more strategy"?? All you do is go after the players that you think will score the most, and you don't have to make decisions based on position at all, ever.
I am not saying it's a bad setup or anything. I don't prefer leagues like that, but I can see the appeal. Just saying, there is NO strategy for collecting RBs, WRs (and TEs if they also apply) if you simply need to just figure out who will score more. Positional stratagy as far as allocating resources is a huge strategy, and this lacks that.
Backside Pursuit
Footballguy
In the end, I think it all depends upon how you feel about PPR. Earlier, you asked to “show the flaw” in a specific 2RB system and I think the implicit answer to that request is: PPR.
PPR is intrinsically flawed and intrinsically corrupting. PPR was designed to artificially inflate the value of one subset of players (WR/TEs) in order to offset the already artificially inflated value of another subset of players (RBs)—all at the cost of artificially depressing the value of the most important subset of players (QBs). We don’t have points-per-completion for QBs. We don’t have points-per-carry for RBs. We don’t have points-per-kickoff for kickers.
If someone finds PPR acceptable or preferable, then there is nothing wrong with the 2RB system. (Other than that it absurdly and unrealistically depresses the relative value of QBs.) If, on the other hand, someone believes PPR to be the most mind-numbingly foolish, artificial, and corrupting innovation in this great game, one’s feelings about the 2RB system are quite different.
TheMathNinja
Footballguy
Wow, I'm excited this thread has been revived. Lots of awesome discussion going on. Backside, I've enjoyed your comments the most, because I think it reflects an accurate understanding of the evolution of Fantasy Football. I much prefer taking away the 2RB system because it was an artificial construct to begin with (i.e. many teams simply do not feature 2 RB's. Consider Washington, Minnesota, Tampa Bay, etc.) to messing around w/ scoring rules to de-value the RB position. So forcing it is a bit silly IMO, and many people agree, which is why this thread is getting such mileage.
However, that being said, the PPR bonus may not be quite as arbitrary or artificial as you suspect, Backside. I've done some pretty extensive research on the Estimated Points Added value of a rushing attempt, a reception, etc. Turns out that a WR who catches 5 balls for 100 yards contributes less on average to his team's success than a WR who catches 10 balls for 100 yards if both have the same number of targets. The reason is, the one who catches 10 balls likely earned more first downs for his team, and ultimately, first downs is what moves you down the field and gets you points. In the above example, say I'm at my own 20 yard line. I have a RB who runs for 5 yards on average, but is known to get 3-yard gains quite often (RB's have a right-skewed distribution in yards gained meaning that they might average 5 yards per game, but because one of those was a 30-yard breakout, the median run was closer to 3 yards). I'm going to run 2 run plays and 3 pass plays, targeting the same WR on all 3 pass plays, and either get 3ydRun/20yardPass/Incomp/3ydRun/Incomp or I'm going to get 3ydRun/8ydPass/Incomp/3ydRun/12ydPass. In the first scenario, I have a WR who gets 1 catch for 20 yards on 3 targets, but I end up only reaching my own 46 and punting on 4th and 7. In the second scenario, I have a WR who gets 2 catches for 20 yards on 3 targets, and because those two each earned me first downs, I have a 1st and 10 at my own 46 instead of punting there. MAJOR value added.
While the above situation is a bit contrived, something like this is often repeated enough to make the Estimated-Points-Added value of a reception about 0.6 points on average (*while holding yards, TD's, fumbles, and Targets constant*). Specifically, my data shows and average of:
0.490 pts per reception for RB's (+/- 0.145 w/ standard error)
0.768 pts per reception for WR's (+/- 0.083 w/ standard error)
0.549 pts per reception for TE's (+/- 0.158 w/ standard error)
The reason we see a higher value on receptions for WR's is that they average more yards per catch than either RB's or TE's, and thus, a higher proportion of their catches earn first downs, which like I said, is ultimately what gives value. What this means is that if you have two WR's who have a 100-yard game on equal targets, the one who has 2 more catches than the other will have, on average, earned 1.5 more points for his team. When looking at RB's and TE's, it will mean he on average scored 1 more point. It's worth noting, however, that if you do not hold targets constant (and most fantasy scoring systems do not account for targets) but only hold yards, TD's, and fumbles constant, a reception is worth, on average, -0.4 points. This is because more receptions for a set amount of yards, usually means more targets, and more targets for the same amount of yards generally means less efficiency (more incompletions and drops, which are "negative plays"). So if your system does not include a value for targets, maybe you should have negative PPR ;-)
But I include targets in my scoring (at -1 pt per target, which I know is controversial, but I stand by it for now), and for this reason, I prefer a system that gives 0.5 pts PPR combined with starting only 1 RB. In particular I like:
1 QB
1 RB
2 WR
1 TE
2 RB/WR/TE Flex
because I think the options this system give most resembles what teams will likely have for "Feature players". I don't like options where you can start zero at a position. People here have reasoned that starting zero is legitimate because many formations include zero RB's or zero WR's, etc. Though this is true, Fantasy Football is not a system where you switch up formations in a single game; you go with one formation for a whole game. And over the course of a season (and usually, even over the course of a game), no NFL team will be in a position where they have NO feature RB or feature WR or feature TE. Every team, by season's end, will have at the very minimum 1 feature RB, two feature WR's, 1 feature TE, and two other feature players (most commonly 1 RB and 1 WR, but often a number of other combinations). That's why I like starting 7, and that's why I like the above format.
But I still like 0.5 PPR, not because I need to artificially inflate my WR and TE positions to account for the scarcity of RB's in my league because I require 2 to start, but because the data shows that players who catch more balls, at a fixed targets and yardage level, are still more valuable to a team than players who catch fewer balls.
Backside Pursuit
Footballguy
Excellent work, as usual, TMN. Your stuff is always a pleasure to read. I apologize for making you repeat yourself as you already covered most of that upthread. Here’s my question: Is a 3-yard carry on 3rd and 2 more valuable than a 3-yard carry on 1st and 10? If so, how do we account for that? Would you, ideally, prefer to award a bonus directly for first downs rather than extrapolate imperfectly from statistical tendencies? Last year, Adrian Peterson had 85 rushing first downs; Calvin Johnson had 92 receiving first downs. Do we only reward the latter? Last year, Michael Turner and Josh Gordon both picked up first downs 40 times. Is Calvin Johnson’s value greater than Josh Gordon’s both because of yardage gained and first downs gained? Is Adrian Peterson’s value greater than Michael Turner’s (or DeMarco Murray’s with 39) both because of yardage gained and for moving the chains? If so, do we also award a .5 bonus for every x carries? And what about Matt Ryan’s 79 first downs gained on third down attempts?
How about if we award an x-point bonus for every first down gained on a third- or fourth-down attempt? Now that’s something that feels intuitively gratifying to me. (But it would apply equally to all positions.)
ETA: If the argument in favor of PPR is contingent upon third-down conversions, it still feels arbitrary and discriminatory to me to reward one position for third-down proficiency while ignoring the other positions. PPR may be internally consistent within the WR subset but it also seems externally inconsistent across subsets.
How about if we award an x-point bonus for every first down gained on a third- or fourth-down attempt? Now that’s something that feels intuitively gratifying to me. (But it would apply equally to all positions.)
ETA: If the argument in favor of PPR is contingent upon third-down conversions, it still feels arbitrary and discriminatory to me to reward one position for third-down proficiency while ignoring the other positions. PPR may be internally consistent within the WR subset but it also seems externally inconsistent across subsets.
Last edited by a moderator:
Mario Kart
Footballguy
I had the idea of scoring first downs ten years ago from the link in this page but the technology (MFL etc.) was not up to date to score those kinds of things. The idea of "targets" were not even looked at ten years ago so scoring those would obviously not been able to be done.Backside Pursuit said:
The above quote by Backside is, possibly, looking into scoring a player in a similar way that NBA fantasy is. Each position is, essentially, scored the same way but each position scores more at what that position is ideal for. Guards get the same amount of points for a block as do centers but, obviously, centers have more blocks... but guards typically have more steals, etc.
The point is, that a league should designate the same scoring for each position no matter what. Some will have a problem with that because it may place a higher value on one position or the other however, I forget who in this thread, pointed out that WR/TE tend to have higher yard receiving averages... okay, for every reception over 10 yards, each position gets a point for that. Forget screen passes and dump offs (unless they run for more than 10 yards). What this scoring will do is ensure, most of the time, a first down is gained but also will not reward RB's who catch screens for 5 yards (same goes for WR's but who tends to catch those short passes more?)
Instead of looking at how to tweak PPR or "devalue" the RB position due to the draft tendencies, tweak how things are scored for all to fit the whole paradigm. Why are RB's more valuable? (The answer is debatable but lets say we figure it out) Okay, lets not devalue that aspect, instead figure out why WR's are valuable at what they do and emphasize that more.
SaintsInDome2006
Footballguy
Somewhere along the line I read a version of the story or the legend of the beginnings of fantasy football, around 1961-62 or so, that it started amongst sports writers covering the AFL, they would get together in bars and hotels wherever as they travelled covering the teams and the whole thing started over "who is better", this RB or that one, this QB or that one, etc.
Actually I think the only way to really even it out to reflect modern reality for coaches would be to expand the active FF rosters, because that is what has happened in real football since those early days. It would I think put an emphasis on going beyobnd real "studs" and finding sleepers and talent. I haven't found any takers on this, but I know such leagues exist.
Warrior
Footballguy
I think this is a silly topic.
Similar to how a wartime video game like Call of Duty would be extremely boring if your character died after a single shot, fantasy football would be very boring if it were set up like a real NFL team.
Fantasy teams have multiple RBs because it's more fun and accomplishes the goals of fantasy players better than 1 RB.
Similar to how a wartime video game like Call of Duty would be extremely boring if your character died after a single shot, fantasy football would be very boring if it were set up like a real NFL team.
Fantasy teams have multiple RBs because it's more fun and accomplishes the goals of fantasy players better than 1 RB.
BigSteelThrill
Footballguy
Yeah i don't see the problem with allowing teams to start 2... I just see the benefit of allowing some to start 0 at that position if they choose.
cobalt_27
Footballguy
It's dumb. Moreover, it's silly to defend a league setup that has a 10:1 ratio of RB selected over QB in the first 30 picks.
I know a lot of you are unable to see beyond just the "this is the way it's always been done" argument. But, when you really, really think about it, you know it's moronic.
mjr
Footballguy
Not when its a cap league. And even when it isn't, there's no lack in the need to allocate resources as you say. Its simply allocated in a different way than it would in the more traditional lineup.
In my league some teams will go WR heavy, some RB heavy and some stack up on TEs, which creates imbalances that require just as much strategy to contend with as any other league. The point here being that its not as easy as simply stacking up on 9 of the highest scoring guys as you would think, most particularly if those coveted players are eating up large portions of your cap, as they're bound to do.
TheMathNinja
Footballguy
To answer your question: yes, a 3-yard carry on 3rd and 2 is far more valuable than a 3-yard carry on 1st and 10. In fact, a 3-yard carry on 3rd and 2 is on average [barely] more valuable than a 9-yard carry on 1st and 10. In the first situation, you take your team from a somewhat dire situation to a great position. In the latter, you take your team from a great position to another great position, further down field than the first example. This is roughly how they balance each other.Backside Pursuit said:
To answer your question regarding whether this makes me want to award points for a first down (which more customizable sites like MFL allow for), I've definitely considered it. I would certainly create a more accurate points-added component to fantasy points. But after looking at the data closely, I noticed that incorporating first downs into the calculations only very minimally adds accuracy to how fantasy players are ranked in terms of estimated points added. The basic reason for this is because first downs are distributed somewhat randomly among players (i.e. there aren't as many "Third down backs" as you might think...meaning guys who post pretty poor yardage totals but rack up major quantities of first downs). So leaving out the first down data doesn't really compromise accuracy much at all, and just introduces some added complexity. In terms of R^2 values, we're talking an increase on the magnitude of 1% of explaining EPA in a model that otherwise incorporates yards, attempts, etc. if we add first downs. In my estimation, just not worth it. My gut says doing this for only 3rd-and-4th-downs wouldn't change much of anything either.
And, for running backs, I actually prefer to give -0.5 points per rush, for the simple reason that most runs do not add to their team's Estimated Points position until the RB gets to the 5-yard mark. Obviously, the exception to this is when he gets the first down before that point. But otherwise, if you're going to lose a down, you better gain at least 5 if you want to be in a statistically better position than before. This is the same dynamic for why WR's have a negative point total per reception if you hold yards constant. So, in order to separate why Calvin Johnson is better than Josh Gordon (per your example), and to decide whether more of this has to do with first downs gained or yards gained, multiple linear regression analysis is the only thing that can tell us for sure, and with numerical proof. But, by and large, like I mentioned, a WR's first downs will basically just follow from their yardage total; there isn't a ton of variation with things like "3rd down receivers".
But PPR is unequal across positions, like you mentioned. I gave the average value for a reception in my last post, and it's true that it's more valuable for WR's. For a league that counts targets and not first downs, I think there is some wisdom in giving a WR 0.75 PPR and a RB 0.5 PPR, for example. I personally don't like to add complexity where I don't have to, but the data is there to support it.
I don't think anyone on this forum is saying there's something wrong with allowing teams to start 2 RB's. In fact, I think every one of the proposed flex line-ups allowed for 2 RB's (and this is consistent with NFL practices, also). What this thread is against is requiring teams to start 2 RB's, when the impact is to overvalue the position in drafts and force teams into roster setups that do not resemble all NFL teams' strategies.
The Dynasty
Footballguy
As a 16 team league, we were forced to move to 1 RB. Requiring 32 starting RBs each week would make the bye weeks impossible to navigate through. So we run a
1 QB
1 RB
2 WR
1 TE
1 K
1 DST
1 Flex RB/WR/TE
lineup. Personally, I think a 16 team league is the ideal setup for a league. We were hesitant at first, but we've grown to love it. It's large enough to mimic the NFL, but small enough to keep it fun. QBs go early and often because no one wants to be the guy having to start Andy Dalton every week. Although RBs still maintain their value as well. We avergae 8 RBs and 7 QBs in the first round. This year we added one point per first down reception to keep the value of WRs and TEs while still reflecting the importance of a player's on-the-field, real-life performance. Also, I don't see how having so many flex spots increases strategy. I can just imagine the level of trading in those leagues to be almost nonexistent.
1 QB
1 RB
2 WR
1 TE
1 K
1 DST
1 Flex RB/WR/TE
lineup. Personally, I think a 16 team league is the ideal setup for a league. We were hesitant at first, but we've grown to love it. It's large enough to mimic the NFL, but small enough to keep it fun. QBs go early and often because no one wants to be the guy having to start Andy Dalton every week. Although RBs still maintain their value as well. We avergae 8 RBs and 7 QBs in the first round. This year we added one point per first down reception to keep the value of WRs and TEs while still reflecting the importance of a player's on-the-field, real-life performance. Also, I don't see how having so many flex spots increases strategy. I can just imagine the level of trading in those leagues to be almost nonexistent.
