AlNFL19 Posted February 23, 2019 Share Posted February 23, 2019 I am well aware of the arguments for why WAR is stupid and has no place in football. That being said, I decided to try anyway. There are a couple of things to address before going through the data this year, however, about this specific WAR model: It does not separate players from the other 21 on the field on every play. That is impossible. Truely, a "player's" WAR in this case is the *wins added by plays on which the player in question touched the ball or was the target of a pass.* Obviously, 22 players and even more coaches go into every play, so nothing could ever numerically separate them all. As a result, passing and receiving WAR values overlap. Take Gurley for example. His rushing WAR is pretty high, but it's really the WAR added by plays where Gurley ran the ball, so it's crediting Gurley, his offensive line, other blockers, the defense's mistakes, etc., not just Gurley. "Replacement level" isn't really replacement level, it's the average from the last ten years. So really, it's WAA, but WAR sounds better. It is, just like the baseball version, approximate. That should be self-explanatory. If Quarterback X's WAR is 2 and Quarterback's Y's WAR is 1, and Quarterback X was put on Team Y, they wouldn't be expected to just win 1 more game. That's not how the model works. Again, it's the wins added by plays where the player in question touched the ball. It's not an apples-to-apples comparison because they have different players around them and played against different teams. I've only found WAR for offensive skill players (or anybody else who had a carry or target) because that's the only publicly available data. Okay. First, here's the 1-WAR club (the players whose team production when they had the ball added 1 or more wins): Player Position Team Total WAR Patrick Mahomes QB Kansas City 4.97 Drew Brees QB New Orleans 4.34 Matt Ryan QB Atlanta 3.91 Todd Gurley II RB Los Angeles 3.15 Aaron Rodgers QB Green Bay 2.74 Russell Wilson QB Seattle 2.73 Melvin Gordon RB Los Angeles 2.29 Alvin Kamara RB New Orleans 2.22 Saquon Barkley RB New York 2.16 Jared Goff QB Los Angeles 2.1 Philip Rivers QB Los Angeles 1.95 Tom Brady QB New England 1.81 Derrick Henry RB Tennessee 1.81 Deshaun Watson QB Houston 1.78 Kareem Hunt RB Kansas City 1.7 Ben Roethlisberger QB Pittsburgh 1.67 Andrew Luck QB Indianapolis 1.66 Mike Williams WR Los Angeles 1.58 James Conner RB Pittsburgh 1.56 Tyler Boyd WR Cincinnati 1.56 Tyler Lockett WR Seattle 1.54 Carson Wentz QB Philadelphia 1.47 Aaron Jones RB Green Bay 1.46 Kirk Cousins QB Minnesota 1.3 Phillip Lindsay RB Denver 1.28 Mark Andrews TE Baltimore 1.23 Nick Chubb RB Cleveland 1.23 Michael Thomas WR New Orleans 1.2 Davante Adams WR Green Bay 1.12 Dak Prescott QB Dallas 1.12 Damien Williams RB Kansas City 1.08 James White RB New England 1.07 A.J. Green WR Cincinnati 1.03 Here's the minus-1-WAR club: Blake Bortles QB Jacksonville -1.04 J.J. Nelson WR Arizona -1.06 Ricky Seals-Jones TE Arizona -1.07 Chad Williams WR Arizona -1.08 Case Keenum QB Denver -1.12 Bilal Powell RB New York -1.12 Alfred Morris RB San Francisco -1.19 Dion Lewis RB Tennessee -1.25 Willie Snead IV WR Baltimore -1.26 Nick Mullens QB San Francisco -1.26 Frank Gore RB Miami -1.43 Maurice Harris WR Washington -1.46 Kelvin Benjamin WR BUF / KC -1.60 Jermaine Kearse WR New York -1.61 Sam Darnold QB New York -1.88 Josh Rosen QB Arizona -3.76 Here's the passing leaders: Quarterback WAR 1. Patrick Mahomes +4.79 2. Drew Brees +3.72 3. Matt Ryan +3.29 4. Russell Wilson +2.96 5. Aaron Rodgers +2.32 6. Phillip Rivers +2.28 7. Jared Goff +1.95 8. Andrew Luck +1.87 9. Kirk Cousins +1.55 10. Tom Brady +1.54 Here's the rushing leaders: Player Position Team Rushing WAR T. Gurley II RB Los Angeles 2.90 M. Gordon RB Los Angeles 1.96 A. Kamara RB New Orleans 1.96 S. Barkley RB New York 1.81 D. Henry RB Tennessee 1.75 J. Conner RB Pittsburgh 1.47 J. Allen QB Buffalo 1.45 A. Jones RB Green Bay 1.40 P. Lindsay RB Denver 1.26 I. Crowell RB New York 1.08 Here's the receiving leaders: Player Position Team Receiving WAR T. Lockett WR Seattle 1.60 T. Boyd WR Cincinnati 1.59 M. Williams WR Los Angeles 1.27 M. Andrews TE Baltimore 1.23 M. Thomas WR New Orleans 1.20 D. Adams WR Green Bay 1.12 A.J. Green WR Cincinnati 1.03 K. Hunt RB Kansas City 1.02 C. Ridley WR Atlanta 0.92 E. Ebron TE Indianapolis 0.91 And just for kicks, the top ten fullbacks: Player Team Total WAR J. Develin New England 0.51 A. Armah Carolina 0.41 A. Janovich Denver 0.20 A. Sherman Kansas City 0.19 Z. Line New Orleans 0.18 P. DiMarco Buffalo 0.04 M. Burton Chicago 0.00 J. Olawale Dallas -0.01 D. Vitale Green Bay -0.01 R. Nix Pittsburgh -0.02 Call it feeble guesswork if you want, I think it turned out pretty good for its purposes. I have WAR data for 574 players from last year, so if there's a number you're curious about, ask (except I was too lazy to check passing WAR for anybody other than the 32 majority starting quarterbacks). Link to comment Share on other sites More sharing options...
LETSGOBROWNIES Posted February 23, 2019 Share Posted February 23, 2019 I think that last bit finally puts an end to the age old question of “who’s better, J. Develin or A. Armah?”. Link to comment Share on other sites More sharing options...
DoleINGout Posted February 23, 2019 Share Posted February 23, 2019 46 minutes ago, LETSGOBROWNIES said: I think that last bit finally puts an end to the age old question of “who’s better, J. Develin or A. Armah?”. Armah have to debate you on that one... Link to comment Share on other sites More sharing options...
ramssuperbowl99 Posted February 23, 2019 Share Posted February 23, 2019 1 hour ago, LETSGOBROWNIES said: I think that last bit finally puts an end to the age old question of “who’s better, J. Develin or A. Armah?”. Good to see some conclusive proof that Goff deserved the SB MVP over Brady tbh. Link to comment Share on other sites More sharing options...
childofpudding Posted February 23, 2019 Share Posted February 23, 2019 Looks OK at first glance, but 1) How exactly do you calculate it for each position; and 2) How does it differ from FO's DVOA (or maybe more accurately, DYAR)? Link to comment Share on other sites More sharing options...
buno67 Posted February 23, 2019 Share Posted February 23, 2019 Two things? How is Chubb in this and not Baker? what is Chubbs Rushing WAR Link to comment Share on other sites More sharing options...
AlNFL19 Posted February 23, 2019 Author Share Posted February 23, 2019 7 minutes ago, buno67 said: Two things? How is Chubb in this and not Baker? what is Chubbs Rushing WAR Chubb’s rushing WAR was +1.06. Baker isn’t there for a couple of reasons: - His passing WAR while Hue was still there was -0.54 (without Hue it was +1.28) - His rushing WAR was -0.31 - He gets penalized for having an interception thrown on his only target Link to comment Share on other sites More sharing options...
Thomas5737 Posted February 23, 2019 Share Posted February 23, 2019 11 minutes ago, buno67 said: Two things? How is Chubb in this and not Baker? what is Chubbs Rushing WAR After 2 and a half games we were starting to wonder if the Browns would win a game last year and then Baker came in and every loss was pretty disappointing. So If I use Tyrod Taylor as the replacement Baker's WAR should be at about 8. Link to comment Share on other sites More sharing options...
AlNFL19 Posted February 23, 2019 Author Share Posted February 23, 2019 27 minutes ago, childofpudding said: Looks OK at first glance, but 1) How exactly do you calculate it for each position; and 2) How does it differ from FO's DVOA (or maybe more accurately, DYAR)? It’s calculated the same for every position - it’s the sum of their passing, rushing, and receiving WAR totals. It determines play-by-play efficiency above or below the average and then assigns a WAR value to that based on volume of touches / targets. It’s different from DVOA because it’s based more on easily available box score data than on context and such. As a result, it’s not perfect; but it’s pretty good I think. What it’s based on as attributed to roughly one win is based on the results of every team going back through the 2007 season. Link to comment Share on other sites More sharing options...
AlNFL19 Posted February 23, 2019 Author Share Posted February 23, 2019 5 minutes ago, Thomas5737 said: After 2 and a half games we were starting to wonder if the Browns would win a game last year and then Baker came in and every loss was pretty disappointing. So If I use Tyrod Taylor as the replacement Baker's WAR should be at about 8. As I said in the OP, it’s kinda not replacement level, it’s the average. Taylor was a good bit worse than average. Link to comment Share on other sites More sharing options...
childofpudding Posted February 23, 2019 Share Posted February 23, 2019 4 minutes ago, AlNFL19 said: It’s calculated the same for every position - it’s the sum of their passing, rushing, and receiving WAR totals. It determines play-by-play efficiency above or below the average and then assigns a WAR value to that based on volume of touches / targets. It’s different from DVOA because it’s based more on easily available box score data than on context and such. As a result, it’s not perfect; but it’s pretty good I think. What it’s based on as attributed to roughly one win is based on the results of every team going back through the 2007 season. And how do you calculate their passing, rushing and receiving WAR totals? What's the formula, for someone who is not familiar with WAR? Link to comment Share on other sites More sharing options...
KellChippy Posted February 23, 2019 Share Posted February 23, 2019 So this is kinda like production per touch? Impressive to see Thomas and Adams on there for wrs considering they get force fed the ball. Link to comment Share on other sites More sharing options...
AlNFL19 Posted February 23, 2019 Author Share Posted February 23, 2019 18 minutes ago, childofpudding said: And how do you calculate their passing, rushing and receiving WAR totals? What's the formula, for someone who is not familiar with WAR? So, first of all it uses different statistics like ANY/A and touchdown rate, as two examples, and finds the percent change from the average over the last few years. These numbers are all averaged, while being weighted so some numbers are more important than others because, for example, ANY/A correlates more to wins than touchdown rate does. Passing and receiving are basically the same thing for obvious reasons, while rushing uses the comparable statistics that are rush based rather than pass based (for example, yards per carry and touchdown rate). At the end of this step, you have a percent change from zero. To convert this to WAR, first the number is weighted by volume. What this means is that if Player X had a +10% score on 10 touches, each touch registers an additional 10%. This calculation is then weighted by what percentage of the offense was contributed to by this player (for example, if Player X touched the ball 10 times out of the team’s 900, his score will end up pretty small unless his percentage score was really high). This number is then converted from percentage added to wins added based on the average percentage correlating with a win since 2007. 14 minutes ago, KellChippy said: So this is kinda like production per touch? Impressive to see Thomas and Adams on there for wrs considering they get force fed the ball. That’s half of it. It’s the production per touch weighted by the volume of touches. Link to comment Share on other sites More sharing options...
LETSGOBROWNIES Posted February 23, 2019 Share Posted February 23, 2019 53 minutes ago, buno67 said: Two things? How is Chubb in this and not Baker? what is Chubbs Rushing WAR The first half under Hue killed Baker. Link to comment Share on other sites More sharing options...
candyman93 Posted February 23, 2019 Share Posted February 23, 2019 1 hour ago, buno67 said: Two things? How is Chubb in this and not Baker? what is Chubbs Rushing WAR Hue Jackson was a turdsuckbar Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.