Power = 50 + W+ 0.43*HW + 0.57*RW + TO/2 + PM/G + W%
50 is an arbitrary number used so that I do not obtain negative values (they're a pain to work with).
W = wins which is set to a value of 5
HW = home wins - weighted to 0.43 because home teams win 57% on average over the last 10 years; they receive less points for having a higher average of winning.
RW = road wins - weighted to 0.57 because a team that wins on the road is better than a team that wins at home.
TO = turnover margin - it's divided by 2 to lessen its effect on total power (i.e. it's not weighted the same as point margin)
PM/G = point margin per game - denotes the average points scored minus points allowed per game
W% = win percentage - values were divided by ten to make results more-or-less equivalent to the rest of the variables (there is not an over inflation caused by W% just a point value - 100% = 10 points).
| 1 | GB | 92.63 |
| 2 | SF | 83.97 |
| 3 | DET | 77.86 |
| 4 | HOU | 76.75 |
| 5 | NO | 76.54 |
| 6 | BAL | 76.53 |
| 7 | NE | 74.60 |
| 8 | BUF | 73.86 |
| 9 | ATL | 72.57 |
| 10 | NYJ | 71.59 |
| 11 | CHI | 69.21 |
| 12 | PIT | 68.10 |
| 13 | DAL | 65.90 |
| 14 | NYG | 65.83 |
| 15 | TB | 64.93 |
| 16 | CLE | 61.56 |
| 17 | OAK | 59.14 |
| 18 | CIN | 58.83 |
| 19 | CAR | 58.71 |
| 20 | SD | 58.03 |
| 21 | KC | 57.48 |
| 22 | WAS | 57.10 |
| 23 | TEN | 56.10 |
| 24 | MIN | 53.50 |
| 25 | DEN | 52.22 |
| 26 | JAX | 50.60 |
| 27 | SEA | 49.67 |
| 28 | ARI | 44.60 |
| 29 | PHI | 42.47 |
| 30 | MIA | 37.17 |
| 31 | STL | 34.67 |
| 32 | IND | 24.00 |
cool power rankings dude, if you look at ESPN's power ranking system it's pretty similar, its just cool how you show those same standings mathematically. It's interesting how your formula puts New England 7th, making them the second worst 1 loss team in front of San Diego. I only say that because they seem like the best team period behind Green Bay, its interesting how the math does not support that.
ReplyDeletealso, I like how you differentiate wins between road and home games. I wonder if you could even expound on that more and differentiate in between road games as distance seems to be a much bigger factor in the NFL than in any other professional sport. For example, it's a more impressive road win to see the Niner's win in New York than if they win in Oakland.
ReplyDeleteYea i found that interesting too. Im hoping to create actual statistics for each variable and how significant they are in the next couple of weeks. That should give me precise percentages to use rather than arbitrarily choosing weights. Nice! thats an awesome idea. This week im going to incorporate QBR and try to figure out distance traveled also. Thanks for the idea kyle!! =D
ReplyDelete