That is exactly why Arizona is where they are and Buffalo is where they are. My current rank calculation is SQRT(Winner's Rank)/Loser's Rank + 1. This gives it an exponential quality to the modifier. I'm open to suggestions on better ways to quantify it as this was the formula that had no statistical reasoning behind it, just sort of felt right, if ya dig? In my opinion the team you play is probably the most important part of the calculation but maybe it is too highly weighted. For example, Arizona's win was worth 2.04 points for winning by 24 and 3.35 points for beating the #2 ranked team as the number 22 team (these are multiplied by 1.72 for winning at home)
That doesn't make much sense to me as the Jets beat two teams ranked ranked way above them that both have the same record (2-2). edit: oops - the cards are 3-2 now.
Our one win against the Dolphins was essentially negligible because it was week 1 and no teams were ranked and it was by a slim margin. Our "Credit" came from winning on the road.
The idea would be along the lines of was Buffalo that bad in their last loss to Arizona or was JP Losman that bad that he lost it for Buffalo? In other words should a bad loss with a big injury factored in count as much as a bad loss with a fully healthy team. Obviously not but how to factor that in is the difficulty.
Week 6 Power Rankings Code: Rank Team Sum Win Loss % H-W H-L H-% R-W R-L R-% PF PA PD S.O.S. 1 CLE 12.31 2 3 0.400 1 2 0.333 1 1 0.500 16.2 18.4 -2.2 0.540 2 ARI 11.22 4 2 0.667 3 0 1.000 1 2 0.333 29.5 24.0 5.5 0.483 3 TEN 10.77 5 0 1.000 3 0 1.000 2 0 1.000 23.0 11.2 11.8 0.407 4 TB 9.25 4 2 0.667 3 0 1.000 1 2 0.333 23.5 16.2 7.3 0.489 5 CHI 8.54 3 3 0.500 1 1 0.500 2 2 0.500 24.7 18.2 6.5 0.478 6 DAL 8.08 4 2 0.667 2 1 0.667 2 1 0.667 29.2 23.5 5.7 0.506 7 SD 6.75 3 3 0.500 2 1 0.667 1 2 0.333 29.7 23.2 6.5 0.535 8 PHI 6.19 3 3 0.500 2 1 0.667 1 2 0.333 27.8 20.5 7.3 0.528 9 NO 5.98 3 3 0.500 3 1 0.750 0 2 0.000 28.7 22.2 6.5 0.516 10 MIA 4.83 2 3 0.400 1 1 0.500 1 2 0.333 21.4 20.6 0.8 0.476 11 WAS 4.65 4 2 0.667 2 1 0.667 2 1 0.667 21.0 19.5 1.5 0.466 12 PIT 3.98 4 1 0.800 2 0 1.000 2 1 0.667 20.6 15.8 4.8 0.471 13 IND 3.19 3 2 0.600 1 2 0.333 2 0 1.000 22.8 19.4 3.4 0.471 14 CAR 2.65 4 2 0.667 3 0 1.000 1 2 0.333 19.5 16.2 3.3 0.522 15 NYJ 1.74 3 2 0.600 2 1 0.667 1 1 0.500 28.2 26.0 2.2 0.471 16 GB 1.30 3 3 0.500 1 2 0.333 2 1 0.667 26.7 24.2 2.5 0.489 17 ATL 0.89 4 2 0.667 3 0 1.000 1 2 0.333 23.2 21.2 2.0 0.478 18 DEN 0.25 4 2 0.667 3 1 0.750 1 1 0.500 27.7 25.7 2.0 0.483 19 JAX -0.17 3 3 0.500 1 2 0.333 2 1 0.667 20.7 21.3 -0.7 0.506 20 MIN -0.44 3 3 0.500 2 1 0.667 1 2 0.333 18.8 19.8 -1.0 0.522 21 BUF -2.71 4 1 0.800 2 0 1.000 2 1 0.667 25.2 20.8 4.4 0.447 22 BAL -2.79 2 3 0.400 2 1 0.667 0 2 0.000 15.6 17.4 -1.8 0.488 23 NYG -3.81 4 1 0.800 3 0 1.000 1 1 0.500 28.2 16.8 11.4 0.495 24 NE -5.31 3 2 0.600 1 1 0.500 2 1 0.667 17.8 21.8 -4.0 0.500 25 SF -6.75 2 4 0.333 1 3 0.250 1 1 0.500 23.5 27.8 -4.3 0.488 26 HOU -7.05 1 4 0.200 1 1 0.500 0 3 0.000 22.4 31.6 -9.2 0.488 27 SEA -8.50 1 4 0.200 1 2 0.333 0 2 0.000 20.0 30.2 -10.2 0.539 28 OAK -9.27 1 4 0.200 0 2 0.000 1 2 0.333 16.2 27.0 -10.8 0.523 29 KC -9.61 1 4 0.200 1 1 0.500 0 3 0.000 13.0 26.2 -13.2 0.539 30 CIN -10.91 0 6 0.000 0 2 0.000 0 4 0.000 14.7 24.0 -9.3 0.560 31 DET -16.89 0 5 0.000 0 2 0.000 0 3 0.000 15.2 31.8 -16.6 0.548 32 STL -18.36 1 4 0.200 0 2 0.000 1 2 0.333 12.4 32.8 -20.4 0.539 Notes: ? Ranks are calculated based on Points Margin, Home/Away and Strength of Opponent for each game and summed over the season. ? Ranks are based on an absolute sum not averaged over games played (teams with less games played will be at a disadvantage). ? The rank differential portion of the calculation is based on the prior weeks ranks. Early season game rank multipliers might be significantly different from their late season counterparts ? Win Loss, Home/Away, PPG and Stength of Schedule are shown as reference only and are not explicitly used in the calculations.
Another week another outlier… Cleveland is obviously not the number 1 team in the league and it is now a certainty that I have to tweak the rank modifier. Arizona managed to back up their status much the same as Miami did when they jumped up and its possible Cleveland does the same and we all point to this victory as the moment their season changed. However, that is unlikely. Just by beating the then #1 team doesn’t mean you get to take their place. Similarly, because you had one loss to a bad team doesn’t mean you are one of the ten worst teams in the league. Was Cleveland’s victory 60% as impressive as Arizona’s shellacking of Buffalo? Maybe but that doesn’t mean Arizona’s victory was properly weighted. Again, I am open to suggestions on how to determine how important a victory is based on the rank of the two teams playing or citations of how other rankings have dealt with it. As of now, I expect to tweak it and restate my rankings beginning this week.
Rambo I have a feeling as the season gets closer to a close your formula will be pretty much dead on. But in the early portions of the season, it is off simply because there isnt enough data to run with. As you obtain more data say 10 games per team, my bet is your rankings become nearly dead on.
Well that was always kinda the plan in the back of my head. Volatile up front and better on the back end once both the teams stabilize (losing teams tend maintain that losing mentality and vice versa). However, I think it is still too off but maybe I'm wrong. I'm going to tinker a little this week and maybe even come up with multiple strategies for dealing with things going forward. I appreciate the enthusiasm and confidence. It is nice to hear.
Amended and Restated Week 6 Rankings Code: Rank Team Sum Win Loss % H-W H-L H-% R-W R-L R-% PF PA PD S.O.S. 1 TEN 10.36 5 0 1.000 3 0 1.000 2 0 1.000 23.0 11.2 11.8 0.407 2 CHI 8.39 3 3 0.500 1 1 0.500 2 2 0.500 24.7 18.2 6.5 0.478 3 TB 8.19 4 2 0.667 3 0 1.000 1 2 0.333 23.5 16.2 7.3 0.489 4 ARI 8.01 4 2 0.667 3 0 1.000 1 2 0.333 29.5 24.0 5.5 0.483 5 DAL 7.62 4 2 0.667 2 1 0.667 2 1 0.667 29.2 23.5 5.7 0.506 6 SD 6.39 3 3 0.500 2 1 0.667 1 2 0.333 29.7 23.2 6.5 0.535 7 PHI 6.12 3 3 0.500 2 1 0.667 1 2 0.333 27.8 20.5 7.3 0.528 8 NO 5.61 3 3 0.500 3 1 0.750 0 2 0.000 28.7 22.2 6.5 0.516 9 NYG 4.04 4 1 0.800 3 0 1.000 1 1 0.500 28.2 16.8 11.4 0.495 10 CLE 4.00 2 3 0.400 1 2 0.333 1 1 0.500 16.2 18.4 -2.2 0.540 11 PIT 3.98 4 1 0.800 2 0 1.000 2 1 0.667 20.6 15.8 4.8 0.471 12 WAS 3.46 4 2 0.667 2 1 0.667 2 1 0.667 21.0 19.5 1.5 0.466 13 MIA 3.20 2 3 0.400 1 1 0.500 1 2 0.333 21.4 20.6 0.8 0.476 14 CAR 2.54 4 2 0.667 3 0 1.000 1 2 0.333 19.5 16.2 3.3 0.522 15 IND 2.29 3 2 0.600 1 2 0.333 2 0 1.000 22.8 19.4 3.4 0.471 16 GB 1.95 3 3 0.500 1 2 0.333 2 1 0.667 26.7 24.2 2.5 0.489 17 DEN 1.78 4 2 0.667 3 1 0.750 1 1 0.500 27.7 25.7 2.0 0.483 18 NYJ 1.70 3 2 0.600 2 1 0.667 1 1 0.500 28.2 26.0 2.2 0.471 19 ATL 1.37 4 2 0.667 3 0 1.000 1 2 0.333 23.2 21.2 2.0 0.478 20 BUF 1.32 4 1 0.800 2 0 1.000 2 1 0.667 25.2 20.8 4.4 0.447 21 JAX -0.31 3 3 0.500 1 2 0.333 2 1 0.667 20.7 21.3 -0.7 0.506 22 MIN -0.71 3 3 0.500 2 1 0.667 1 2 0.333 18.8 19.8 -1.0 0.522 23 BAL -2.09 2 3 0.400 2 1 0.667 0 2 0.000 15.6 17.4 -1.8 0.488 24 NE -4.14 3 2 0.600 1 1 0.500 2 1 0.667 17.8 21.8 -4.0 0.500 25 SF -6.24 2 4 0.333 1 3 0.250 1 1 0.500 23.5 27.8 -4.3 0.488 26 HOU -6.92 1 4 0.200 1 1 0.500 0 3 0.000 22.4 31.6 -9.2 0.488 27 SEA -8.36 1 4 0.200 1 2 0.333 0 2 0.000 20.0 30.2 -10.2 0.539 28 OAK -9.28 1 4 0.200 0 2 0.000 1 2 0.333 16.2 27.0 -10.8 0.523 29 KC -10.05 1 4 0.200 1 1 0.500 0 3 0.000 13.0 26.2 -13.2 0.539 30 CIN -10.27 0 6 0.000 0 2 0.000 0 4 0.000 14.7 24.0 -9.3 0.560 31 DET -16.04 0 5 0.000 0 2 0.000 0 3 0.000 15.2 31.8 -16.6 0.548 32 STL -17.89 1 4 0.200 0 2 0.000 1 2 0.333 12.4 32.8 -20.4 0.539
I essentially cut the points gained based on the rank of your opponent in half and I think it is a much better representation. 1. Tennessee is the clear cut #1 team until they lose. 2. Chicago is very strong. Their losses have been by 3, 3, and 2 points. 3. TB, as much as I hate to admit, is equally strong having lost by 4 and 3 points and playing only top 20 teams. They beat Chi but are hurt because they have 3 of 4 wins at home. 4. Arizona, another surging team with 2 big weeks. The Dallas game actually wasn't that impressive in my calculations as they won by a small margin at home. The Buffalo shellacking accounts for much of their ranking 5. Dallas is sliding with only 1 decent win over GB and a couple close losses. 6. San Diego, boosted by their last win they have been hurt by the level of competition they played haven't been dinged too much on their losses. 7. Philadelphia, when they win they tend to win big and when they lose they tend to lose by very little against good opponents. 8. New Orleans, their points have come almost exclusively from their last win over Oakland and show the most potential to drop. On the other hand they have some injuries that may determine their success one way or the other. 9. New York Giants, would be number 1 had they not got embarrassed on Monday Night Football 10. Cleveland, gained two slots from not playing week 5(!) and 16 from beating the Giants. Two game win streak has been solid and they look to be returning to form. If they keep it up they will justify a top ten ranking and if they don't it is back to the cellar.
Week 7 Rankings Code: Rank Team Sum Win Loss % H-W H-L H-% R-W R-L R-% PF PA PD S.O.S. 1 TEN 15.43 6 0 1.000 3 0 1.000 3 0 1.000 24.8 11.0 13.8 0.410 2 PIT 10.32 5 1 0.833 2 0 1.000 3 1 0.750 23.5 14.8 8.7 0.475 3 CHI 9.55 4 3 0.571 2 1 0.667 2 2 0.500 28.0 21.4 6.6 0.485 4 TB 9.06 5 2 0.714 4 0 1.000 1 2 0.333 23.0 15.3 7.7 0.472 5 ARI 7.47 4 2 0.667 3 0 1.000 1 2 0.333 29.5 24.0 5.5 0.480 6 NYG 6.03 5 1 0.833 4 0 1.000 1 1 0.500 28.3 16.8 11.5 0.490 7 CAR 6.00 5 2 0.714 4 0 1.000 1 2 0.333 21.0 14.9 6.1 0.514 8 PHI 5.87 3 3 0.500 2 1 0.667 1 2 0.333 27.8 20.5 7.3 0.534 9 GB 4.95 4 3 0.571 2 2 0.500 2 1 0.667 27.7 22.7 5.0 0.481 10 SD 4.49 3 4 0.429 2 1 0.667 1 3 0.250 27.4 23.1 4.3 0.525 11 WAS 3.82 5 2 0.714 3 1 0.750 2 1 0.667 20.0 18.3 1.7 0.455 12 BUF 3.31 5 1 0.833 3 0 1.000 2 1 0.667 24.8 19.7 5.2 0.424 13 DAL 3.22 4 3 0.571 2 1 0.667 2 2 0.500 27.0 25.0 2.0 0.529 14 CLE 2.95 2 4 0.333 1 2 0.333 1 2 0.333 15.3 17.7 -2.3 0.554 15 BAL 1.65 3 3 0.500 2 1 0.667 1 2 0.333 17.5 16.7 0.8 0.490 16 NO 1.43 3 4 0.429 3 1 0.750 0 3 0.000 25.6 23.3 2.3 0.523 17 ATL 1.39 4 2 0.667 3 0 1.000 1 2 0.333 23.2 21.2 2.0 0.486 18 NYJ 1.04 3 3 0.500 2 1 0.667 1 2 0.333 25.7 24.3 1.3 0.470 19 NE 0.93 4 2 0.667 2 1 0.667 2 1 0.667 21.7 19.3 2.3 0.475 20 MIA -0.09 2 4 0.333 1 2 0.333 1 2 0.333 20.0 21.7 -1.7 0.485 21 JAX -0.24 3 3 0.500 1 2 0.333 2 1 0.667 20.7 21.3 -0.7 0.515 22 IND -0.98 3 3 0.500 1 2 0.333 2 1 0.667 21.3 21.8 -0.5 0.495 23 MIN -1.60 3 4 0.429 2 1 0.667 1 3 0.250 22.0 23.9 -1.9 0.544 24 DEN -2.97 4 3 0.571 3 1 0.750 1 2 0.333 24.7 27.9 -3.1 0.475 25 HOU -5.53 2 4 0.333 2 1 0.667 0 3 0.000 23.3 29.8 -6.5 0.490 26 SF -8.11 2 5 0.286 1 3 0.250 1 2 0.333 22.6 28.0 -5.4 0.485 27 OAK -8.62 2 4 0.333 1 2 0.333 1 2 0.333 16.2 24.7 -8.5 0.505 28 SEA -9.39 1 5 0.167 1 2 0.333 0 3 0.000 18.3 28.5 -10.2 0.549 29 STL -12.90 2 4 0.333 1 2 0.333 1 2 0.333 16.0 29.7 -13.7 0.525 30 KC -15.04 1 5 0.167 1 2 0.333 0 3 0.000 12.5 27.5 -15.0 0.529 31 CIN -16.51 0 7 0.000 0 3 0.000 0 4 0.000 14.0 26.0 -12.0 0.561 32 DET -16.95 0 6 0.000 0 2 0.000 0 4 0.000 16.2 31.2 -15.0 0.561
If you did a simple list based on point differential, I think the rankings would be more intuitive. Note how NYG and Buffalo get bumped up while Chicago and Arizona get knocked down. I think most fans would agree with that. And NE gets a big raise (which they deserve after last night) and Cleveland gets dumped. All of that makes sense. new_rank old_Rank Team PD 1 1 TEN 13.8 2 6 NYG 11.5 3 2 PIT 8.7 4 4 TB 7.7 5 8 PHI 7.3 6 3 CHI 6.6 7 7 CAR 6.1 8 5 ARI 5.5 9 12 BUF 5.2 10 9 GB 5 11 10 SD 4.3 12 19 NE 2.3 13 16 NO 2.3 14 17 ATL 2 15 13 DAL 2 16 11 WAS 1.7 17 18 NYJ 1.3 18 15 BAL 0.8 19 22 IND -0.5 20 21 JAX -0.7 21 20 MIA -1.7 22 23 MIN -1.9 23 14 CLE -2.3 24 24 DEN -3.1 25 26 SF -5.4 26 25 HOU -6.5 27 27 OAK -8.5 28 28 SEA -10.2 29 31 CIN -12 30 29 STL -13.7 31 30 KC -15 32 32 DET -15
I disagree. You have to be cognizant as to where those points were scored and how. For example, New England has had one good game pointswise (and that was against a team missing their starting QB and RB). Their weekly differentials have been 7, 9, -25, 9, -20, 34. The average team wins by 12.4 points each game. I would consider last nights win an aberration at this point, an outlier that I am seeking to average out rather than let it significantly influence the rankings. On the other hand, the Giants have won two games by 28 and 38 points which are the reasons their point differential is so high. Who did they beat by those wide margins? #29 St. Louis and #28 Seattle. Tampa Bay has won by more than average against the #7 and #17 teams and has only lost by a combined 7 points (NYG lost by 3 times as much in their 1 loss!).
