I took a peak down the football “Wins Above Replacement” rabbit hole recently and, although I’ve never been much for fantasy football (at least, not as it’s generally played today), I wondered if there couldn’t be a ‘better’- dare I say “more ottoneu”- method to scoring fantasy football.
The challenges with creating a football “WAR” (or a similar, all-encompassing advanced metric, such as one that could be used for fantasy purposes) are many. Without delving into all that here, my thinking instead is to translate and scale different measures for fantasy football outputs into what we already know to be an incredibly perfect measure: the ottoneu fantasy baseball offensive formula (which, of course, is built from weighted Runs Created (wRC)). The idea is to effectively create a “wRC” for fantasy football… or, what I’ll call a ‘weighted Points Created’ measure.
Because of the game’s statistical complexity, all offensive and defensive stats will be valued within the same “wRC” translation; that is, there won’t be a separate formula for offensive and defensive scores. Instead, the goal here is to translate our baseball stat ‘categories’ from the ottoneu formula into football stat ‘categories’. From there, we can balance the two translations with a relationship factor between how many occurrences occur for each football stat versus baseball stat in each category. After that, we set a weighting for each football stat within a category by multiplying the ‘balance factor’ between sports by the weighted measure from the baseball formula.
One important note: there is no measure that factors offensive touchdowns of any kind- passing, rushing, or receiving (although there are measures for Defensive and Return TD’s). As with wRC, where “Runs” aren’t factored but instead predicted, the same is true here- where offensive scores aren’t factored, but are (hopefully) predicted. The philosophical comparison is that baseball isn’t a game of R’s, but a game of Outs, and avoiding enough Outs will eventually score R’s (and securing Outs efficiently will prevent them); likewise, the point of football isn’t to score, but rather to move the chains toward a fresh set of downs (or to prevent first downs defensively), and in doing so eventually the offense will score naturally. This is important to remember as I work through the figuring of each statistical category’s relative translation between baseball and football.
As a further side note, I like to see H2H scores that reflect actual “real-life” scores; which is to say, I’m dividing the ottoneu Hitting formula by 10 as it relates to the “balance factor” between the two sports. For example, instead of a -1 AB value, I prefer -0.1. The idea is that each player’s score will reflect his actual football/game points value, based on the stats and weights assigned to them, with each team’s score better reflecting the sort of score you’d see in an actual NFL game.
Another aside, I’m using categories that are available on Yahoo’s platform as that is arguably the most customizable game engine. Whether these stats (or better stats) are available with ottoneu’s service, I wouldn’t know.
Also, although it’s a little late to start a league I’d be remiss if I didn’t invite anyone who would like to join the one I’ve set up on Yahoo to test this format. You can follow the league here; or join with this link. The league is free and set for 12 Teams with an Auction Draft scheduled on TUE, 9/11 @ 8:15pm ET (estimated draft time is ~2 hrs). Scoring will start on Week 2, with a 12-Week regular season and Playoffs starting on Week 14-16. I’ll also create a Divisional schedule once the league is full, just for fun.
With all that said, let’s dive in…
ROSTERS
I argue that the purpose- or goal- of fantasy sports is (or should be) to intuitively build a complete team that is reflective of ‘real life’ with ‘realistic’ scores. The first step to accomplishing this goal is with Roster settings. For this reason, I’m keeping “Kickers” and “Team Defense/Spec. Teams” as part of the setup, while also having roster spots for individual defensive players. The full roster categories are:
Offense - QB, RB, TE, WRx2, WR/TE, WR/RB, WR/RB/TE, K
Defense - DT, DE, LB, CB, S, DB, Dx2, DEF
(“D” as a ‘UTIL’-type spot for any defensive player)
So that’s 9 “players” on both sides of the ball. I’m not sure if the number of Bench/IR slots really matter here (at least not to the extent as they do in a H2H baseball format), so I’m only concerned with a balanced Active roster design.
SCORING
The fun stuff. I’ve added a philosophical explanation for the inclusion of the football stats for each formula category. The mathematical computations are made using 2017 stats only; I’m sure the differences year-to-year would change some of the balance factors somewhat, but I’m not sure to what degree it really matters. In any case, I wasn’t going to go deeper than last year as I’m offering this mainly as a proof of concept.
.
AB, -0.1 (baseball)
Football equivalents: Pass Attempt, Rush Attempt, Passer Sack, Passer Interception, Incomplete Pass, Fumbles, and Fumbles Lost
Philosophically, AB’s are the impetus for every play- as are Passing and Rushing Attempts in football. That is, no play happens without one of these occurring (ignoring the AB vs PA factor here). In football terms, every play takes time off the clock and moves the game closer to an end- thus the negative values. The Passer Sack, Passer Interception, Incomplete Pass, Fumbles, and Fumbles Lost stats further reflect unsuccessful offensive plays which work to undermine the team’s chances to score points before the end of the game.
Mathematically, there were 165,567 AB’s in the MLB last year. To make the balance translation, all of the occurrences in the categories above are added up, which equals 40,250 PassATT + RushATT + PassSACK + PassINT + INCpass + FUM + FumLOSS. The balance factor is then 40250 / 165567 = 0.243. The final step is to multiply the baseball weighting by the balance factor; so in this case, 0.243 * -0.1 = -0.024. The weighting, then, for the seven football categories here is “-0.024”.
.
H’s, 0.56 (baseball)
Football equivalents: Receptions, Completed Pass, Pass Defended, Tackles Solo, and Tackles Assisted
Philosophically, H’s are the basic measure for a Hitter’s success. Likewise, these are the basic category measures for offensive and defensive success.
Mathematically, there were 42,215 H’s in the MLB last year. In the NFL, there were 55,982 occurrences of the above categories. The balance factor is then 42215 / 55982 = 0.754. Multiplied by the weighting factor, 0.754 * 0.56 = 0.422. So the “weighted Points Created” factor for these five football categories is “0.422”.
.
2B’s, 0.29 (baseball)
Football equivalents: Receiving Yards, Defensive Sack, Tackle for Loss, and Fumbles Forced
Philosophically, these are all an additional ‘degree’ of offensive and defensive success. They describe events that either move the ball “more” forward (offensively) or backward (defensively).
Mathematically, there were 8,397 2B’s in the MLB last year. The NFL saw 119,034 occurrences of the stats above. 8397 / 119034 = 0.071 balance factor, * 0.29 = 0.020. So these four football categories here have a “0.020” weighting. In Yahoo, this translates to a “48.88 Yards = 1 Point” for the Rec YDS category.
.
3B’s, 0.57 (baseball)
Football equivalents: Receiving 1st Down, Defensive Fumble Recovered, Defensive Interception, Blocked Kick, 3 & Outs Forced (team stat), 4th Down Stops (team stat)
Philosophically, these are all plays that describe a further ‘degree’ of success. They describe events that either move the chains (for receivers) or turn the ball over (for defense). Two of the three team stats that are being counted in the formula are in this category.
Mathematically, there were 795 3B’s in the MLB in 2017; in the NFL, there were 8,314 plays for the football categories above. 795/8314 = 0.096 *0.57 = 0.055 weighting for the football stats in this category.
.
HR’s, 0.94 (baseball)
Football equivalents: Passing 1st Down, Rushing 1st Down, Defensive TD, Safety, Kickoff+Punt Return TD (team stat)
Philosophically, as with HR’s, these are all the “best possible” plays for both sides of the ball. As I detailed above, the goal of the offense is to establish 1st Downs (with all plays starting as either a Pass or Rush play), while any scoring play on defense and special teams is the ultimate reward there. The KO/PR TD stat is the third of three team stats factored in this exercise.
Mathematically, there were 6,105 HR’s last year; in football, the number of plays above totaled 8,955. So 6105/8955 = 0.6817 * 0.94= 0.641 weighting for the five football stats here.
.
BB+HBP, 0.3 (baseball)
Football equivalents: Pass Yards, Rush Yards, Return Yards, Turn-Over Return Yards
Philosophically, these are all describing the degree to which the player who starts the play (offense) or ends the play (defense) moves the ball in their direction.
Mathematically, there were 17,592 BB’s+HBP’s for Hitters last season, and 210,127 occurrences of the above football stats. 17592/210127 = 0.084 * 0.3 = 0.025 weighting. As all of these describe Yardage, the translation is 39.81 Yards = 1 Point for these four football stats.
.
SB, 0.19 (baseball)
Football equivalents: FG Made, 2-Point Conversions, Extra-Point Made
Philosophically, the SB/CS dynamic is being used here to describe all Point-After attempts; as the MLB average for SB% tracks relatively closely to the success percentage of all PAT attempts, this makes for a logical plug-in for football.
Mathematically, there were 2,527 SB’s last season and 1,967 successful PAT’s last season. 1967/2527 = 0.778 * 0.19 = 0.148 weighting for these PAT categories.
.
CS, -0.28 (baseball)
Football equivalents: FG Missed, Extra Point Missed
Philosophically, these balance out the successful PAT categories above (excluding unsuccessful 2-Pt Conv’s).
Mathematically, there were 937 CS and 229 missed PAT’s last year. 229/937 = 0.244 balance factor * -0.28 = -0.068 is the factor to score these football stats.
.
FULL “weighted Points Created” FACTOR LIST
“AB Category”
Offense = Points
PassATT = -0.024
RushATT = -0.024
PassSACK = -0.024
PassINT = -0.024
INCPass = -0.024
FUMB = -0.024
FumLOSS = -0.024.
“H Category”
Offense = Points
REC = 0.422
CompPASS = 0.422Defense = Points
PassDEF = 0.422
TackSOLO = 0.422
TackASST = 0.422.
“2B Category”
Offense = Points
RecYDS = 48.88 YDS = 1 PtDefense = Points
SACK = 0.020
TackLOSS = 0.020
FumFORCED = 0.020.
“3B Category”
Offense = Points
REC1stDown = 0.055Defense = Points
FumREC = 0.055
INT = 0.055
BlocKICK = 0.055
3&OForced(TEAM) = 0.055
4thDownSTP(TEAM) = 0.055.
“HR Category”
Offense = Points
PASS1stDown = 0.641
RUSH1stDown = 0.641Defense = Points
DefTD = 0.641
SAFETY = 0.641
KO+PRTD = 0.641.
“BB+HBP Category”
Offense = Points
PassYDS = 39.81 YDS = 1 Pt
RushYDS = 39.81 YDS = 1 Pt
ReturnYDS = 39.81 YDS = 1 PtDefense = Points
TORetYDS = 39.81 YDS = 1 Pt.
“SB Category”
PAT Attempts = Points
FGMade = 0.148
2PTConv = 0.148
XPMade = 0.148.
“CS Category”
PAT Attempts = Points
FGMissed = -0.068
XPMissed = -0.068
.
You can check-out the player pool for my Yahoo league linked above here (sorted by Offense and 2017 stats). Some positional comparisons:
QB
#1 T. Brady, 409.84
#5 B. Roethlisberger, 376.82
#15 J. Goff, 319.41
#25 J. McCown, 268.47
RB
#1 L. Bell, 123.01
#5 M. Gordon III, 93.52
#25 J. Ajayi, 56.04
#50 R. Burkhead, 34.79
WR
#1 A. Brown, 79.50
#5 L. Fitzgerald, 74.94
#25 D. Amendola, 48.45
#50 A. Cooper, 37.67
TE
#1 T. Kelce, 61.05
#5 J. Doyle, 50.49
#15 A. Hooper, 33.27
#25 D. Njoku, 22.76
LB
#1 B. Martinez, 64.62
#5 P. Brown, 62.17
#25 R. Shazier, 42.70
#50 T. Davis, 34.70
DB
#1 A. Jackson, 61.76
#5 L. Collins, 48.08
#25 D. Slay, 38.64
#50 T. Jefferson, 34.83
DL
#1 K. Mack, 34.77
#5 C. Campbell, 31.12
#15 A. Robinson, 25.86
#25 B. Graham, 22.31
K
#1 W. Lutz, 12.42
#5 C. Boswell, 10.74
#15 B. Walsh, 7.97
#25 A. Rosas, 6.21
TEAM (2 Def Cat’s + 1 Spc Tm Cat)
#1 Baltimore, 4.47
#5 Denver, 3.96
#15 New Orleans, 3.17
#25 Miami, 2.59
Naturally, QB’s by far have the most influence on “points created” with both the remaining offensive and defensive player pools relatively balanced between each other at some factor less than QB’s, and K’s and Team Defenses/Special Teams having the least weighted influence- though still enough to perhaps affect the outcome of any match. Or in other words, about what you’d expect from each roster component in terms of their overall impact on the game if you were building a more complete team.
Anyway, just throwing it out there- but I think establishing a more unique and intuitive method for scoring fantasy football would be the “ottoneu way” to further distinguish the service as an innovator within the field. As people continue to struggle to find an adequate football version of WAR, I propose we could do just that here for fantasy purposes by working from the wRC baseball formula that ottoneu was built around.
In any case, I’m now looking forward to this fantasy football season for the first time in like 15 years or so, if only just to see how this format could play out…
Thanks for reading!